Semantic_similarity.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import pandas as pd\n",
    "import requests\n",
    "# import nltk\n",
    "# import nltk.data\n",
    "# import gensim\n",
    "# from nltk.tokenize import word_tokenize, sent_tokenize\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Comparing text similarity "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "messages = [\n",
    "    \n",
    "    # Energy \n",
    "     \"Implement massive energy efficiency practices\",\n",
    "     \"Replace fossil fuels with low-carbon renewables\",\n",
    "     \"Apply strong carbon taxes to cut fossil fuel use\",\n",
    "     \"We should leave remaining stocks of fossil fuels in the ground \",\n",
    "     \"We must swiftly eliminate subsidies for fossil fuels\",\n",
    "     \n",
    "     \n",
    "    # Diets\n",
    "     \"Eat mostly plant-based foods while reducing the global consumption of animal products\",\n",
    "     \"Encourage vegeterian and vegan diets\",\n",
    "     \"Free up croplands for growing human plant food instead of livestock feed\", \n",
    "     \"Release some grazing land to support natural climate solution\",\n",
    "    \n",
    "    \"Increase use of renewable energy sources for housing and public amenities\",\n",
    "    \n",
    "    # COM[0] text\n",
    "    \"Fossil free energy production\",\n",
    "\n",
    "    \n",
    "    # Random unrelated text (control)\n",
    "    (\"Word and sentence embeddings have become an essential part of any Deep-Learning-based natural language \"\n",
    "    \"processing systems.\"),\n",
    "]\n",
    "\n",
    "urgent_messages = [\n",
    "    \n",
    "    # example of urgent climate-related language\n",
    "    \"we have to act now\",\n",
    "    \"time for action is running out\",\n",
    "    \"planet Earth is facing a climate emergency\",\n",
    "    \"alarming trends for climate change made it urgently necessary to act\",\n",
    "    \"funds need to be allocated now\",\n",
    "    \n",
    "    #\"Avoid untold suffering due to the climate crisis\",\n",
    "    \"The climate crisis has arrived and is accelerating faster than most scientists expected\",\n",
    "    \"It is more severe than anticipated, threatening natural ecosystems and the fate of humanity\",\n",
    "    \n",
    "    # Example from Dublin's conclusion \n",
    "    \n",
    "    \"investment must start without delay\",\n",
    "    \"clear that a business-as-usual scenario is not an acceptable option, even in the short term\",    \n",
    "]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Simple document comparison with TF-IDF "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1.         0.71126733]\n",
      " [0.71126733 1.        ]]\n"
     ]
    }
   ],
   "source": [
    "tfidf = TfidfVectorizer()\n",
    "\n",
    "vecs = tfidf.fit_transform([crisis_demo_list[1], crisis_demo_list[3]])\n",
    "corr_matrix = ((vecs * vecs.T).A)\n",
    "print(corr_matrix)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "# Universal Sentence Encoder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Universal Sentence Encoder encodes text into high dimensional vectors that can be used for text classification, semantic similarity, clustering and other natural language tasks. The input is variable length English text and the output is a 512 dimensional vector. \n",
    "\n",
    "The text doesn't need to be preprocessed to get vector embeddings, but it might be worth exploring whether or not this makes a difference."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/sofija/anaconda3/lib/python3.6/site-packages/h5py/__init__.py:34: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import tensorflow_hub as hub\n",
    "import numpy as np \n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'2.0.0'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#module_url_tf2 = \"https://tfhub.dev/google/universal-sentence-encoder/4\"\n",
    "\n",
    "# Import the Universal Sentence Encoder's TF Hub module\n",
    "# USE large model trained with a Transformer encoder\n",
    "\n",
    "module_url_transformer = \"https://tfhub.dev/google/universal-sentence-encoder-large/4\"\n",
    "\n",
    "# Import the Universal Sentence Encoder's TF Hub module\n",
    "\n",
