Seq2Seq(Simple).ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"news_summary_more.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data[0:50000]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>headlines</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>upGrad learner switches to career in ML &amp; Al w...</td>\n",
       "      <td>Saurav Kant, an alumnus of upGrad and IIIT-B's...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Delhi techie wins free food from Swiggy for on...</td>\n",
       "      <td>Kunal Shah's credit card bill payment platform...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>New Zealand end Rohit Sharma-led India's 12-ma...</td>\n",
       "      <td>New Zealand defeated India by 8 wickets in the...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Aegon life iTerm insurance plan helps customer...</td>\n",
       "      <td>With Aegon Life iTerm Insurance plan, customer...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Have known Hirani for yrs, what if MeToo claim...</td>\n",
       "      <td>Speaking about the sexual harassment allegatio...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                           headlines  \\\n",
       "0  upGrad learner switches to career in ML & Al w...   \n",
       "1  Delhi techie wins free food from Swiggy for on...   \n",
       "2  New Zealand end Rohit Sharma-led India's 12-ma...   \n",
       "3  Aegon life iTerm insurance plan helps customer...   \n",
       "4  Have known Hirani for yrs, what if MeToo claim...   \n",
       "\n",
       "                                                text  \n",
       "0  Saurav Kant, an alumnus of upGrad and IIIT-B's...  \n",
       "1  Kunal Shah's credit card bill payment platform...  \n",
       "2  New Zealand defeated India by 8 wickets in the...  \n",
       "3  With Aegon Life iTerm Insurance plan, customer...  \n",
       "4  Speaking about the sexual harassment allegatio...  "
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "text = data[\"text\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Saurav Kant, an alumnus of upGrad and IIIT-B's PG Program in Machine learning and Artificial Intelligence, was a Sr Systems Engineer at Infosys with almost 5 years of work experience. The program and upGrad's 360-degree career support helped him transition to a Data Scientist at Tech Mahindra with 90% salary hike. upGrad's Online Power Learning has powered 3 lakh+ careers.\""
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "text[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk.tokenize import sent_tokenize, word_tokenize\n",
    "from nltk.corpus import stopwords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "clear_text = []\n",
    "for i in range(len(text)):\n",
    "    clear_text.append(text[i].lower().replace(\".\",\"\").replace(\",\",\"\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"saurav kant an alumnus of upgrad and iiit-b's pg program in machine learning and artificial intelligence was a sr systems engineer at infosys with almost 5 years of work experience the program and upgrad's 360-degree career support helped him transition to a data scientist at tech mahindra with 90% salary hike upgrad's online power learning has powered 3 lakh+ careers\""
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clear_text[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package punkt to\n",
      "[nltk_data]     C:\\Users\\Videet\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package punkt is already up-to-date!\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import nltk\n",
    "nltk.download('punkt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "tokens = []\n",
    "for i in range(len(clear_text)):\n",
    "    tokens.append(word_tokenize(clear_text[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "summary = data[\"headlines\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "clear_summary = []\n",
    "for i in range(len(summary)):\n",
    "    clear_summary.append(summary[i].lower().replace(\".\",\"\").replace(\",\",\"\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'upgrad learner switches to career in ml & al with 90% salary hike'"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clear_summary[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_len = 0\n",
    "for i in range(len(tokens)):\n",
    "    if max_len <= len(tokens[i]):\n",
    "        max_len = len(tokens[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "85"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max_len"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "summary_token = []\n",
    "for i in range(len(clear_summary)):\n",
    "    summary_token.append(word_tokenize(clear_summary[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_len_sum = 0\n",
    "for i in range(len(summary_token)):\n",
    "    if max_len_sum <= len(summary_token[i]):\n",
    "        max_len_sum = len(summary_token[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "19"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max_len_sum"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "vocab = []\n",
    "for i in range(len(tokens)):\n",
    "    for w in tokens[i]:\n",
    "        vocab.append(w)\n",
    "vocab = list(set(vocab))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "vocab_size = len(vocab)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "83769"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vocab_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Model\n",
    "from keras.layers import Input\n",
    "from keras.layers import LSTM\n",
    "from keras.layers import Dense\n",
