syntatic_preocessing.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xwi9mzFPPDv7"
      },
      "source": [
        "# Identifying Entities in Healthcare Data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3wwzoMzrUDtr"
      },
      "source": [
        "##Workspace set up: Import and Install useful packages."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "T4hxafMCE1nn"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: pycrf in c:\\users\\moham\\anaconda3\\lib\\site-packages (0.0.1)\n",
            "Requirement already satisfied: sklearn-crfsuite in c:\\users\\moham\\anaconda3\\lib\\site-packages (0.3.6)\n",
            "Requirement already satisfied: python-crfsuite>=0.8.3 in c:\\users\\moham\\anaconda3\\lib\\site-packages (from sklearn-crfsuite) (0.9.9)\n",
            "Requirement already satisfied: tqdm>=2.0 in c:\\users\\moham\\anaconda3\\lib\\site-packages (from sklearn-crfsuite) (4.64.1)\n",
            "Requirement already satisfied: six in c:\\users\\moham\\anaconda3\\lib\\site-packages (from sklearn-crfsuite) (1.16.0)\n",
            "Requirement already satisfied: tabulate in c:\\users\\moham\\anaconda3\\lib\\site-packages (from sklearn-crfsuite) (0.8.10)\n",
            "Requirement already satisfied: colorama in c:\\users\\moham\\anaconda3\\lib\\site-packages (from tqdm>=2.0->sklearn-crfsuite) (0.4.6)\n"
          ]
        }
      ],
      "source": [
        "!pip install pycrf\n",
        "!pip install sklearn-crfsuite\n",
        "\n",
        "import spacy\n",
        "import sklearn_crfsuite\n",
        "from sklearn_crfsuite import metrics\n",
        "import pandas as pd\n",
        "\n",
        "model = spacy.load(\"en_core_web_sm\")\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bE9Hwb3AUHwG"
      },
      "source": [
        "##Data Preprocessing\n",
        "\n",
        "The dataset provided is in the form of one word per line. Let's understand the format of data below:\n",
        "- Suppose there are *x* words in a sentence, then there will be *x* continuous lines with one word in each line. \n",
        "- Further, the two sentences are separated by empty lines. The labels for the data follow the same format.\n",
        "\n",
        "**We need to pre-process the data to recover the complete sentences and their labels.**\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nBU_BNUuK8tm"
      },
      "source": [
        "### Construct the proper sentences from individual words and print the 5 sentences."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "b4OjXEhjhkCW"
      },
      "source": [
        "![image.png](data:image/png;base64,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)\n",
        "\n",
        "![image.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "bz6czVx9R6WJ"
      },
      "outputs": [],
      "source": [
        "#Initializing the train and test sentences and labels\n",
        "with open('train_sent', 'r') as train_sent_file:\n",
        "  train_words = train_sent_file.readlines()\n",
        "\n",
        "with open('train_label', 'r') as train_labels_file:\n",
        "  train_labels_by_word = train_labels_file.readlines()\n",
        "\n",
        "with open('test_sent', 'r') as test_sent_file:\n",
        "  test_words = test_sent_file.readlines()\n",
        "\n",
        "with open('test_label', 'r') as test_labels_file:\n",
        "  test_labels_by_word = test_labels_file.readlines()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "TE31q8yxVdU-"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Number of tokens in training set \n",
            "No. of words:  48501 \n",
            "No. of labels:  48501\n",
            "\n",
            "Number of tokens in test set \n",
            "No. of words:  19674 \n",
            "No. of labels:  19674\n"
          ]
        }
      ],
      "source": [
        "#Checking to see that the number of tokens and number of corresponding labels are same.\n",
        "print(\"Number of tokens in training set\",\"\\nNo. of words: \",len(train_words),\"\\nNo. of labels: \",len(train_labels_by_word))\n",
        "print(\"\\nNumber of tokens in test set\",\"\\nNo. of words: \",len(test_words),\"\\nNo. of labels: \",len(test_labels_by_word))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "k4gdPBHwLdNX"
      },
      "outputs": [],
      "source": [
        "#Combining tokens belonging to the same sentence. Sentences are separated by \"\\n\" in the dataset.\n",
        "def convert_to_sentences(dataset):\n",
        "    sent_list = []\n",
        "    sent = \"\"\n",
        "    for entity in dataset:\n",
        "        if entity != '\\n':\n",
        "            sent = sent + entity[:-1] + \" \"       \n",
        "        else: \n",
        "            sent_list.append(sent[:-1])           \n",
        "            sent = \"\"\n",
        "    return sent_list"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {},
      "outputs": [],
      "source": [
        "#Converting tokens to sentences and individual labels to sequences of corresponding labels.\n",
        "train_sentences = convert_to_sentences(train_words)\n",
        "train_labels = convert_to_sentences(train_labels_by_word)\n",
        "test_sentences = convert_to_sentences(test_words)\n",
        "test_labels = convert_to_sentences(test_labels_by_word)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "First five training sentences and their labels:\n",
            "\n",
            "All live births > or = 23 weeks at the University of Vermont in 1995 ( n = 2395 ) were retrospectively analyzed for delivery route , indication for cesarean , gestational age , parity , and practice group ( to reflect risk status ) \n",
            " O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O \n",
            "\n",
            "The total cesarean rate was 14.4 % ( 344 of 2395 ) , and the primary rate was 11.4 % ( 244 of 2144 ) \n",
            " O O O O O O O O O O O O O O O O O O O O O O O O O \n",
            "\n",
            "Abnormal presentation was the most common indication ( 25.6 % , 88 of 344 ) \n",
            " O O O O O O O O O O O O O O O \n",
            "\n",
            "The `` corrected '' cesarean rate ( maternal-fetal medicine and transported patients excluded ) was 12.4 % ( 273 of 2194 ) , and the `` corrected '' primary rate was 9.6 % ( 190 of 1975 ) \n",
            " O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O \n",
            "\n",
            "Arrest of dilation was the most common indication in both `` corrected '' subgroups ( 23.4 and 24.6 % , respectively ) \n",
            " O O O O O O O O O O O O O O O O O O O O O O \n",
            "\n"
          ]
        }
      ],
      "source": [
        "print(\"First five training sentences and their labels:\\n\")\n",
        "for i in range(5):\n",
        "    print(train_sentences[i],\"\\n\",train_labels[i],\"\\n\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "First five test sentences and their labels:\n",
            "\n",
            "Furthermore , when all deliveries were analyzed , regardless of risk status but limited to gestational age > or = 36 weeks , the rates did not change ( 12.6 % , 280 of 2214 ; primary 9.2 % , 183 of 1994 ) \n",
            " O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O \n",
            "\n",
            "As the ambient temperature increases , there is an increase in insensible fluid loss and the potential for dehydration \n",
            " O O O O O O O O O O O O O O O O O O O \n",
            "\n",
            "The daily high temperature ranged from 71 to 104 degrees F and AFI values ranged from 1.7 to 24.7 cm during the study period \n",
            " O O O O O O O O O O O O O O O O O O O O O O O O \n",
            "\n",
            "There was a significant correlation between the 2- , 3- , and 4-day mean temperature and AFI , with the 4-day mean being the most significant ( r = 0.31 , p & # 60 ; 0.001 ) \n",
            " O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O \n",
            "\n",
            "Fluctuations in ambient temperature are inversely correlated to changes in AFI \n",
            " O O O O O O O O O O O \n",
            "\n"
          ]
        }
      ],
      "source": [
        "print(\"First five test sentences and their labels:\\n\")\n",
        "for i in range(5):\n",
        "    print(test_sentences[i],\"\\n\",test_labels[i],\"\\n\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "urCxIOBTMBwG"
      },
      "source": [
        "### Count the number of sentences in the processed train and test dataset "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "zwa18H_IMFXV"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Number of sentences in the train dataset: 2599\n",
            "Number of sentences in the test dataset: 1056\n"
          ]
        }
      ],
      "source": [
        "print(\"Number of sentences in the train dataset: {}\".format(len(train_sentences)))\n",
        "print(\"Number of sentences in the test dataset: {}\".format(len(test_sentences)))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Jjg2DK7hLcO_"
      },
      "source": [
        "### Count the number of lines of labels in the processed train and test dataset."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "unXoD2bWLPZD"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Number of lines of labels in the train dataset: 2599\n",
            "Number of lines of labels in the test dataset: 1056\n"
          ]
        }
      ],
      "source": [
        "print(\"Number of lines of labels in the train dataset: {}\".format(len(train_labels)))\n",
        "print(\"Number of lines of labels in the test dataset: {}\".format(len(test_labels)))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "r9Hy2tECbd9a"
      },
      "source": [
        "##Concept Identification\n",
        "\n",
        "We will first explore what are the various concepts present in the dataset. For this, we will use PoS Tagging. \n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3YGpfOK7Mk3X"
      },
      "source": [
        "### Extract those tokens which have NOUN or PROPN as their PoS tag and find their frequency"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "7w8nJSPlZqjw"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Number of sentences in combined dataset (training + test): 3655\n"
          ]
        }
      ],
      "source": [
        "#Creating a combined dataset from training and test sentences, since this is an Exploratory analysis.\n",
        "combined = train_sentences + test_sentences\n",
        "print(\"Number of sentences in combined dataset (training + test): {}\".format(len(combined)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "No. of tokens in combined dataset with PoS tag of 'NOUN' or 'PROPN': 24376\n"
          ]
        }
      ],
      "source": [
        "#Creating a list of tokens which have PoS tag of 'NOUN' or 'PROPN'\n",
        "noun_propn = []         \n",
        "pos_tag = []            \n",
        "for sent in combined:\n",
        "    for token in model(sent):\n",
        "        if token.pos_ in ['NOUN', 'PROPN']:\n",
