cosine_similarity.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "8716180d",
   "metadata": {},
   "source": [
    "# Course Recommendation using Cosine Similarity\n",
    "Cosine similarity is a commonly used similarity metric in machine learning and natural language processing. It measures the similarity between two vectors by computing the cosine of the angle between them. The two vectors are typically represented as lists of numerical values, and each element in the list corresponds to a different feature or dimension.\n",
    "![Vector Similarity](https://www.machinelearningplus.com/wp-content/uploads/2018/10/soft-cosine.png)\n",
    "\n",
    "The cosine similarity metric is particularly useful for comparing text documents or word embeddings in natural language processing. In this context, each document or word is represented as a vector of numerical values, where each value represents the importance or frequency of a particular word or feature. By computing the cosine similarity between two vectors, we can determine how similar the two documents or words are based on the words they share in common.\n",
    "\n",
    "The cosine similarity metric ranges from -1 to 1, with -1 indicating completely opposite vectors, 0 indicating vectors that are orthogonal (perpendicular) to each other, and 1 indicating identical vectors.\n",
    "\n",
    "To calculate the cosine similarity between two vectors, we use the following formula:\n",
    "$$\\text{cosine_similarity}(\\mathbf{x}, \\mathbf{y}) = \\frac{\\mathbf{x} \\cdot \\mathbf{y}}{\\|\\mathbf{x}\\|\\|\\mathbf{y}\\|}$$\n",
    "\n",
    "Then, we calculate the cosine distance between two vectors using the formula:\n",
    "$$\\text{cosine_distance}(\\mathbf{x}, \\mathbf{y}) = 1 - \\text{cosine_similarity}(\\mathbf{x}, \\mathbf{y})$$\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "id": "68fd5af0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load data analysis packages\n",
    "import pandas as pd\n",
    "import neattext.functions as nfx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "id": "5f7093cd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load ML packages\n",
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer\n",
    "from sklearn.metrics.pairwise import cosine_similarity, linear_kernel\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "id": "9a13df93",
   "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>course_name</th>\n",
       "      <th>university</th>\n",
       "      <th>difficulty_level</th>\n",
       "      <th>course_rating</th>\n",
       "      <th>course_url</th>\n",
       "      <th>course_description</th>\n",
       "      <th>course_skills</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Write A Feature Length Screenplay For Film Or ...</td>\n",
       "      <td>Michigan State University</td>\n",
       "      <td>Beginner</td>\n",
       "      <td>4.8</td>\n",
       "      <td>https://www.coursera.org/learn/write-a-feature...</td>\n",
       "      <td>Write a Full Length Feature Film Script  In th...</td>\n",