    "embed = hub.load(module_url_transformer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_embeddings(texts, output=None):\n",
    "    \"\"\"Uses Universal Sentence Encoder to get 512-dimensional\n",
    "        vector representations of the input texts with TF2.0.\n",
    "        To return numpy array set output = 'numpy' \"\"\"\n",
    "    if type(texts) is str:\n",
    "        texts = [texts]  \n",
    "    if output == 'numpy':\n",
    "        return embed(texts)['outputs'].numpy()\n",
    "    else:\n",
    "        return embed(texts)\n",
    "    \n",
    "def cosine_similarity(v1, v2):\n",
    "    \"\"\"Calculates (cosine) similarity between vectors v1 and v2.\"\"\"\n",
    "    mag1 = np.linalg.norm(v1)\n",
    "    mag2 = np.linalg.norm(v2)\n",
    "    if (not mag1) or (not mag2):\n",
    "        return 0\n",
    "    return np.dot(v1, v2) / (mag1 * mag2)\n",
    "\n",
    "\n",
    "def test_similarity(text1, text2):\n",
    "    vec1 = get_embeddings(text1)['outputs'].numpy()\n",
    "    vec2 = get_embeddings(text2)['outputs'].numpy()\n",
    "    return cosine_similarity(vec1, vec2)\n",
    "    \n",
    "    \n",
    "def semantic_search(query, data, vectors):\n",
    "    \"\"\"Finds the text in data that's most semantically \n",
    "        similar to the query.\"\"\"\n",
    "    #query = process_text(query)\n",
    "    print(\"Extracting features...\")\n",
    "    query_vec = get_embeddings(query)['outputs'].numpy().ravel()\n",
    "    res = []\n",
    "    for i, d in enumerate(data):\n",
    "        qvec = vectors[i].ravel()\n",
    "        sim = cosine_similarity(query_vec, qvec)\n",
    "        res.append((sim, d[:100], i))\n",
    "    return sorted(res, key=lambda x : x[0], reverse=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Implement massive energy efficiency practices',\n",
       " 'Replace fossil fuels with low-carbon renewables',\n",
       " 'Apply strong carbon taxes to cut fossil fuel use',\n",
       " 'We should leave remaining stocks of fossil fuels in the ground ',\n",
       " 'We must swiftly eliminate subsidies for fossil fuels',\n",
       " 'Eat mostly plant-based foods while reducing the global consumption of animal products',\n",
       " 'Encourage vegeterian and vegan diets',\n",
       " 'Free up croplands for growing human plant food instead of livestock feed',\n",
       " 'Release some grazing land to support natural climate solution',\n",
       " 'Increase use of renewable energy sources for housing and public amenities',\n",
       " 'Fossil free energy production',\n",
       " 'Word and sentence embeddings have become an essential part of any Deep-Learning-based natural language processing systems.']"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "messages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "example_message_embeddings = get_embeddings(messages, 'numpy') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "urgent_message_embeddings = get_embeddings(urgent_messages, 'numpy') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdAAAAF8CAYAAAB/g+qVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZbBcEDjIHgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1666f7c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def plot_similarity(texts, embeddings, rotation):\n",
    "    corr = np.inner(embeddings, embeddings)\n",
    "    sns.set(font_scale=1.2)\n",
    "    g = sns.heatmap(\n",
    "      corr,\n",
    "      xticklabels=texts,\n",
    "      yticklabels=texts,\n",
    "      vmin=0,\n",
    "      vmax=1,\n",
    "      cmap=\"YlOrRd\")\n",
    "    g.set_xticklabels([elem[:20] for elem in texts], rotation=rotation)\n",
    "    g.set_yticklabels([elem[:20] for elem in texts], rotation=0)\n",
    "    g.set_title(\"Semantic Textual Similarity\")\n",
    "    \n",
    "#plot_similarity(urgent_messages, urgent_message_embeddings, 90)\n",
    "plot_similarity(messages, example_message_embeddings, 90)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# References\n",
    "\n",
    "Semantic similarity using USE: \n",
    "https://www.nishanpantha.com.np/programming/universal-sentence-encoder-semantic-search.html\n",
    "\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "guardian",
   "language": "python",
   "name": "guardian"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}