        "           noun_propn.append(token.text)\n",
        "           pos_tag.append(token.pos_)\n",
        "print(\"No. of tokens in combined dataset with PoS tag of 'NOUN' or 'PROPN': {}\".format(len(noun_propn)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "24376\n"
          ]
        }
      ],
      "source": [
        "print(len(pos_tag))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-SQG8wRxMo6A"
      },
      "source": [
        "### Print the top 25 most common tokens with NOUN or PROPN PoS tags"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "id": "DhDY4jjoawTa"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Top 25 comon tokens with PoS tag of 'NOUN' or 'PROPN' \n",
            "\n",
            "patients        492\n",
            "treatment       281\n",
            "%               247\n",
            "cancer          200\n",
            "therapy         175\n",
            "study           154\n",
            "disease         142\n",
            "cell            140\n",
            "lung            116\n",
            "group            94\n",
            "chemotherapy     88\n",
            "gene             87\n",
            "effects          85\n",
            "results          79\n",
            "women            77\n",
            "use              74\n",
            "TO_SEE           74\n",
            "surgery          71\n",
            "cases            71\n",
            "risk             71\n",
            "analysis         70\n",
            "rate             67\n",
            "response         66\n",
            "survival         65\n",
            "children         64\n",
            "Name: NOUN_PROPN, dtype: int64\n"
          ]
        }
      ],
      "source": [
        "noun_pos = pd.DataFrame({\"NOUN_PROPN\":noun_propn,\"POS_tag\":pos_tag})\n",
        "print(\"Top 25 comon tokens with PoS tag of 'NOUN' or 'PROPN' \\n\")\n",
        "print(noun_pos[\"NOUN_PROPN\"].value_counts().head(25))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bnfAlKgA5fd4"
      },
      "source": [
        "## Defining features for CRF\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "aS6P1Gog3JMf"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "The total cesarean rate was 14.4 % ( 344 of 2395 ) , and the primary rate was 11.4 % ( 244 of 2144 )\n",
            "\n",
            "PoS tag of word in isolation\n",
            "Word: cesarean -- VERB \n",
            "\n",
            "PoS tag of all words in sentence with context in tact.\n",
            "The -- DET\n",
            "total -- ADJ\n",
            "cesarean -- ADJ\n",
            "rate -- NOUN\n",
            "was -- AUX\n",
            "14.4 -- NUM\n",
            "% -- NOUN\n",
            "( -- PUNCT\n",
            "344 -- NUM\n",
            "of -- ADP\n",
            "2395 -- NUM\n",
            ") -- PUNCT\n",
            ", -- PUNCT\n",
            "and -- CCONJ\n",
            "the -- DET\n",
            "primary -- ADJ\n",
            "rate -- NOUN\n",
            "was -- AUX\n",
            "11.4 -- NUM\n",
            "% -- NOUN\n",
            "( -- PUNCT\n",
            "244 -- NUM\n",
            "of -- ADP\n",
            "2144 -- NUM\n",
            ") -- PUNCT\n",
            "\n",
            "Result of modified workflow to obtain PoS tag of word at a specific position while keeping context within sentence in-tact.\n",
            "Word: cesarean POSTAG: ADJ\n"
          ]
        }
      ],
      "source": [
        "# Let's define the features to get the feature value for one word and do analysis of PoS tags.\n",
        "sentence = train_sentences[1]   \n",
        "sent_list = sentence.split()      # Splitting the sentence into its constituent words.\n",
        "position = 2                      # Choosing position of word within sentence. Index starts at 0.\n",
        "\n",
        "word = sent_list[position]        # Extracting word for PoS tag analysis.\n",
        "\n",
        "print(sentence)\n",
        "\n",
        "\n",
        "print(\"\\nPoS tag of word in isolation\\nWord:\",word,\"--\",model(word)[0].pos_,\"\\n\")\n",
        "\n",
        "print(\"PoS tag of all words in sentence with context in tact.\")\n",
        "for token in model(sentence):\n",
        "    print(token.text, \"--\", token.pos_)\n",
        "\n",
        "# Modified workflow to obtain PoS tag of specific word in question while keeping sentence context in tact.\n",
        "print(\"\\nResult of modified workflow to obtain PoS tag of word at a specific position while keeping context within sentence in-tact.\")\n",
        "cnt = 0                           # Count of the word position within sentence.\n",
        "for token in model(sentence):\n",
        "      postag = token.pos_\n",
        "      if (token.text == word) and (cnt == position):\n",
        "          break\n",
        "      cnt += 1\n",
        "print(\"Word:\", word,\"POSTAG:\",postag)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {},
      "outputs": [],
      "source": [
        "# Function to obtain contextual PoS tagger.\n",
        "def contextual_pos_tagger(sent_list,position):\n",
        "    '''Obtaining PoS tag for individual word with sentence context in-tact. \n",
        "       If the PoS tag is obtained for a word individually, it may not capture the context of use in the sentence and may assign the incorrect PoS tag.'''\n",
        "\n",
        "    sentence = \" \".join(sent_list)          \n",
        "    posit = 0                               \n",
        "    for token in model(sentence):\n",
        "        postag = token.pos_\n",
        "        if (token.text == word) and (posit == position):\n",
        "            break\n",
        "        posit += 1\n",
        "    return postag"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {},
      "outputs": [],
      "source": [
        "# Define the features to get the feature values for one word.\n",
        "def getFeaturesForOneWord(sent_list, position):\n",
        "  word = sent_list[position]\n",
        "    \n",
        "  # Obtaining features for current word\n",
        "  features = [\n",
        "    'word.lower=' + word.lower(),                                   \n",
        "    'word.postag=' + contextual_pos_tagger(sent_list, position),    \n",
        "    'word[-3:]=' + word[-3:],                                      \n",
        "    'word[-2:]=' + word[-2:],                                       \n",
        "    'word.isupper=%s' % word.isupper(),                            \n",
        "    'word.isdigit=%s' % word.isdigit(),                            \n",
        "    'words.startsWithCapital=%s' % word[0].isupper()                \n",
        "  ]\n",
        " \n",
        "  if(position > 0):\n",
        "    prev_word = sent_list[position-1]\n",
        "    features.extend([\n",
        "    'prev_word.lower=' + prev_word.lower(),                               \n",
        "    'prev_word.postag=' + contextual_pos_tagger(sent_list, position - 1), \n",
        "    'prev_word.isupper=%s' % prev_word.isupper(),                         \n",
        "    'prev_word.isdigit=%s' % prev_word.isdigit(),                         \n",
        "    'prev_words.startsWithCapital=%s' % prev_word[0].isupper()            \n",
        "  ])\n",
        "  else:\n",
        "    features.append('BEG')                                                \n",
        " \n",
        "  if(position == len(sent_list)-1):\n",
        "    features.append('END')                                               \n",
        " \n",
        "  return features\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "haBEOsZGKhb9"
      },
      "source": [
        "## Getting the features"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Fwb8YUdlNcG1"
      },
      "source": [
        "### Write a code/function to get the features for a sentence"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "id": "emBIm5uK7vtK"
      },
      "outputs": [],
      "source": [
        "def getFeaturesForOneSentence(sentence):\n",
        "  sentence_list = sentence.split()\n",
        "  return [getFeaturesForOneWord(sentence_list, position) for position in range(len(sentence_list))]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Cesarean rates at tertiary care hospitals should be compared with rates at community hospitals only after correcting for dissimilar patient groups or gestational age\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "['word.lower=cesarean',\n",
              " 'word.postag=NOUN',\n",
              " 'word[-3:]=ean',\n",
              " 'word[-2:]=an',\n",
              " 'word.isupper=False',\n",
              " 'word.isdigit=False',\n",
              " 'words.startsWithCapital=True',\n",
              " 'BEG']"
            ]
          },
          "execution_count": 18,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Checking feature extraction\n",
        "example_sentence = train_sentences[5]\n",
        "print(example_sentence)\n",
        "\n",
        "features = getFeaturesForOneSentence(example_sentence)\n",
        "features[0]\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "['word.lower=care',\n",
              " 'word.postag=NOUN',\n",
              " 'word[-3:]=are',\n",
              " 'word[-2:]=re',\n",
              " 'word.isupper=False',\n",
              " 'word.isdigit=False',\n",
              " 'words.startsWithCapital=False',\n",
              " 'prev_word.lower=tertiary',\n",
              " 'prev_word.postag=NOUN',\n",
              " 'prev_word.isupper=False',\n",
              " 'prev_word.isdigit=False',\n",
              " 'prev_words.startsWithCapital=False']"
            ]
          },
          "execution_count": 19,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "features[4]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A7fs-B8CNein"
      },
      "source": [
        "### Write a code/function to get the labels of a sentence"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "id": "tiA-A_ws8I8-"
      },
      "outputs": [],
      "source": [
        "def getLabelsInListForOneSentence(labels):\n",
        "  return labels.split()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O']\n"
          ]
        }
      ],
      "source": [
        "# Checking label extraction\n",
        "example_labels = getLabelsInListForOneSentence(train_labels[5])\n",
        "print(example_labels)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QG6hYY6rNjS8"
      },
      "source": [
        "## Define input and target variables\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DZmjKbXgN44o"
      },
      "source": [
        "Correctly computing X and Y sequence matrices for training and test data.\n",
        "Check that both sentences and labels are processed"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U1C4djFXNpqN"
      },
      "source": [
        "### Define the features' values for each sentence as input variable  for CRF model in test and the train dataset "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "id": "8Nuz-MjW8LWm"
      },