       "      <td>Drama  Comedy  peering  screenwriting  film  D...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Business Strategy: Business Model Canvas Analy...</td>\n",
       "      <td>Coursera Project Network</td>\n",
       "      <td>Beginner</td>\n",
       "      <td>4.8</td>\n",
       "      <td>https://www.coursera.org/learn/canvas-analysis...</td>\n",
       "      <td>By the end of this guided project, you will be...</td>\n",
       "      <td>Finance  business plan  persona (user experien...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Silicon Thin Film Solar Cells</td>\n",
       "      <td>�cole Polytechnique</td>\n",
       "      <td>Advanced</td>\n",
       "      <td>4.1</td>\n",
       "      <td>https://www.coursera.org/learn/silicon-thin-fi...</td>\n",
       "      <td>This course consists of a general presentation...</td>\n",
       "      <td>chemistry  physics  Solar Energy  film  lambda...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Finance for Managers</td>\n",
       "      <td>IESE Business School</td>\n",
       "      <td>Intermediate</td>\n",
       "      <td>4.8</td>\n",
       "      <td>https://www.coursera.org/learn/operational-fin...</td>\n",
       "      <td>When it comes to numbers, there is always more...</td>\n",
       "      <td>accounts receivable  dupont analysis  analysis...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Retrieve Data using Single-Table SQL Queries</td>\n",
       "      <td>Coursera Project Network</td>\n",
       "      <td>Beginner</td>\n",
       "      <td>4.6</td>\n",
       "      <td>https://www.coursera.org/learn/single-table-sq...</td>\n",
       "      <td>In this course you�ll learn how to effectively...</td>\n",
       "      <td>Data Analysis  select (sql)  database manageme...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         course_name   \n",
       "0  Write A Feature Length Screenplay For Film Or ...  \\\n",
       "1  Business Strategy: Business Model Canvas Analy...   \n",
       "2                      Silicon Thin Film Solar Cells   \n",
       "3                               Finance for Managers   \n",
       "4       Retrieve Data using Single-Table SQL Queries   \n",
       "\n",
       "                  university difficulty_level course_rating   \n",
       "0  Michigan State University         Beginner           4.8  \\\n",
       "1   Coursera Project Network         Beginner           4.8   \n",
       "2        �cole Polytechnique         Advanced           4.1   \n",
       "3       IESE Business School     Intermediate           4.8   \n",
       "4   Coursera Project Network         Beginner           4.6   \n",
       "\n",
       "                                          course_url   \n",
       "0  https://www.coursera.org/learn/write-a-feature...  \\\n",
       "1  https://www.coursera.org/learn/canvas-analysis...   \n",
       "2  https://www.coursera.org/learn/silicon-thin-fi...   \n",
       "3  https://www.coursera.org/learn/operational-fin...   \n",
       "4  https://www.coursera.org/learn/single-table-sq...   \n",
       "\n",
