05-collaborative-filtering-embedding.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Collaborative Filtering Embedding\n",
    "\n",
    "Now we will take advantage of the user interactions to represent movies in an embedding space. We don't need any movie attribute here, only the user ratings."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from scipy.sparse import coo_matrix\n",
    "from lightfm import LightFM\n",
    "from sklearn.preprocessing import normalize\n",
    "\n",
    "from src.factorization import describe_csr_matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Building the Utility Matrix\n",
    "\n",
    "First of all, we need to put data into an user/movie matrix format. Users will be represented as rows, movies as columns and like(+1)/dislike(-1) as values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/recommender/anaconda3/envs/movie-similarity/lib/python3.7/site-packages/numpy/lib/arraysetops.py:580: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
      "  mask |= (ar1 == a)\n"
     ]
    },
    {
     "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>id</th>\n",
       "      <th>user</th>\n",
       "      <th>rating</th>\n",
       "      <th>title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Jumanji</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>14</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Jumanji</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>39</td>\n",
       "      <td>3.5</td>\n",
       "      <td>Jumanji</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>42</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Jumanji</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2</td>\n",
       "      <td>45</td>\n",
       "      <td>0.5</td>\n",
       "      <td>Jumanji</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id  user  rating    title\n",
       "0   2     4     4.0  Jumanji\n",
       "1   2    14     4.0  Jumanji\n",
       "2   2    39     3.5  Jumanji\n",
       "3   2    42     3.0  Jumanji\n",
       "4   2    45     0.5  Jumanji"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratings = pd.read_csv('data/movie_ratings.csv', index_col='id')\n",
    "movies = pd.read_csv('data/movie_info.csv', usecols=['id','title'], index_col='id')\n",
    "\n",
    "ratings = ratings.join(movies, how='inner')\n",
    "ratings.reset_index(drop=False, inplace=True)\n",
    "\n",
    "ratings.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We need to convert idss to indexes in order to crete the sparse matrix:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>id</th>\n",
       "      <th>user</th>\n",
       "      <th>rating</th>\n",
       "      <th>title</th>\n",
       "      <th>user_idx</th>\n",
       "      <th>movie_idx</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Jumanji</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>14</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Jumanji</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>39</td>\n",
       "      <td>3.5</td>\n",
       "      <td>Jumanji</td>\n",
       "      <td>35</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>42</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Jumanji</td>\n",
       "      <td>37</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2</td>\n",
       "      <td>45</td>\n",
       "      <td>0.5</td>\n",
       "      <td>Jumanji</td>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id  user  rating    title  user_idx  movie_idx\n",
       "0   2     4     4.0  Jumanji         3          0\n",
       "1   2    14     4.0  Jumanji        12          0\n",
       "2   2    39     3.5  Jumanji        35          0\n",
       "3   2    42     3.0  Jumanji        37          0\n",
       "4   2    45     0.5  Jumanji        40          0"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratings['user_idx'] = ratings['user'].astype('category').cat.codes\n",
    "ratings['movie_idx'] = ratings['id'].astype('category').cat.codes\n",
    "\n",
    "ratings.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ratings above 3.5 will be considered as likes (+1) and the remaining ones as dislikes (-1):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " 1.0    15810170\n",
       "-1.0     9643061\n",
       "Name: value, dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratings['value'] = np.where(ratings['rating'] >= 3.5, 1.0, -1.0)\n",
    "ratings['value'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So we will crete a +1/-1 sparse matrix to represent user interactions with the movies:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "273627 x 10698 sparse matrix with 99.13% of sparsity.\n"
     ]
    }
   ],
   "source": [
    "values = ratings['value'].values\n",
    "idx = (ratings.user_idx.values, ratings.movie_idx.values,)\n",
    "dim = (ratings.user_idx.max()+1, ratings.movie_idx.max()+1)\n",
    "x = coo_matrix((values, idx), shape=dim)\n",
    "\n",
    "print(describe_csr_matrix(x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Factorizing the Utility Matrix\n",
    "\n",
    "By factorizing the utility matrix we will get user factors and movie factors. The last ones will be used as embedding features to represent movies. As in the content based approach, we will get 300 factors:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 33min 13s, sys: 1.06 s, total: 33min 14s\n",
      "Wall time: 4min 17s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<lightfm.lightfm.LightFM at 0x7f12fe7959d0>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "N_COMPONENTS = 300\n",
    "lfm = LightFM(no_components=N_COMPONENTS, loss='logistic', random_state=0)\n",
    "lfm.fit(x, epochs=5, num_threads=8, verbose=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since we have +1/-1 values, we applied a Logistic Loss. Then, we can extract movie factors and bias:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10698, 300)\n",