      "outputs": [],
      "source": [
        "X_train = [getFeaturesForOneSentence(sentence) for sentence in train_sentences]\n",
        "X_test = [getFeaturesForOneSentence(sentence) for sentence in test_sentences]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "q88-s6BQNr0M"
      },
      "source": [
        "### Define the labels as the target variable for test and the train dataset"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "id": "HrzLkWs0Nvx4"
      },
      "outputs": [],
      "source": [
        "Y_train = [getLabelsInListForOneSentence(labels) for labels in train_labels]\n",
        "Y_test = [getLabelsInListForOneSentence(labels) for labels in test_labels]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YrIAz8gqN_Js"
      },
      "source": [
        "## Build the CRF Model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "k2oGSiqQ8Otd"
      },
      "outputs": [
        {
          "ename": "AttributeError",
          "evalue": "'CRF' object has no attribute 'keep_tempfiles'",
          "output_type": "error",
          "traceback": [
            "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[1;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\IPython\\core\\formatters.py\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self, obj, include, exclude)\u001b[0m\n\u001b[0;32m    968\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    969\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mmethod\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 970\u001b[1;33m                 \u001b[1;32mreturn\u001b[0m \u001b[0mmethod\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minclude\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0minclude\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mexclude\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mexclude\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    971\u001b[0m             \u001b[1;32mreturn\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    972\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\base.py\u001b[0m in \u001b[0;36m_repr_mimebundle_\u001b[1;34m(self, **kwargs)\u001b[0m\n\u001b[0;32m    612\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m_repr_mimebundle_\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    613\u001b[0m         \u001b[1;34m\"\"\"Mime bundle used by jupyter kernels to display estimator\"\"\"\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 614\u001b[1;33m         \u001b[0moutput\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m{\u001b[0m\u001b[1;34m\"text/plain\"\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mrepr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m}\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    615\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mget_config\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m\"display\"\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m==\u001b[0m \u001b[1;34m\"diagram\"\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    616\u001b[0m             \u001b[0moutput\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m\"text/html\"\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mestimator_html_repr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\base.py\u001b[0m in \u001b[0;36m__repr__\u001b[1;34m(self, N_CHAR_MAX)\u001b[0m\n\u001b[0;32m    277\u001b[0m         )\n\u001b[0;32m    278\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 279\u001b[1;33m         \u001b[0mrepr_\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    280\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    281\u001b[0m         \u001b[1;31m# Use bruteforce ellipsis when there are a lot of non-blank characters\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\pprint.py\u001b[0m in \u001b[0;36mpformat\u001b[1;34m(self, object)\u001b[0m\n\u001b[0;32m    151\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mpformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mobject\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    152\u001b[0m         \u001b[0msio\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_StringIO\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 153\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_format\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobject\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msio\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m{\u001b[0m\u001b[1;33m}\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    154\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0msio\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mgetvalue\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    155\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\pprint.py\u001b[0m in \u001b[0;36m_format\u001b[1;34m(self, object, stream, indent, allowance, context, level)\u001b[0m\n\u001b[0;32m    168\u001b[0m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_readable\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    169\u001b[0m             \u001b[1;32mreturn\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 170\u001b[1;33m         \u001b[0mrep\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_repr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobject\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlevel\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    171\u001b[0m         \u001b[0mmax_width\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_width\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mindent\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mallowance\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    172\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mrep\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m>\u001b[0m \u001b[0mmax_width\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\pprint.py\u001b[0m in \u001b[0;36m_repr\u001b[1;34m(self, object, context, level)\u001b[0m\n\u001b[0;32m    429\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    430\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m_repr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mobject\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlevel\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 431\u001b[1;33m         repr, readable, recursive = self.format(object, context.copy(),\n\u001b[0m\u001b[0;32m    432\u001b[0m                                                 self._depth, level)\n\u001b[0;32m    433\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mreadable\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\utils\\_pprint.py\u001b[0m in \u001b[0;36mformat\u001b[1;34m(self, object, context, maxlevels, level)\u001b[0m\n\u001b[0;32m    187\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    188\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mobject\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmaxlevels\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlevel\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 189\u001b[1;33m         return _safe_repr(\n\u001b[0m\u001b[0;32m    190\u001b[0m             \u001b[0mobject\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmaxlevels\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlevel\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mchanged_only\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_changed_only\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    191\u001b[0m         )\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\utils\\_pprint.py\u001b[0m in \u001b[0;36m_safe_repr\u001b[1;34m(object, context, maxlevels, level, changed_only)\u001b[0m\n\u001b[0;32m    438\u001b[0m         \u001b[0mrecursive\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    439\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mchanged_only\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 440\u001b[1;33m             \u001b[0mparams\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_changed_params\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobject\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    441\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    442\u001b[0m             \u001b[0mparams\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mobject\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_params\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdeep\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\utils\\_pprint.py\u001b[0m in \u001b[0;36m_changed_params\u001b[1;34m(estimator)\u001b[0m\n\u001b[0;32m     91\u001b[0m     estimator with non-default values.\"\"\"\n\u001b[0;32m     92\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 93\u001b[1;33m     \u001b[0mparams\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mestimator\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_params\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdeep\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     94\u001b[0m     \u001b[0minit_func\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mestimator\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m__init__\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m\"deprecated_original\"\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mestimator\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m__init__\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     95\u001b[0m     \u001b[0minit_params\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0minspect\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msignature\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minit_func\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mparameters\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\base.py\u001b[0m in \u001b[0;36mget_params\u001b[1;34m(self, deep)\u001b[0m\n\u001b[0;32m    208\u001b[0m         \u001b[0mout\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mdict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    209\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mkey\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_get_param_names\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 210\u001b[1;33m             \u001b[0mvalue\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    211\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mdeep\u001b[0m \u001b[1;32mand\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvalue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m\"get_params\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    212\u001b[0m                 \u001b[0mdeep_items\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mvalue\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_params\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;31mAttributeError\u001b[0m: 'CRF' object has no attribute 'keep_tempfiles'"
          ]
        },
        {
          "ename": "AttributeError",
          "evalue": "'CRF' object has no attribute 'keep_tempfiles'",
          "output_type": "error",
          "traceback": [