       "                                  course_description   \n",
       "0  Write a Full Length Feature Film Script  In th...  \\\n",
       "1  By the end of this guided project, you will be...   \n",
       "2  This course consists of a general presentation...   \n",
       "3  When it comes to numbers, there is always more...   \n",
       "4  In this course you�ll learn how to effectively...   \n",
       "\n",
       "                                       course_skills  \n",
       "0  Drama  Comedy  peering  screenwriting  film  D...  \n",
       "1  Finance  business plan  persona (user experien...  \n",
       "2  chemistry  physics  Solar Energy  film  lambda...  \n",
       "3  accounts receivable  dupont analysis  analysis...  \n",
       "4  Data Analysis  select (sql)  database manageme...  "
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load the CSV data\n",
    "df = pd.read_csv(\"data/coursera_data.csv\")\n",
    "df.head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "74120e6a",
   "metadata": {},
   "source": [
    "### Cleaning Data\n",
    "As can be seen above, the `course_name` consists of many filler words like `and, or, the`. We don't want these filler words in our vectors. So, we use the `neattext` library to remove such words from our data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "aa5eb91b",
   "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>course_name</th>\n",
       "      <th>clean_course_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Write A Feature Length Screenplay For Film Or ...</td>\n",
       "      <td>Write Feature Length Screenplay Film Television</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Business Strategy: Business Model Canvas Analy...</td>\n",
       "      <td>Business Strategy Business Model Canvas Analys...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Silicon Thin Film Solar Cells</td>\n",
       "      <td>Silicon Thin Film Solar Cells</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Finance for Managers</td>\n",
       "      <td>Finance Managers</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Retrieve Data using Single-Table SQL Queries</td>\n",
       "      <td>Retrieve Data SingleTable SQL Queries</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3517</th>\n",
       "      <td>Capstone: Retrieving, Processing, and Visualiz...</td>\n",
       "      <td>Capstone Retrieving Processing Visualizing Dat...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3518</th>\n",
       "      <td>Patrick Henry: Forgotten Founder</td>\n",
       "      <td>Patrick Henry Forgotten Founder</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3519</th>\n",
       "      <td>Business intelligence and data analytics: Gene...</td>\n",
       "      <td>Business intelligence data analytics Generate ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3520</th>\n",
       "      <td>Rigid Body Dynamics</td>\n",
       "      <td>Rigid Body Dynamics</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3521</th>\n",
       "      <td>Architecting with Google Kubernetes Engine: Pr...</td>\n",