      "[[ 0.13717581 -0.03386943 -0.03017038  0.05062018  0.03954934]\n",
      " [ 0.03843264  0.02700372 -0.01565084  0.02153345 -0.09214582]\n",
      " [ 0.07393657 -0.06563244 -0.05454138  0.03856608  0.0649413 ]\n",
      " [ 0.05018987 -0.07951837 -0.03726879  0.05435565 -0.09216952]\n",
      " [ 0.00977158  0.00257648 -0.04626083 -0.03330302 -0.04443587]]\n"
     ]
    }
   ],
   "source": [
    "movie_bias, movie_embedding = lfm.get_item_representations()\n",
    "movie_embedding = normalize(movie_embedding, norm=\"l2\", axis=1, copy=False)\n",
    "\n",
    "print(movie_embedding.shape)\n",
    "print(movie_embedding[:5, :5])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's check the bias term against the number of votes, number of likes and average rating:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 3.46 s, sys: 1.57 s, total: 5.03 s\n",
      "Wall time: 4.63 s\n"
     ]
    },
    {
     "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>id</th>\n",
       "      <th>title</th>\n",
       "      <th>num_votes</th>\n",
       "      <th>num_likes</th>\n",
       "      <th>num_dislikes</th>\n",
       "      <th>avg_rating</th>\n",
       "      <th>bias</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>Jumanji</td>\n",
       "      <td>27138</td>\n",
       "      <td>12345.0</td>\n",
       "      <td>14793.0</td>\n",
       "      <td>3.246499</td>\n",
       "      <td>-0.311027</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3</td>\n",
       "      <td>Grumpier Old Men</td>\n",
       "      <td>15579</td>\n",
       "      <td>5861.0</td>\n",
       "      <td>9718.0</td>\n",
       "      <td>3.173792</td>\n",
       "      <td>-0.631766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4</td>\n",
       "      <td>Waiting to Exhale</td>\n",
       "      <td>2989</td>\n",
       "      <td>848.0</td>\n",
       "      <td>2141.0</td>\n",
       "      <td>2.874540</td>\n",
       "      <td>-0.952873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5</td>\n",
       "      <td>Father of the Bride Part II</td>\n",
       "      <td>15466</td>\n",
       "      <td>5064.0</td>\n",
       "      <td>10402.0</td>\n",
       "      <td>3.077169</td>\n",
       "      <td>-1.132931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>Heat</td>\n",
       "      <td>28680</td>\n",
       "      <td>20049.0</td>\n",
       "      <td>8631.0</td>\n",
       "      <td>3.844125</td>\n",
       "      <td>0.970788</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id                        title  num_votes  num_likes  num_dislikes  \\\n",
       "0   2                      Jumanji      27138    12345.0       14793.0   \n",
       "1   3             Grumpier Old Men      15579     5861.0        9718.0   \n",
       "2   4            Waiting to Exhale       2989      848.0        2141.0   \n",
       "3   5  Father of the Bride Part II      15466     5064.0       10402.0   \n",
       "4   6                         Heat      28680    20049.0        8631.0   \n",
       "\n",
       "   avg_rating      bias  \n",
       "0    3.246499 -0.311027  \n",
       "1    3.173792 -0.631766  \n",
       "2    2.874540 -0.952873  \n",
       "3    3.077169 -1.132931  \n",
       "4    3.844125  0.970788  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "ratings['like'] = np.where(ratings['rating'] >= 3.5, 1.0, 0.0)\n",
    "ratings['dislike'] = np.where(ratings['rating'] < 3.5, 1.0, 0.0)\n",
    "\n",
    "movie_votes = ratings.groupby(['id']).agg(\n",
    "    title = pd.NamedAgg(column='title', aggfunc='first'),\n",
    "    num_votes = pd.NamedAgg(column='user_idx', aggfunc='count'),\n",
    "    num_likes = pd.NamedAgg(column='like', aggfunc='sum'),\n",
    "    num_dislikes = pd.NamedAgg(column='dislike', aggfunc='sum'),\n",
    "    avg_rating = pd.NamedAgg(column='rating', aggfunc='mean'),\n",
    ").reset_index(drop=False)\n",
    "movie_votes['bias'] = movie_bias\n",
    "\n",
    "movie_votes.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='num_votes', ylabel='bias'>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "movie_votes.plot.scatter(x='num_votes', y='bias')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='num_likes', ylabel='bias'>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "movie_votes.plot.scatter(x='num_likes', y='bias')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='num_dislikes', ylabel='bias'>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "movie_votes.plot.scatter(x='num_dislikes', y='bias')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='avg_rating', ylabel='bias'>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "movie_votes.plot.scatter(x='avg_rating', y='bias')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can observe a strong correlation between average rating and the bias term learned in the factorization step."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualizing the Embedding Space\n",
    "\n",
    "Let's check the learned factors using the Embedding Projector:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 4.6 s, sys: 168 ms, total: 4.77 s\n",
      "Wall time: 4.76 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "movie_votes[['id','title']].to_csv('output/cb_embedding_meta.tsv', sep='\\t', header=True, index=False)\n",
    "pd.DataFrame(movie_embedding).to_csv('output/collab_filt_vectors.tsv', sep='\\t', \n",
    "                                     float_format='%.5f', header=False, index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](img/collab_filt_lord_rings.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](img/collab_filt_star_wars.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](img/collab_filt_pulp_fiction.png)"
   ]
  }
 ],
 "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": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}