            "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[1;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\IPython\\core\\formatters.py\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self, obj)\u001b[0m\n\u001b[0;32m    700\u001b[0m                 \u001b[0mtype_pprinters\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtype_printers\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    701\u001b[0m                 deferred_pprinters=self.deferred_printers)\n\u001b[1;32m--> 702\u001b[1;33m             \u001b[0mprinter\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpretty\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    703\u001b[0m             \u001b[0mprinter\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mflush\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    704\u001b[0m             \u001b[1;32mreturn\u001b[0m \u001b[0mstream\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mgetvalue\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\IPython\\lib\\pretty.py\u001b[0m in \u001b[0;36mpretty\u001b[1;34m(self, obj)\u001b[0m\n\u001b[0;32m    392\u001b[0m                         \u001b[1;32mif\u001b[0m \u001b[0mcls\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mobject\u001b[0m\u001b[0;31m \u001b[0m\u001b[0;31m\\\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    393\u001b[0m                                 \u001b[1;32mand\u001b[0m \u001b[0mcallable\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mcls\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m__dict__\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'__repr__'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 394\u001b[1;33m                             \u001b[1;32mreturn\u001b[0m \u001b[0m_repr_pprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcycle\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    395\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    396\u001b[0m             \u001b[1;32mreturn\u001b[0m \u001b[0m_default_pprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcycle\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\IPython\\lib\\pretty.py\u001b[0m in \u001b[0;36m_repr_pprint\u001b[1;34m(obj, p, cycle)\u001b[0m\n\u001b[0;32m    698\u001b[0m     \u001b[1;34m\"\"\"A pprint that just redirects to the normal repr function.\"\"\"\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    699\u001b[0m     \u001b[1;31m# Find newlines and replace them with p.break_()\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 700\u001b[1;33m     \u001b[0moutput\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mrepr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobj\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    701\u001b[0m     \u001b[0mlines\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0moutput\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msplitlines\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    702\u001b[0m     \u001b[1;32mwith\u001b[0m \u001b[0mp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mgroup\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\base.py\u001b[0m in \u001b[0;36m__repr__\u001b[1;34m(self, N_CHAR_MAX)\u001b[0m\n\u001b[0;32m    277\u001b[0m         )\n\u001b[0;32m    278\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 279\u001b[1;33m         \u001b[0mrepr_\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    280\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    281\u001b[0m         \u001b[1;31m# Use bruteforce ellipsis when there are a lot of non-blank characters\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\pprint.py\u001b[0m in \u001b[0;36mpformat\u001b[1;34m(self, object)\u001b[0m\n\u001b[0;32m    151\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mpformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mobject\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    152\u001b[0m         \u001b[0msio\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_StringIO\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 153\u001b[1;33m         \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_format\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobject\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msio\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m{\u001b[0m\u001b[1;33m}\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    154\u001b[0m         \u001b[1;32mreturn\u001b[0m \u001b[0msio\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mgetvalue\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    155\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\pprint.py\u001b[0m in \u001b[0;36m_format\u001b[1;34m(self, object, stream, indent, allowance, context, level)\u001b[0m\n\u001b[0;32m    168\u001b[0m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_readable\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    169\u001b[0m             \u001b[1;32mreturn\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 170\u001b[1;33m         \u001b[0mrep\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_repr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobject\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlevel\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    171\u001b[0m         \u001b[0mmax_width\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_width\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mindent\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mallowance\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    172\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mrep\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m>\u001b[0m \u001b[0mmax_width\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\pprint.py\u001b[0m in \u001b[0;36m_repr\u001b[1;34m(self, object, context, level)\u001b[0m\n\u001b[0;32m    429\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    430\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0m_repr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mobject\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlevel\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 431\u001b[1;33m         repr, readable, recursive = self.format(object, context.copy(),\n\u001b[0m\u001b[0;32m    432\u001b[0m                                                 self._depth, level)\n\u001b[0;32m    433\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0mreadable\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\utils\\_pprint.py\u001b[0m in \u001b[0;36mformat\u001b[1;34m(self, object, context, maxlevels, level)\u001b[0m\n\u001b[0;32m    187\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    188\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mformat\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mobject\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmaxlevels\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlevel\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 189\u001b[1;33m         return _safe_repr(\n\u001b[0m\u001b[0;32m    190\u001b[0m             \u001b[0mobject\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcontext\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmaxlevels\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlevel\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mchanged_only\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_changed_only\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    191\u001b[0m         )\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\utils\\_pprint.py\u001b[0m in \u001b[0;36m_safe_repr\u001b[1;34m(object, context, maxlevels, level, changed_only)\u001b[0m\n\u001b[0;32m    438\u001b[0m         \u001b[0mrecursive\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mFalse\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    439\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mchanged_only\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 440\u001b[1;33m             \u001b[0mparams\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_changed_params\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mobject\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    441\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    442\u001b[0m             \u001b[0mparams\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mobject\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_params\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdeep\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\utils\\_pprint.py\u001b[0m in \u001b[0;36m_changed_params\u001b[1;34m(estimator)\u001b[0m\n\u001b[0;32m     91\u001b[0m     estimator with non-default values.\"\"\"\n\u001b[0;32m     92\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 93\u001b[1;33m     \u001b[0mparams\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mestimator\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_params\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdeep\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     94\u001b[0m     \u001b[0minit_func\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mestimator\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m__init__\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m\"deprecated_original\"\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mestimator\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m__init__\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     95\u001b[0m     \u001b[0minit_params\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0minspect\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msignature\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0minit_func\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mparameters\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\base.py\u001b[0m in \u001b[0;36mget_params\u001b[1;34m(self, deep)\u001b[0m\n\u001b[0;32m    208\u001b[0m         \u001b[0mout\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mdict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    209\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mkey\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_get_param_names\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 210\u001b[1;33m             \u001b[0mvalue\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    211\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mdeep\u001b[0m \u001b[1;32mand\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mvalue\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m\"get_params\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    212\u001b[0m                 \u001b[0mdeep_items\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mvalue\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mget_params\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
            "\u001b[1;31mAttributeError\u001b[0m: 'CRF' object has no attribute 'keep_tempfiles'"
          ]
        }
      ],
      "source": [
        "crf = sklearn_crfsuite.CRF(max_iterations=300)\n",
        "crf.fit(X_train, Y_train)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5VPFrPZJOJMq"
      },
      "source": [
        "## Evaluation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "84gUagMZOP9B"
      },
      "source": [