       "      <td>Architecting Google Kubernetes Engine Production</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3522 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                            course_name   \n",
       "0     Write A Feature Length Screenplay For Film Or ...  \\\n",
       "1     Business Strategy: Business Model Canvas Analy...   \n",
       "2                         Silicon Thin Film Solar Cells   \n",
       "3                                  Finance for Managers   \n",
       "4          Retrieve Data using Single-Table SQL Queries   \n",
       "...                                                 ...   \n",
       "3517  Capstone: Retrieving, Processing, and Visualiz...   \n",
       "3518                   Patrick Henry: Forgotten Founder   \n",
       "3519  Business intelligence and data analytics: Gene...   \n",
       "3520                                Rigid Body Dynamics   \n",
       "3521  Architecting with Google Kubernetes Engine: Pr...   \n",
       "\n",
       "                                      clean_course_name  \n",
       "0       Write Feature Length Screenplay Film Television  \n",
       "1     Business Strategy Business Model Canvas Analys...  \n",
       "2                         Silicon Thin Film Solar Cells  \n",
       "3                                      Finance Managers  \n",
       "4                 Retrieve Data SingleTable SQL Queries  \n",
       "...                                                 ...  \n",
       "3517  Capstone Retrieving Processing Visualizing Dat...  \n",
       "3518                    Patrick Henry Forgotten Founder  \n",
       "3519  Business intelligence data analytics Generate ...  \n",
       "3520                                Rigid Body Dynamics  \n",
       "3521   Architecting Google Kubernetes Engine Production  \n",
       "\n",
       "[3522 rows x 2 columns]"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Remove stop words from data\n",
    "df['clean_course_name'] = df['course_name'].apply(nfx.remove_stopwords)\n",
    "# Remove special characters\n",
    "df['clean_course_name'] = df['clean_course_name'].apply(nfx.remove_special_characters)\n",
    "df[['course_name','clean_course_name']]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "2536b19c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Vectorize our Text\n",
    "count_vect = CountVectorizer()\n",
    "cv_mat = count_vect.fit_transform(df['clean_course_name'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "id": "6dcd479d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "matrix([[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]])"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Dense Matrix\n",
    "cv_mat.todense()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "id": "28f3ff12",
   "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>10</th>\n",
       "      <th>10000</th>\n",
       "      <th>101</th>\n",
       "      <th>102</th>\n",
       "      <th>1650</th>\n",
       "      <th>17002000</th>\n",
       "      <th>1760</th>\n",