    "from keras.utils.vis_utils import plot_model\n",
    "from random import random\n",
    "from numpy import array\n",
    "from numpy import cumsum\n",
    "from matplotlib import pyplot\n",
    "from pandas import DataFrame\n",
    "from keras.models import Sequential\n",
    "from keras.layers import LSTM, Embedding, RepeatVector\n",
    "from keras.layers import Dense\n",
    "from keras.layers import TimeDistributed\n",
    "from keras.layers import Bidirectional\n",
    "import keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs = Input(shape=(max_len,))\n",
    "encoder1 = Embedding(vocab_size+1, 128)(inputs)\n",
    "encoder2 = LSTM(128)(encoder1)\n",
    "encoder3 = RepeatVector(max_len_sum)(encoder2)\n",
    "# decoder output model\n",
    "decoder1 = LSTM(128, return_sequences=True)(encoder3)\n",
    "outputs = TimeDistributed(Dense(vocab_size+1, activation='softmax'))(decoder1)\n",
    "# tie it together\n",
    "model = Model(inputs=inputs, outputs=outputs)\n",
    "model.compile(loss='categorical_crossentropy', optimizer='adam')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(len(summary_token)):\n",
    "    for _ in range(max_len_sum-len(summary_token[i])):\n",
    "                   summary_token[i].append(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "tokens = np.asarray(tokens)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(50000,)"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tokens.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "33285\n"
     ]
    }
   ],
   "source": [
    "vocab_sum = []\n",
    "for i in range(len(summary_token)):\n",
    "    for w in summary_token[i]:\n",
    "        vocab_sum.append(w)\n",
    "vocab_sum = list(set(vocab_sum))\n",
    "vocab_sum_size = len(vocab_sum)\n",
    "print(vocab_sum_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(len(tokens)):\n",
    "    tokens[i] = np.asarray(tokens[i])\n",
    "    #tokens[i] = np.reshape(tokens[i], (len(tokens[i]),1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sum_final = []\n",
    "for i in range(len(summary_token)):\n",
    "    single = []\n",
    "    for w in summary_token[i]:\n",
    "        temp = [0]*(vocab_size+1)\n",
    "        flag = False\n",
    "        for k in range(len(vocab)):\n",
    "            if w == vocab[k]:\n",
    "                temp[k] = 1\n",
    "                flag = True\n",
    "                break\n",
    "        if not flag:\n",
    "            temp[vocab_size] = 1\n",
    "        temp = np.asarray(temp)\n",
    "        single.append(temp)\n",
    "    sum_final.append(single)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tk = []\n",
    "for i in range(len(tokens)):\n",
    "    single = []\n",
    "    for w in tokens[i]:\n",
    "        for k in range(len(vocab)):\n",
    "            if w == vocab[k]:\n",
    "                single.append(k+1)\n",
    "                break\n",
    "    #single = np.asarray(single)\n",
    "    tk.append(single)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(len(tk)):\n",
    "    for _ in range(max_len-len(tk[i])):\n",
    "                   tk[i].append(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sum_final = np.asarray(sum_final)\n",
    "tk = np.asarray(tk)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(len(tk)):\n",
    "    tk[i] = np.asarray(tk[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.fit([tk], sum_final,\n",
    "          batch_size=256,\n",
    "          epochs=15,\n",
    "          validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(sum_final)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tk.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sum_final.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(\"news_summary_more.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = data[50000:50001]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = a[\"text\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = data[\"text\"][50002]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "b = []\n",
    "for w in a:\n",
    "    for k in range(len(vocab)):\n",
    "        if w == vocab[k]:\n",
    "            b.append(k+1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(max_len-len(b)):\n",
    "    b.append(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_len"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "s = a.lower().replace(\".\",\"\").replace(\",\",\"\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "j = word_tokenize(s)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(j)\n",
    "b = []\n",
    "for w in j:\n",
    "    for k in range(len(vocab)):\n",
    "        if w == vocab[k]:\n",
    "            b.append(k+1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(max_len-len(b)):\n",
    "    b.append(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "c = model.predict(b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "b = np.asarray(b)\n",
    "b = np.reshape(b, (1,78))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "b.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "v = np.max(a[0],axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "v"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "v[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = np.array(a)\n",
    "n = []\n",
    "for i in range(18):\n",
    "    for j in range(10248):\n",
    "        if v[i] == a[0][i][j]:\n",
    "            n.append(1)\n",
    "        else:\n",
    "            n.append(0)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n = np.reshape(n,(1,18,10248))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}