        "### Predict the labels of each of the tokens in each sentence of the test dataset that has been pre processed earlier."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "id": "D_sDJs1n8-sI"
      },
      "outputs": [],
      "source": [
        "Y_pred = crf.predict(X_test)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qIneYrqZOUiJ"
      },
      "source": [
        "### Calculate the f1 score using the actual labels and the predicted labels of the test dataset."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "id": "Lr_JNIgrOTcr"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "0.9073062342668767"
            ]
          },
          "execution_count": 26,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "metrics.flat_f1_score(Y_test, Y_pred, average='weighted')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Sentence:  The objective of this study was to determine if the rate of preeclampsia is increased in triplet as compared to twin gestations\n",
            "Actual labels:     ['O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'D', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O']\n",
            "Predicted labels:  ['O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'D', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O', 'O']\n"
          ]
        }
      ],
      "source": [
        "# Example test sentence and corresponding actual and predicted labels \n",
        "print(\"Sentence: \",test_sentences[13])\n",
        "print(\"Actual labels:    \", Y_test[13])\n",
        "print(\"Predicted labels: \", Y_pred[13])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "[['word.lower=the', 'word.postag=NOUN', 'word[-3:]=The', 'word[-2:]=he', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=True', 'BEG'], ['word.lower=objective', 'word.postag=NOUN', 'word[-3:]=ive', 'word[-2:]=ve', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=the', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=True'], ['word.lower=of', 'word.postag=NOUN', 'word[-3:]=of', 'word[-2:]=of', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=objective', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=this', 'word.postag=NOUN', 'word[-3:]=his', 'word[-2:]=is', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=of', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=study', 'word.postag=NOUN', 'word[-3:]=udy', 'word[-2:]=dy', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=this', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=was', 'word.postag=NOUN', 'word[-3:]=was', 'word[-2:]=as', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=study', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=to', 'word.postag=NOUN', 'word[-3:]=to', 'word[-2:]=to', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=was', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=determine', 'word.postag=NOUN', 'word[-3:]=ine', 'word[-2:]=ne', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=to', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=if', 'word.postag=NOUN', 'word[-3:]=if', 'word[-2:]=if', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=determine', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=the', 'word.postag=NOUN', 'word[-3:]=the', 'word[-2:]=he', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=if', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=rate', 'word.postag=NOUN', 'word[-3:]=ate', 'word[-2:]=te', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=the', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=of', 'word.postag=NOUN', 'word[-3:]=of', 'word[-2:]=of', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=rate', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=preeclampsia', 'word.postag=NOUN', 'word[-3:]=sia', 'word[-2:]=ia', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=of', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=is', 'word.postag=NOUN', 'word[-3:]=is', 'word[-2:]=is', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=preeclampsia', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=increased', 'word.postag=NOUN', 'word[-3:]=sed', 'word[-2:]=ed', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=is', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=in', 'word.postag=NOUN', 'word[-3:]=in', 'word[-2:]=in', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=increased', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=triplet', 'word.postag=NOUN', 'word[-3:]=let', 'word[-2:]=et', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=in', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=as', 'word.postag=NOUN', 'word[-3:]=as', 'word[-2:]=as', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=triplet', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=compared', 'word.postag=NOUN', 'word[-3:]=red', 'word[-2:]=ed', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=as', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=to', 'word.postag=NOUN', 'word[-3:]=to', 'word[-2:]=to', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=compared', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=twin', 'word.postag=NOUN', 'word[-3:]=win', 'word[-2:]=in', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=to', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False'], ['word.lower=gestations', 'word.postag=NOUN', 'word[-3:]=ons', 'word[-2:]=ns', 'word.isupper=False', 'word.isdigit=False', 'words.startsWithCapital=False', 'prev_word.lower=twin', 'prev_word.postag=NOUN', 'prev_word.isupper=False', 'prev_word.isdigit=False', 'prev_words.startsWithCapital=False', 'END']]\n"
          ]
        }
      ],
      "source": [
        "# Feature list of sentence above\n",
        "print(X_test[13])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ORDyDw6sYHf5"
      },
      "source": [
        "##Identifying Diseases and Treatments using Custom NER\n",
        "\n",
        "We now use the CRF model's prediction to prepare a record of diseases identified in the corpus and treatments used for the diseases.\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "V9aOa7LEOcWu"
      },
      "source": [
        "### Create the logic to get all the predicted treatments (T) labels corresponding to each disease (D) label in the test dataset."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sWxZ2TaOqN-u"
      },
      "source": [
        "![image.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "id": "6reMqClg97Ry"
      },
      "outputs": [],
      "source": [
        "#Extracting a dictionary of all the predicted diseases from our test data and the corresponding treatments.\n",
        "#Assumption: For each identified disease, one of the treatments is in the same sentence as the disease exists.\n",
        "disease_treatment = {}           \n",
        "for i in range(len(Y_pred)):\n",
        "    cnt_disease = 0           \n",
        "    cnt_treatment = 0         \n",
        "    diseases = [\"\"]           \n",
        "    treatment = [\"\"]          \n",
        "    length = len(Y_pred[i])   \n",
        "    for j in range(length):\n",
        "        if (Y_pred[i][j] == 'D'):                                                     # Checking for label indicating disease for current word ('D')\n",
        "            diseases[cnt_disease] += (X_test[i][j][0].split('=')[1] + \" \")            # Adding word to diseases list.\n",
        "            if j < length - 1:\n",
        "                if (Y_pred[i][j+1] != 'D'):                                           # Check for name of disease extending over multiple words. \n",
        "                    # If next word does not have label 'D', then truncate the space added at the end of the last word.\n",
        "                    diseases[cnt_disease] = diseases[cnt_disease][:-1]\n",
        "                    cnt_disease += 1\n",
        "                    diseases.append(\"\")                                               # Adding a placeholder for the next disease in the current sentence.\n",
        "            else:\n",
        "                diseases[cnt_disease] = diseases[cnt_disease][:-1]\n",
        "                cnt_disease += 1\n",
        "                diseases.append(\"\")\n",
        "                            \n",
        "        if (Y_pred[i][j] == 'T'):                                                     # Checking for label indicating treatment for current word ('T')\n",
        "            treatment[cnt_treatment] += (X_test[i][j][0].split('=')[1] + \" \") # Adding word to corresponding treatment list.\n",
        "            if j < length - 1:\n",
        "                if (Y_pred[i][j+1] != 'T'):                                           # Check for name of treatment extending over multiple words. \n",
        "                    # If next word does not have label 'T', then truncate the space added at the end of the last word.\n",
        "                    treatment[cnt_treatment] = treatment[cnt_treatment][:-1]\n",
        "                    cnt_treatment += 1\n",
        "                    treatment.append(\"\")                                              # Adding a placeholder for the next treatment in the current sentence.\n",
        "            else:\n",
        "                treatment[cnt_treatment] = treatment[cnt_treatment][:-1]\n",
        "                cnt_treatment += 1\n",
        "                treatment.append(\"\")\n",
        "\n",
        "    diseases.pop(-1)    # Getting rid of the last empty placeholder in diseases list\n",
        "    treatment.pop(-1)   # Getting rid of the last empty placeholder in treatments list\n",
        "\n",
        "    # To our dictionary, add or append treatments to the diseases identified from the current sentence, if any.\n",
        "    if len(diseases) > 0:       # Checking if any diseases have been identified for the current sentence.\n",
        "        for disease in diseases:\n",
        "            if disease in disease_treatment.keys():\n",
        "                # Extend treatment list if other treatments for the particular disease already exist\n",
        "                disease_treatment[disease].extend(treatment)\n",
        "            else:\n",
        "                # Creating list of treatments for particular disease if it doesn not exist already.\n",
        "                disease_treatment[disease] = treatment"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "{'gestational diabetes cases': [],\n",
              " 'preeclampsia': [],\n",
              " 'severe preeclampsia': [],\n",
              " 'asymmetric double hemiplegia': [],\n",
              " 'reversible nonimmune hydrops fetalis': [],\n",
              " 'breast and/or ovarian cancer': [],\n",
              " 'breast cancer': ['hormone replacement therapy',\n",
              "  'undergone subcutaneous mastectomy'],\n",
              " 'ovarian cancer': ['hormone replacement therapy',\n",
              "  'undergone subcutaneous mastectomy'],\n",
              " 'prostate cancer': ['radical prostatectomy and iodine 125 interstitial radiotherapy'],\n",
              " 'mutated prostate cancer': ['radical prostatectomy and iodine 125 interstitial radiotherapy'],\n",