       "      <th>1910</th>\n",
       "      <th>19621974</th>\n",
       "      <th>1e</th>\n",
       "      <th>...</th>\n",
       "      <th>young</th>\n",
       "      <th>your</th>\n",
       "      <th>yourself</th>\n",
       "      <th>youth</th>\n",
       "      <th>yuan</th>\n",
       "      <th>zap</th>\n",
       "      <th>zcurve</th>\n",
       "      <th>zero</th>\n",
       "      <th>zoning</th>\n",
       "      <th>zos</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 3602 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   10  10000  101  102  1650  17002000  1760  1910  19621974  1e  ...  young   \n",
       "0   0      0    0    0     0         0     0     0         0   0  ...      0  \\\n",
       "1   0      0    0    0     0         0     0     0         0   0  ...      0   \n",
       "2   0      0    0    0     0         0     0     0         0   0  ...      0   \n",
       "3   0      0    0    0     0         0     0     0         0   0  ...      0   \n",
       "4   0      0    0    0     0         0     0     0         0   0  ...      0   \n",
       "\n",
       "   your  yourself  youth  yuan  zap  zcurve  zero  zoning  zos  \n",
       "0     0         0      0     0    0       0     0       0    0  \n",
       "1     0         0      0     0    0       0     0       0    0  \n",
       "2     0         0      0     0    0       0     0       0    0  \n",
       "3     0         0      0     0    0       0     0       0    0  \n",
       "4     0         0      0     0    0       0     0       0    0  \n",
       "\n",
       "[5 rows x 3602 columns]"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cv_words = pd.DataFrame(cv_mat.todense(),columns=count_vect.get_feature_names_out())\n",
    "df_cv_words.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "c2d02500",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cosine Similarity Matrix\n",
    "cosine_sim_mat = cosine_similarity(cv_mat)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "id": "8dd693ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.        , 0.        , 0.18257419, ..., 0.        , 0.        ,\n",
       "        0.        ],\n",
       "       [0.        , 1.        , 0.        , ..., 0.27216553, 0.        ,\n",
       "        0.        ],\n",
       "       [0.18257419, 0.        , 1.        , ..., 0.        , 0.        ,\n",
       "        0.        ],\n",
       "       ...,\n",
       "       [0.        , 0.27216553, 0.        , ..., 1.        , 0.        ,\n",
       "        0.        ],\n",
       "       [0.        , 0.        , 0.        , ..., 0.        , 1.        ,\n",
       "        0.        ],\n",
       "       [0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
       "        1.        ]])"
      ]
     },
     "execution_count": 137,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cosine_sim_mat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "id": "37ebe0af",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Use seaborn to generate heatmap\n",
    "import seaborn as sns\n",