              " 'hereditary prostate cancer': [],\n",
              " 'multiple sclerosis ( ms )': [],\n",
              " 'hereditary retinoblastoma': ['radiotherapy'],\n",
              " 'epilepsy': [],\n",
              " 'unstable angina or non-q-wave myocardial infarction': ['roxithromycin'],\n",
              " 'coronary-artery disease': ['antichlamydial antibiotics'],\n",
              " 'early-stage cervical carcinoma': [],\n",
              " 'advanced disease': [],\n",
              " 'cerebral palsy': ['hyperbaric oxygen therapy'],\n",
              " 'severe pain': [],\n",
              " 'myofascial trigger point pain': [],\n",
              " 'infections': [],\n",
              " 'primary pulmonary hypertension ( pph )': ['fenfluramines'],\n",
              " 'essential hypertension': [],\n",
              " 'osteoporosis among older persons': [],\n",
              " 'foot infections': [],\n",
              " 'cellulitis': ['g-csf therapy', 'intravenous antibiotic treatment'],\n",
              " 'foot infection': [],\n",
              " 'hemorrhagic stroke': [],\n",
              " 'cardiac disease': [],\n",
              " 'rheumatoid arthritis': [],\n",
              " 'stroke victims': [],\n",
              " \"early parkinson 's disease\": ['ropinirole monotherapy'],\n",
              " 'sore throat': ['antibiotics'],\n",
              " 'upper respiratory conditions': [],\n",
              " 'sore throat increase reattendance': [],\n",
              " 'postoperative pain relief': [],\n",
              " 'analgesia when equianalgesic doses': [],\n",
              " 'chronic neutropenia': [],\n",
              " 'female stress urinary incontinence': ['surgical treatment'],\n",
              " 'stress urinary incontinence': ['therapy'],\n",
              " 'corpal gastritis': [],\n",
              " 'preeclampsia ( proteinuric hypertension )': ['intrauterine insemination with donor sperm versus intrauterine insemination'],\n",
              " 'mild preeclampsia': [],\n",
              " 'myocardial amyloidosis': [],\n",
              " 'heart failure': [],\n",
              " 'amyloidosis': [],\n",
              " 'cardiac amyloid heart disease': [],\n",
              " 'retinopathy': [],\n",
              " 'intra-abdominal injury': [],\n",
              " 'tangential gunshot wounds': [],\n",
              " 'intra-abdominal hemorrhage': [],\n",
              " 'severe acquired hyperammonemia': ['organ transplantation and chemotherapy',\n",
              "  'matrix metalloproteinase inhibitors'],\n",
              " 'cancer': ['organ transplantation and chemotherapy',\n",
              "  'matrix metalloproteinase inhibitors'],\n",
              " 'major pulmonary embolism': ['thrombolytic treatment'],\n",
              " 'congestive heart failure': [],\n",
              " 'chronic pulmonary disease': [],\n",
              " 'malignant pleural mesothelioma': ['thoracotomy , radiotherapy , and chemotherapy'],\n",
              " 'additional edema': [],\n",
              " 'dyskinesias': [],\n",
              " 'darier disease': [],\n",
              " 'tumor necrosis factor-alpha': [],\n",
              " 'syndrome of hemorrhagic shock and encephalopathy': [],\n",
              " 'tss': [],\n",
              " 'pulmonary symptoms attributable': ['chemotherapy'],\n",
              " 'non-obstructive azoospermia': ['testicular sperm retrieval by testicular fine needle aspiration ( tefna )',\n",
              "  'open biopsy and testicular sperm extraction ( tese )'],\n",
              " 'mantle cell lymphoma': [],\n",
              " 'acute lymphoblastic leukaemias': [],\n",
              " 'endometriosis': [],\n",
              " 'hyperlipidemias': [],\n",
              " 'deep vein thrombosis ( dvt )': [],\n",
              " 'hepatocerebral degeneration ( portal systemic encephalopathy )': [],\n",
              " 'mastitis': [],\n",
              " 'cervical carcinoma': [],\n",
              " 'duodenal ulcer': [],\n",
              " 'childhood schizophrenia': [],\n",
              " 'congenital adrenal hyperplasia': [],\n",
              " \"laennec 's cirrhosis\": [],\n",
              " 'renal stones': [],\n",
              " 'stone disease': [],\n",
              " 'congenital anomalies': [],\n",
              " 'ovarian carcinoma': [],\n",
              " 'heart congestive failure': [],\n",
              " 'biliary indication': [],\n",
              " 'piglet influenza': [],\n",
              " 'chronic progressive external ophthalmoplegia': [],\n",
              " 'cystic periventricular leukomalacia': [],\n",
              " 'metastasis of laryngeal carcinomas': [],\n",
              " 'hiv infection': [],\n",
              " 'graft spasm': [],\n",
              " 'renal dysfunction': [],\n",
              " 'colorectal cancer': ['elective surgery',\n",
              "  'docosahexanoic acid ( dha ) -concentrated fish oil'],\n",
              " \"leukaemia and non-hodgkin 's lymphoma\": [],\n",
              " 'childhood acute lymphoblastic leukaemia': [],\n",
              " 'gastrointestinal tumours': ['elective surgery'],\n",
              " 'chlamydia trachomatis': [],\n",
              " 'anemia of end-stage renal disease': [],\n",
              " 'malignant lymphomas': [],\n",
              " 'polymyalgia rheumatica and temporal arteritis': [],\n",
              " 'coronary heart disease': [],\n",
              " 'restenosis after coronary angioplasty': [],\n",
              " \"oral crohn 's disease\": [],\n",
              " 'gestational diabetes': [],\n",
              " 'bronchial asthma': [],\n",
              " 'disconnection syndrome': [],\n",
              " 'severe accidental hypothermia victims': [],\n",
              " 'hiv/aids': [],\n",
              " 'gunshot wounds': [],\n",
              " 'blastocystis hominis infection': [],\n",
              " 'does hyperglycemia really cause coronary heart disease': [],\n",
              " 'primary stomach lymphoma': [],\n",
              " 'systemic lupus erythematosus : mite dermatitis': [],\n",
              " 'lung cancer': [],\n",
              " 'metastatic malignant mesothelioma presenting as colonic polyps': [],\n",
              " \"parkinson 's disease\": ['microelectrode-guided posteroventral pallidotomy'],\n",
              " 'murine hepatitis virus strain 3 infection': [],\n",
              " 'synovial sarcoma': [],\n",
              " 'small cell lung cancer': [],\n",
              " 'acute respiratory distress syndrome': [],\n",
              " 'diabetes': [],\n",
              " 'chronic hepatitis': [],\n",
              " 'postraumatic stress disorder': [],\n",
              " 'kaposi sarcoma-associated herpesvirus infection': [],\n",
              " 'diabetes and hyperlipidemias': [],\n",
              " 'cardiac insufficiency': [],\n",
              " 'malignant soft tissue sarcomas': [],\n",
              " 'soft tissue sarcomas': ['radiotherapy'],\n",
              " 'intramedullary metastases of bronchogenic carcinoma': [],\n",
              " 'pseudo-lymphoma': [],\n",
              " 'auditory perception disorders due to bilateral cortical lesions': [],\n",
              " 'cerebral injuries': [],\n",
              " 'coronary endothelial dysfunction': [],\n",
              " 'myocardial perfusion defects': [],\n",
              " 'ectopic pregnancy': [],\n",
              " \"alzheimer 's disease\": [],\n",
              " \"eisenmenger 's syndrome\": ['laparoscopic cholecystectomy'],\n",
              " 'continuous ambulatory peritoneal dialysis': [],\n",
              " 'advanced esophageal cancer': ['adjuvant chemoradiotherapy with cddp'],\n",
              " 'sickle cell anemia': [],\n",
              " 'intractable seizures': [],\n",
              " 'new-onset angina preceding acute myocardial infarction': [],\n",
              " 'congenital bilateral aplasia': [],\n",
              " 'insulin-dependent diabetes mellitus': [],\n",
              " 'aids exceptionalism': [],\n",
              " 'myoepithelial carcinoma': [],\n",
              " 'pleomorphic adenoma': [],\n",
              " 'homocystinuria': [],\n",
              " 'malaria': [],\n",
              " 'severe psoriasis': [],\n",
              " 'leukemia': ['trisomy'],\n",
              " 'acute obstructive hydrocephalus complicating bacterial meningitis': [],\n",
              " 'refractory pouchitis': [],\n",
              " 'some malignant tumors such as non-small cell lung cancer': [],\n",
              " 'inoperable advanced malignancies such as colorectal cancer': ['combination with leucovorin or cisplatin'],\n",
              " 'advanced nsclc with manageable toxicities': [],\n",
              " 'recurrent cervical adenocarcinoma': [],\n",
              " 'influenza provides': [],\n",
              " 'influenza control': [],\n",
              " 'abdominal pain': [],\n",
              " 'pain': [],\n",
              " 'pleuritic pain': [],\n",
              " 'fungal infections': [],\n",
              " 'inflammatory skin diseases': ['topical corticosteroids'],\n",
              " 'unresectable stage iii nsclc': [],\n",
              " 'nsclc': ['got surgical treatment',\n",
              "  'radiotherapy',\n",
              "  'got surgical treatment',\n",
              "  'radiotherapy',\n",
              "  'platinum-based chemotherapy'],\n",
              " 'sclc': ['got surgical treatment',\n",
              "  'radiotherapy',\n",
              "  'got surgical treatment',\n",
              "  'radiotherapy',\n",
              "  'platinum-based chemotherapy'],\n",
              " 'locally advanced non-small-cell lung cancer ( la-nsclc )': ['chemotherapy and radiotherapy )'],\n",
              " 'radiation-induced myelopathy': ['heparin and enoxaparin'],\n",
              " 'limited stage small cell lung cancer': ['vip combination chemotherapy'],\n",
              " 'malignant pleural effusions from nsclc': ['systemic chemotherapy'],\n",
              " 'small-cell lung cancer': ['chemotherapy', 'combination chemotherapy'],\n",
              " 'intraluminal early-stage cancer': [],\n",
              " 'pathologic stage i non-small cell lung cancer': [],\n",
              " 'supraclavicular node metastases in nsclc': ['chemoradiotherapy'],\n",
              " 'non-small-cell-lung-cancer ( nsclc )': ['cisplatin and radiotherapy'],\n",
              " 'esophageal cancer': [],\n",
              " 'lung cancer )': [],\n",
              " 'osteogenic sarcoma': [],\n",
              " 'lung carcinoma': ['videothoracoscopic lobectomy or partial resection',\n",
              "  'open thoracotomy'],\n",
              " 'stage 0 lung carcinoma': ['curative therapy'],\n",
              " 'melanoma': ['surgical resection'],\n",
              " 'ovarian carcinoma brain metastasis': ['surgical resection'],\n",
              " 'colorectal metastases': ['therapeutic vats metastasectomy'],\n",
              " 'limited-stage small-cell carcinoma': [],\n",
              " 'advanced nsclc': ['assessing combination chemotherapy of cisplatin , ifosfamide and irinotecan with rhg-csf support'],\n",
              " 'metastatic colorectal cancer': ['intravenous oxaliplatin'],\n",
              " 'platinum-pretreated ovarian cancer': ['oxaliplatin'],\n",
              " \"non-hodgkin 's lymphoma , breast cancer\": ['oxaliplatin'],\n",
              " 'mesothelioma and non-small cell lung cancer': ['oxaliplatin'],\n",
              " 'bronchogenic carcinoma': [],\n",
              " 'non-small cell lung cancer': ['combination with paclitaxel and carboplatin'],\n",
              " 'advanced hormone refractory prostate cancer': ['combination with paclitaxel and carboplatin'],\n",
              " 'primary lung cancer': ['resection'],\n",
              " 'adenocarcinoma': ['resection'],\n",
              " 'squamous cell carcinoma': ['resection'],\n",
              " 'stage iii nsclc': ['chemotherapy administered before surgery'],\n",
              " 'primary cancer': ['adjuvant radiation therapy', 'immunotherapy'],\n",
              " 'advanced non -- small-cell lung cancer': ['paclitaxel plus carboplatin ( pc )',\n",
              "  'vinorelbine plus cisplatin ( vc )'],\n",
              " 'extensive disease': ['platinum dose ( cisplatin plus carboplatin ) in combination chemotherapy',\n",
              "  'combination therapy with carboplatin alone'],\n",
              " 'untreated small cell lung cancer ( sclc )': [],\n",
              " 'untreated sclc': [],\n",
              " 'pulmonary hypertension': [],\n",
              " 'neutropenia during cancer': ['chemotherapy'],\n",
              " 'head and neck cancer': ['intravenous amifostine'],\n",
              " ' xerostomia': ['intravenous amifostine'],\n",
              " 'carbon monoxide poisoning': [],\n",
              " 'acute myocardial infarction': ['thrombolytic therapy'],\n",
              " 'psoriasis': ['topical application of active vitamin d3 analogue , 1 alpha , 24-dihydroxycholecalciferol'],\n",
              " 'disseminated malignant melanoma': ['leukocyte a recombinant interferon ( rifn-alpha'],\n",
              " 'advanced stage ( tnm iib-ivb ) mycosis fungoides': ['a combination chemotherapy program consisting of bleomycin and methotrexate weekly , doxorubicin every'],\n",
              " 'ventricular tachycardia': ['guiding surgical therapy'],\n",
              " 'cholestasis': [],\n",
              " 'severe acute hepatitis accompanying cholestasis': [],\n",
              " 'autoimmune hepatitis': [],\n",
              " 'spinal adhesive arachnoiditis': ['surgical management'],\n",
              " 'bronchiectasis': ['antibiotics and surgery', 'current surgical therapy'],\n",
              " 'biliary colic symptoms': ['cholecystectomy'],\n",
              " 'biliary dyskinesia': ['cholecystectomy'],\n",
              " 'the common cold': ['pseudoephedrine plus acetaminophen'],\n",
              " 'pulmonary aspergilloma': [],\n",
              " 'viral infections': [],\n",
              " 'asthma': ['fluticasone propionate'],\n",
              " 'pneumocystis carinii pneumonia': [],\n",
              " 'acute nasopharyngitis ( anp )': ['antibiotic treatment'],\n",
              " 'symptoms of a common cold': ['macrolide antibiotics'],\n",
              " 'rhinovirus colds': ['clarithromycin'],\n",
              " 'rhinovirus': ['clarithromycin'],\n",
              " 'acute exacerbation of chronic bronchitis': [],\n",
              " 'pneumonia': [],\n",
              " 'influenza': [],\n",
              " 'influenza a2 ( h2n2 ) virus': [],\n",
              " 'infection': ['a combination of omeprazole , amoxicillin , and clarithromycin'],\n",
              " 'carcinoma': [],\n",
              " 'persistent asthma': ['contemporary asthma management guidelines list inhaled corticosteroids'],\n",
              " 'chronic hepatitis c': ['combination therapy with interferon-alpha ( ifn alpha ) plus ribavirin'],\n",
              " 'hepatitis c viremia': ['combination therapy'],\n",
              " 'severe hypoxemia': ['glucocorticoid pulse therapy'],\n",
              " 'drug-resistant s. pneumoniae': [],\n",
              " 'bacterial meningitis': [],\n",
              " 'haemophilus influenzae meningitis': [],\n",
              " 'asthmatic bronchitis': [],\n",
              " 'rsv bronchiolitis': [],\n",
              " 'primary sclerosing cholangitis ( psc )': [],\n",
              " 'chronic renal failure': [],\n",
              " 'ischemic heart disease': [],\n",
              " 'peritoneal tumors': ['subcutaneous injection of irradiated llc-il2'],\n",
              " 'acute occlusion of the middle cerebral artery': ['thrombolytic therapy'],\n",
              " 'benign prostatic hyperplasia': [],\n",
              " 'rheumatic pain herbal': [],\n",
              " 'inflammatory and autoimmune diseases': ['high-dose intravenous immunoglobulin ( hdivig )'],\n",
              " 'large midline incisional hernias': [],\n",
              " 'low-grade nhl': [],\n",
              " 'large-bowel cancer': [],\n",
              " 'phaeochromocytoma': ['adrenalectomy'],\n",
              " 'malignant melanoma': ['interferon alfa-2a'],\n",
              " 'advanced renal cell carcinoma': ['various interferon alpha preparations',\n",
              "  'interferon alfa-n1 , interferon alfa-2a , and interferon alfa-2b'],\n",
              " 'hairy cell leukemia': [],\n",
              " \"low-grade non-hodgkin 's lymphoma\": ['interferon alpha'],\n",
              " \"low-grade non-hodgkin 's lymphomas\": [],\n",
              " 'chronic myelogenous leukemia ( cml )': [],\n",
              " 'partial seizures': ['lamotrigine ( ltg )', 'lamotrigine monotherapy'],\n",
              " 'esophageal achalasia': ['botulinum toxin injection , pneumatic dilation , and laparoscopic myotomy'],\n",
              " 'irritable bowel syndrome': ['chinese herbal medicine'],\n",
              " 'proximal hypospadias': ['tubularized incised plate hypospadias repair'],\n",
              " 'tumors': ['immunotherapy'],\n",
              " 'sickle cell disease': [],\n",
              " 'haematoma block versus intravenous regional anaesthesia': [],\n",
              " 'stroke': [],\n",
              " 'mitomycin-resistant bladder cancer': ['photodynamic therapy in combination with mitomycin c'],\n",
              " 'b16 melanoma': ['adenosine triphosphate and treatment with buthionine sulfoximine'],\n",
              " 'primary uveal melanoma': ['transpupillary thermotherapy'],\n",
              " 'advanced rectal cancer': [],\n",
              " 'noncirrhotic portal hypertension': [],\n",
              " 'midline incisional hernias': [],\n",
              " 'empyema': [],\n",
              " \"non-hodgkin 's lymphoma\": [],\n",
              " 'acute cerebral ischemia': ['antiplatelet therapy'],\n",
              " 'hyperopia': [],\n",
              " 'renal cell carcinoma': ['interferon treatment'],\n",
              " 'odontogenic infections': [],\n",
              " 'myocardial angiogenesis': ['gene therapy'],\n",
              " 'autoimmune hemolytic anemia': ['heparin'],\n",
              " 'epithelial ovarian cancer': ['high-dose chemotherapy'],\n",
              " 'atrial fibrillation': ['antiplatelet therapy'],\n",
              " 'diabetes mellitus': [],\n",
              " 'lymphoma': [],\n",
              " 'acute repetitive seizures': [],\n",
              " 'moderately symptomatic benign prostatic hyperplasia': ['surgical resection'],\n",
              " 'multiple sclerosis': ['interferon beta treatment',\n",
              "  'intravenous immunoglobulin treatment'],\n",
              " 'acoustic neuroma': ['stereotactic radiosurgery'],\n",
              " 'postvitrectomy diabetic vitreous hemorrhage': ['peripheral retinal cryotherapy'],\n",
              " 'acute carbon monoxide poisoning': [],\n",
              " 'hepatitis b': ['vaccine containing mf59'],\n",
              " 'temporomandibular joint arthropathy': [],\n",
              " 'acute colonic pseudo-obstruction': [],\n",
              " 'severe secondary peritonitis': ['surgical management'],\n",
              " 'hepatic metastases from colorectal cancer': ['hepatic arterial infusion of chemotherapy after resection'],\n",
              " 'severe diverticular hemorrhage': [],\n",
              " 'stable asthma': [],\n",
              " 'arthritis': [],\n",
              " 'live oral poliovirus vaccines': [],\n",
              " 'glaucoma': [],\n",
              " 'responsive multiple myeloma': []}"
            ]
          },
          "execution_count": 30,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Displaying dictionary of extracted diseases and potential treatments.\n",
        "disease_treatment"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "{'breast cancer': ['hormone replacement therapy',\n",
              "  'undergone subcutaneous mastectomy'],\n",
              " 'ovarian cancer': ['hormone replacement therapy',\n",
              "  'undergone subcutaneous mastectomy'],\n",
              " 'prostate cancer': ['radical prostatectomy and iodine 125 interstitial radiotherapy'],\n",
              " 'mutated prostate cancer': ['radical prostatectomy and iodine 125 interstitial radiotherapy'],\n",
              " 'hereditary retinoblastoma': ['radiotherapy'],\n",
              " 'unstable angina or non-q-wave myocardial infarction': ['roxithromycin'],\n",
              " 'coronary-artery disease': ['antichlamydial antibiotics'],\n",
              " 'cerebral palsy': ['hyperbaric oxygen therapy'],\n",
              " 'primary pulmonary hypertension ( pph )': ['fenfluramines'],\n",
              " 'cellulitis': ['g-csf therapy', 'intravenous antibiotic treatment'],\n",
              " \"early parkinson 's disease\": ['ropinirole monotherapy'],\n",
              " 'sore throat': ['antibiotics'],\n",
              " 'female stress urinary incontinence': ['surgical treatment'],\n",
              " 'stress urinary incontinence': ['therapy'],\n",
              " 'preeclampsia ( proteinuric hypertension )': ['intrauterine insemination with donor sperm versus intrauterine insemination'],\n",
              " 'severe acquired hyperammonemia': ['organ transplantation and chemotherapy',\n",
              "  'matrix metalloproteinase inhibitors'],\n",
              " 'cancer': ['organ transplantation and chemotherapy',\n",
              "  'matrix metalloproteinase inhibitors'],\n",
              " 'major pulmonary embolism': ['thrombolytic treatment'],\n",
              " 'malignant pleural mesothelioma': ['thoracotomy , radiotherapy , and chemotherapy'],\n",
              " 'pulmonary symptoms attributable': ['chemotherapy'],\n",
              " 'non-obstructive azoospermia': ['testicular sperm retrieval by testicular fine needle aspiration ( tefna )',\n",
              "  'open biopsy and testicular sperm extraction ( tese )'],\n",
              " 'colorectal cancer': ['elective surgery',\n",
              "  'docosahexanoic acid ( dha ) -concentrated fish oil'],\n",
              " 'gastrointestinal tumours': ['elective surgery'],\n",
              " \"parkinson 's disease\": ['microelectrode-guided posteroventral pallidotomy'],\n",
              " 'soft tissue sarcomas': ['radiotherapy'],\n",
              " \"eisenmenger 's syndrome\": ['laparoscopic cholecystectomy'],\n",
              " 'advanced esophageal cancer': ['adjuvant chemoradiotherapy with cddp'],\n",
              " 'leukemia': ['trisomy'],\n",
              " 'inoperable advanced malignancies such as colorectal cancer': ['combination with leucovorin or cisplatin'],\n",
              " 'inflammatory skin diseases': ['topical corticosteroids'],\n",
              " 'nsclc': ['got surgical treatment',\n",
              "  'radiotherapy',\n",
              "  'got surgical treatment',\n",
              "  'radiotherapy',\n",
              "  'platinum-based chemotherapy'],\n",
              " 'sclc': ['got surgical treatment',\n",
              "  'radiotherapy',\n",
              "  'got surgical treatment',\n",
              "  'radiotherapy',\n",
              "  'platinum-based chemotherapy'],\n",
              " 'locally advanced non-small-cell lung cancer ( la-nsclc )': ['chemotherapy and radiotherapy )'],\n",
              " 'radiation-induced myelopathy': ['heparin and enoxaparin'],\n",
              " 'limited stage small cell lung cancer': ['vip combination chemotherapy'],\n",
              " 'malignant pleural effusions from nsclc': ['systemic chemotherapy'],\n",
              " 'small-cell lung cancer': ['chemotherapy', 'combination chemotherapy'],\n",
              " 'supraclavicular node metastases in nsclc': ['chemoradiotherapy'],\n",
              " 'non-small-cell-lung-cancer ( nsclc )': ['cisplatin and radiotherapy'],\n",
              " 'lung carcinoma': ['videothoracoscopic lobectomy or partial resection',\n",
              "  'open thoracotomy'],\n",
              " 'stage 0 lung carcinoma': ['curative therapy'],\n",
              " 'melanoma': ['surgical resection'],\n",
              " 'ovarian carcinoma brain metastasis': ['surgical resection'],\n",
              " 'colorectal metastases': ['therapeutic vats metastasectomy'],\n",
              " 'advanced nsclc': ['assessing combination chemotherapy of cisplatin , ifosfamide and irinotecan with rhg-csf support'],\n",
              " 'metastatic colorectal cancer': ['intravenous oxaliplatin'],\n",
              " 'platinum-pretreated ovarian cancer': ['oxaliplatin'],\n",
              " \"non-hodgkin 's lymphoma , breast cancer\": ['oxaliplatin'],\n",
              " 'mesothelioma and non-small cell lung cancer': ['oxaliplatin'],\n",
              " 'non-small cell lung cancer': ['combination with paclitaxel and carboplatin'],\n",
              " 'advanced hormone refractory prostate cancer': ['combination with paclitaxel and carboplatin'],\n",
              " 'primary lung cancer': ['resection'],\n",
              " 'adenocarcinoma': ['resection'],\n",
              " 'squamous cell carcinoma': ['resection'],\n",
              " 'stage iii nsclc': ['chemotherapy administered before surgery'],\n",
              " 'primary cancer': ['adjuvant radiation therapy', 'immunotherapy'],\n",
              " 'advanced non -- small-cell lung cancer': ['paclitaxel plus carboplatin ( pc )',\n",
              "  'vinorelbine plus cisplatin ( vc )'],\n",
              " 'extensive disease': ['platinum dose ( cisplatin plus carboplatin ) in combination chemotherapy',\n",
              "  'combination therapy with carboplatin alone'],\n",
              " 'neutropenia during cancer': ['chemotherapy'],\n",
              " 'head and neck cancer': ['intravenous amifostine'],\n",
              " ' xerostomia': ['intravenous amifostine'],\n",
              " 'acute myocardial infarction': ['thrombolytic therapy'],\n",
              " 'psoriasis': ['topical application of active vitamin d3 analogue , 1 alpha , 24-dihydroxycholecalciferol'],\n",
              " 'disseminated malignant melanoma': ['leukocyte a recombinant interferon ( rifn-alpha'],\n",
              " 'advanced stage ( tnm iib-ivb ) mycosis fungoides': ['a combination chemotherapy program consisting of bleomycin and methotrexate weekly , doxorubicin every'],\n",
              " 'ventricular tachycardia': ['guiding surgical therapy'],\n",
              " 'spinal adhesive arachnoiditis': ['surgical management'],\n",
              " 'bronchiectasis': ['antibiotics and surgery', 'current surgical therapy'],\n",
              " 'biliary colic symptoms': ['cholecystectomy'],\n",
              " 'biliary dyskinesia': ['cholecystectomy'],\n",
              " 'the common cold': ['pseudoephedrine plus acetaminophen'],\n",
              " 'asthma': ['fluticasone propionate'],\n",
              " 'acute nasopharyngitis ( anp )': ['antibiotic treatment'],\n",
              " 'symptoms of a common cold': ['macrolide antibiotics'],\n",
              " 'rhinovirus colds': ['clarithromycin'],\n",
              " 'rhinovirus': ['clarithromycin'],\n",
              " 'infection': ['a combination of omeprazole , amoxicillin , and clarithromycin'],\n",
              " 'persistent asthma': ['contemporary asthma management guidelines list inhaled corticosteroids'],\n",
              " 'chronic hepatitis c': ['combination therapy with interferon-alpha ( ifn alpha ) plus ribavirin'],\n",
              " 'hepatitis c viremia': ['combination therapy'],\n",
              " 'severe hypoxemia': ['glucocorticoid pulse therapy'],\n",
              " 'peritoneal tumors': ['subcutaneous injection of irradiated llc-il2'],\n",
              " 'acute occlusion of the middle cerebral artery': ['thrombolytic therapy'],\n",
              " 'inflammatory and autoimmune diseases': ['high-dose intravenous immunoglobulin ( hdivig )'],\n",
              " 'phaeochromocytoma': ['adrenalectomy'],\n",
              " 'malignant melanoma': ['interferon alfa-2a'],\n",
              " 'advanced renal cell carcinoma': ['various interferon alpha preparations',\n",
              "  'interferon alfa-n1 , interferon alfa-2a , and interferon alfa-2b'],\n",
              " \"low-grade non-hodgkin 's lymphoma\": ['interferon alpha'],\n",
              " 'partial seizures': ['lamotrigine ( ltg )', 'lamotrigine monotherapy'],\n",
              " 'esophageal achalasia': ['botulinum toxin injection , pneumatic dilation , and laparoscopic myotomy'],\n",
              " 'irritable bowel syndrome': ['chinese herbal medicine'],\n",
              " 'proximal hypospadias': ['tubularized incised plate hypospadias repair'],\n",
              " 'tumors': ['immunotherapy'],\n",
              " 'mitomycin-resistant bladder cancer': ['photodynamic therapy in combination with mitomycin c'],\n",
              " 'b16 melanoma': ['adenosine triphosphate and treatment with buthionine sulfoximine'],\n",
              " 'primary uveal melanoma': ['transpupillary thermotherapy'],\n",
              " 'acute cerebral ischemia': ['antiplatelet therapy'],\n",
              " 'renal cell carcinoma': ['interferon treatment'],\n",
              " 'myocardial angiogenesis': ['gene therapy'],\n",
              " 'autoimmune hemolytic anemia': ['heparin'],\n",
              " 'epithelial ovarian cancer': ['high-dose chemotherapy'],\n",
              " 'atrial fibrillation': ['antiplatelet therapy'],\n",
              " 'moderately symptomatic benign prostatic hyperplasia': ['surgical resection'],\n",
              " 'multiple sclerosis': ['interferon beta treatment',\n",
              "  'intravenous immunoglobulin treatment'],\n",
              " 'acoustic neuroma': ['stereotactic radiosurgery'],\n",
              " 'postvitrectomy diabetic vitreous hemorrhage': ['peripheral retinal cryotherapy'],\n",
              " 'hepatitis b': ['vaccine containing mf59'],\n",
              " 'severe secondary peritonitis': ['surgical management'],\n",
              " 'hepatic metastases from colorectal cancer': ['hepatic arterial infusion of chemotherapy after resection']}"
            ]
          },
          "execution_count": 31,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Creating a final version of our \"disease_treatment\" dictionary\n",
        "final_dict = {}\n",
        "for key in disease_treatment.keys():\n",
        "    if disease_treatment[key] != []:\n",
        "        final_dict[key] = disease_treatment[key]\n",
        "final_dict"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>disease</th>\n",
              "      <th>treatments</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>breast cancer</td>\n",
              "      <td>[hormone replacement therapy, undergone subcut...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ovarian cancer</td>\n",
              "      <td>[hormone replacement therapy, undergone subcut...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>prostate cancer</td>\n",
              "      <td>[radical prostatectomy and iodine 125 intersti...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>mutated prostate cancer</td>\n",
              "      <td>[radical prostatectomy and iodine 125 intersti...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>hereditary retinoblastoma</td>\n",
              "      <td>[radiotherapy]</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                     disease  \\\n",
              "0              breast cancer   \n",
              "1             ovarian cancer   \n",
              "2            prostate cancer   \n",
              "3    mutated prostate cancer   \n",
              "4  hereditary retinoblastoma   \n",
              "\n",
              "                                          treatments  \n",
              "0  [hormone replacement therapy, undergone subcut...  \n",
              "1  [hormone replacement therapy, undergone subcut...  \n",
              "2  [radical prostatectomy and iodine 125 intersti...  \n",
              "3  [radical prostatectomy and iodine 125 intersti...  \n",
              "4                                     [radiotherapy]  "
            ]
          },
          "execution_count": 32,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Converting dictionary to a dataframe\n",
        "final_df = pd.DataFrame({\"disease\":final_dict.keys(),\"treatments\":final_dict.values()})\n",
        "final_df.head()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TvTodSY5OfZ9"
      },
      "source": [
        "### Predict the treatment for the disease name: 'hereditary retinoblastoma'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "id": "uZNBLMFYVuhv"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Treatments for 'hereditary retinoblastoma' is/are 'radiotherapy'\n"
          ]
        }
      ],
      "source": [
        "search_item = 'hereditary retinoblastoma'\n",
        "treatments = final_dict[search_item]\n",
        "print(\"Treatments for '{0}' is/are \".format(search_item), end = \"\")\n",
        "for i in range(len(treatments)-1):\n",
        "    print(\"'{}'\".format(treatments[i]),\",\", end=\"\")\n",
        "print(\"'{}'\".format(treatments[-1]))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "w_TonOcVAG1P"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "YeNQRLrIYRay"
      },
      "outputs": [],
      "source": []
    }
  ],
  "metadata": {
    "colab": {
      "collapsed_sections": [],
      "name": "Assignment.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.13"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 1
}