    "sns.heatmap(cosine_sim_mat[0:10],annot=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "id": "e22a8f3b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiIAAAGdCAYAAAAvwBgXAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB400lEQVR4nO3deVjU57n/8fcsMOy7oMgiLrgRFY0Ss9lFo1lMamhrPJrF5rRKkqbU1BrbxOTYpMTjifU0vxRtmmgSaTQtGrNKNKc1m6JRUFERV1CRRWRnGIaZ7++PccYZGBAVmAHu13VxKcMw850B5vuZ57mf+1EpiqIghBBCCOECalcfgBBCCCH6LgkiQgghhHAZCSJCCCGEcBkJIkIIIYRwGQkiQgghhHAZCSJCCCGEcBkJIkIIIYRwGQkiQgghhHAZrasPoD1ms5ni4mL8/f1RqVSuPhwhhBBCdICiKNTW1hIZGYla3f6Yh1sHkeLiYqKjo119GEIIIYS4DmfPniUqKqrd67h1EPH39wcsDyQgIMDFRyOEEEKIjqipqSE6Otp2Hm+PWwcR63RMQECABBEhhBCih+lIWYUUqwohhBDCZSSICCGEEMJlJIgIIYQQwmUkiAghhBDCZSSICCGEEMJlJIgIIYQQwmUkiAghhBDCZSSICCGEEMJlJIgIIYQQwmUkiPRCyenfkFNU6fRrOUWVJKd/081HJIQQQjgnQaQXeu7eUfx4zS427SlyuDynqJL56/fy0wnREkaEEEK4BQkivdCOo6X89q54nt18yBZGrCEk5c7BvPDhYZ67d5SLj1IIIYRw803vxPV5/V8n8faAZ2cM59nNhzh9sZ6N350l5c7BpG07BkBiTLCLj1IIIYSQEZFeaXR/f/RGWP3FMf5jUjRrvjzFpNgQWwgZ3f/q2zKL67MyK5/iKr3TrxVX6VmZld/NRySEEO5Ngkgv9EnqnbYwkrHnLLcPDePzo6WAJYR8knqni4+w95qbFMuSzIOtwkhxlZ4lmQeZmxTroiMTQgj3JEGkl3rpwZtsP9yvT1y0Xf7IrYNccjx9RWSQNyuSxziEEWsIWZE8hsggbxcfoRBCuBcJIr2QtTB1yYzhDpcvvVwz0nI1jehc9mFkX+ElCSFCCNEOCSK90EufHHEoTPXUqABLzcizM4bzuy2H2uwzIjpHZJA3qVOHkZy+i9SpwySECCFEGySI9EKvzRnvUJha8PI9DgWsa+ZN4KVPjrj4KHu34io9q3ccJzNlMqt3HG+zgFUIIfo6CSK90Mw/fwk4FqbaF7A+m3mAzJTbXHmIvZp9TciE2JBWNSNCCCGukCDSC1U0NDtdHWMNIxUNzS46st7PWWGqswJWIYQQFipFURRXH0RbampqCAwMpLq6moCAAFcfjhBXtTIrn7lJsU5rQoqr9GRkF7J4+ggXHJkQQnSfazl/SxARQgghRKe6lvO3TM0IIYQQwmUkiAghhBDCZSSICCGEEMJlJIgIIYQQwmUkiAghhBDCZSSICCGEEMJlJIgIIYQQwmUkiAghhBDCZSSICCGEEMJlujSIvPjii6hUKoePESOkvbUQQgghLLRdfQejR49mx44dV+5Q2+V3KYQQQogeostTgVarpX///l19N0IIIYTogbq8RuT48eNERkYyePBg5s6dS1FRUZvXNRgM1NTUOHwIIYQQovfq0iCSlJTE+vXr2bZtG+np6Zw+fZo77riD2tpap9dPS0sjMDDQ9hEdHd2VhyeEEEIIF1MpiqJ0151VVVURGxvLqlWrePzxx1t93WAwYDAYbJ/X1NQQHR3doW2EhRBCCOEeampqCAwM7ND5u1srR4OCgoiPj+fEiRNOv67T6dDpdN15SEIIIYRwoW7tI1JXV8fJkycZMGBAd96tEEIIIdxUlwaR3/zmN+zcuZMzZ87w7bffMmvWLDQaDXPmzOnKuxVCCCFED9GlUzPnzp1jzpw5VFRU0K9fP26//XZ2795Nv379uvJuhRBCCNFDdGkQ2bhxY1fevBBCCCF6ONlrRgghhBAuI0FECCGEEC4jQUQIIYQQLiNBRAghhBAuI0FECCGEEC4jQUQIIYQQLiNBRAghhBAuI0FEXLOVWfkUV+mdfq24Ss/KrPxuPiIhhBA9lQSRPupGwsTcpFiWZB5s9f3FVXqWZB5kblJspx6rEEKI3kuCSA9iHx5aBgn78NCRUYkbCRORQd6sSB7j8P3W71uRPIbIIO/renxCCCH6HgkiPUhtYzOpm3IprtI7/L+4Sk/qplxqG5sd/t+eGw0T9t+/r/CShBAhhBDXRaUoiuLqg2hLTU0NgYGBVFdXExAQ4OrDcZmVWfm2EYrUjTmgUhEd7M1Xxy8S6O1Btd7IoFAflt4zkrTP8jEYTQwN96N/oBeLp49o97at4SN16jBW7zh+zWFiX+ElktN3kZkymQmxITf0OIUQQvQO13L+lhGRHsA6+gGw+qFEUBROlNfh46nmeFkd5XUGZiUOtIUQVHC2Us/cpNirTtNEBnmTOnUYyem7SJ067JpCSHGVntU7jpOZMpnVO463WXMihBBCtEWCSA+wcMoQDhdX859vfwdYwsi5S3rOVOhRAVo1LN2Sx6V6A6hAhYoR/f0Brlrvcb1hwn4aZ0JsSKtpHiGEEKIjJIj0AJFB3qz+6TiOldTwn29/x/KPD1PfZKkBCfXzwGiyXO9UeT3VDUZKa/TMShx41bqN6w0TzmpJnNWcCCGEEFcjQaSHmDa6P2vmTSC/pIZteaU0NZt5cNwALtYZ8bj8UzQrcKaigUcmx7Jqe0GHQ8i1homM7EKnt239/ozsws550EIIIXo9KVbtQZ55P5ePDxZjaFYI9fGkoqGJiABPymua0GpUNJkUdFoVhmaFtQ+PZ/roAW3elrUA1llQKa7Sk5FdeNVCVyGEEMIZKVbthYqr9OwsKMfQrHDXyAgqGppQYylkDQ/U0WRSiI/wxawoeKhhwbv72x3VWDx9RJujJZFB3hJChBBCdAsJIj2AtTdIo9FEQqQ/nx8tJdRHixlobDJTUm1g+sgICkrriQ72wWgGf52G1I05Uq8hhBDCrUkQ6QHW7DwJisLz947icHEt/QN0VDQ0E+hlCSPB3h5kHS0lQKfl1MUGksdFMijMlxEDAjpUPNpT9o7pKccphBCi4ySI9AD+Xlp+fsdglmw+xIBAL8pqDHioobqxmahAHZV6IwA1hmYGBurYcqCY+Ah/lj+Q0KHi0au1e7d2bHWmOwNAbWMzyenfklNU2eoYUjflUlLdKGFECCF6GAkiPcDi6SNY8+VJAIqrG1EAoxlUwLlqg8N1z1cb0KqgoLQWuFLv0V5guFq794VThrjFJncLpwwhPEDHo2/tsYURawgxNJs4e6lBNtwTQogeRoJID5GZcpvt/0qLf1vy9tQSHexDcvo3wOWT9cacdvefaW/vGHfZ5C4yyJv0uROI6+fLo2/tISuvxBZCdBo1qx9KlL1uhBCih5Eg0oP4eag6dL16QzNfHi/nuXtH2UIIKhULpwxp9/vaa/fuLpvcWcPIwGBvFmzYx6V6g4QQIYTowSSI9CCP3j4Y3VV+Yn6eGpoV8PXU8O7uQlI27AOVitWzxzmcqJ1N1Vyt3fuN7EvT2Ty1lifiRFk9j98RJyFECCF6KAkiPcjcpFhC/HXtXqeuycTdoyIoq2ti8/7znLxYz9K7R7QKIS1rOzrS7t0dNrmz1oQAjI0KZOQAfxb/42CrAlYhhBA9gwSRHiQjuxAvj/Z/ZJ5a2He2ihERlk3vooO8Sfssv93ajo60e3eHTe7sC1N1GjXp8ybw5qMTbTUjEkaEEKLnkSDSg9Q2GDl9sf0Tf1MzVNYbqKhvYu28CVTpjRiMJlI35bZZ23G1vWPW7DzpFpvcrdl5slVhqn0B6xMZ7XeTFUII4X4kiPQg72QXOXyu0zq/ntEMQ8J82ZBdyF/mjkfnoaFWb2yztuNq7d79vbRuscmdv5eWsVFBrQpTrWFk2qgI2XBPCCF6GNn0rgcZ9Owntv8HeWvp5+9FaU0jhqZmDGbL5XGhPpyvasBogr8+PIFpo/uTU1TJo+v2cOuQUOoNJpesdhFCCNF3yKZ3vZgWmDQomPvHDcTLQ83bP5vEsT/ei8/l0ZHYUF/GxYQwIyGCNV+epLhKz6rtBbw9fxJD+vm5pLajJ0pO/6bNmpOcokpbjxYhhBA3RkZEepiVWfnMTYp1OqKRlVfCgg37yEyZzITYkDabjrmiGVlPk1NUyfz1e/nhiHCeuWu47XmyXr7usYlEBHiRkV0oOxULIUQLMiLSi7VVz1FcpWdDdqHD0tqrFaFKPUXbEmOCWffYRLYfLSVlwz6Kq/StQkh3trcXQojeSkZEeoGWIxwy4tF5rPU14X46yuoMvD1/ki2EyPMrhBDOyYhIH9KRHiDi+iXGBLNkxghOlNcT7qfDrCgSQoQQohNJEOnhZPqla+UUVbIy6xgvz0qgpNbgFu3thRCiN5Eg0sNdrQeIFFJeP/uakO8PD2dImC9+XlqWf3RERpqEEKKTSBARwglnhanp8ybw7s8mcbqi3lbAKoQQ4sZIEBHCiZc+OeIQQqzTX4kxwbw9f5KtVkTCiBBC3BgJIr3Yyqz8Nk+UxVV6Vmbld/MR9RyZKbeRGBPstAYnMSaYj355h9TgCCFEJ+i2IPLKK6+gUqlITU3trrvs8+YmxTp9125daSM9MK5OanCEEKJrdUsQ2bt3L2vXrmXMmDHdcXfiMmfLeK+lx4iMqAghhOhqXR5E6urqmDt3Lm+88QbBwcFdfXe9zo3ueWIfRvYVXrqmHhgyoiKEEKKrdXkQefLJJ7n33nuZOnVqV99Vr/TcvaOYv35vqzBiXdXx3L2jrnobkUHe9PPXtdkDo63RjRsdURFCCCGupkuDyMaNG9m/fz9paWkdur7BYKCmpsbho69JTv+GZ97PtZ34dxwtZWXyGOav38umPUXMfO0rnnk/l/nr97IyeQw7jpZe9TaLq/Scq9QzNiqQFduOOYxwXG1040ZGVNyFTDEJIYT76rIgcvbsWX71q1+RkZGBl5dXh74nLS2NwMBA20d0dHRXHZ7bGhTqS9bhElIyLH0qahubWfbhYeZOjGbJ5kOcuVjP9iOWcPLG16cpqW5s90RqDRqrZ48jfd4EUBRSN1mCTkdHNyKDvEmdOszlXUWvN1DIFJMQ3UNCv7geXbbp3QcffMCsWbPQaDS2y0wmEyqVCrVajcFgcPgaWEZEDAaD7fOamhqio6P71KZ3xVV6nsjYz4myWoaE+/HizNEs/ucBTpTV46EGoxkWT49nZ8FFDM0mdBo1qx9KbHNH3pZBo7hKT+rGHAwmc7vf6+x2UqcOY/WO4y4bEWkrOHUkUMnGgEJ0vRv5GxW9i1tsevfDH/6QQ4cOkZuba/u4+eabmTt3Lrm5ua1CCIBOpyMgIMDho6+JDPLmL3PHMzTcn5NldSz+5wHOV1reYRjNMHdSDKu2H+dSvcEWJKat2smmPUWtbisju5Dbh4QybdVOh9tfcvcIDpytJirEp8MhZEXyGCbEhrh0M70bqVnpDVNMQrg7qSsT16PLgoi/vz8JCQkOH76+voSGhpKQkNBVd9srRAZ5MyYqkFA/HSfK6tEbzWjVKgYGe5Gxp4gwf09OltXz8zsGExnkzdM/GMqSzYdahZGYYB9e2XaMZfddKWhdtjWPFZ/lk5kymfJaQ6t6EfuhU3fc2fdGAoW7TDEJ4S66YipFQr+4VtJZ1U3dMTSMs5cabJ/3D9BRVW8EoLTawMIpcSzOPMj2wyV8fbKCpTOG86xdGNm0p4hnNx/ilQdvYvakGMDywpJ/oQZUKgYEOgYKZ/US7rqz7/UGiuIqPat3HCczZTKrdxyX9uyiz+uq+ikJ/eJadFmNSGe4ljmm3uSZ93P57HAJAE1GM2F+npTVGtBqoNkEfp5aUMHDSTG88fVp/r34+0QGedvCx8yxkXx0oLhVCLG+MwEc/p+6MQdUKlbPHtcjXjCup2ZFakSEcK4r/jbcpa5MuI5b1IiI65NTVMmHuefRG0xMig3m5R8lUF7XhEatoqkZtGrQaFSYFYXXd57iDw8k2P7AZ0+KYebYSD48UMzMsZG2EAKOoxv2Q6cXqvWgUjGiv3+PeKG4npoVd5xiEsJddPZUijvVlYmeQYKIm/nFO99hNEO4vyffFVXx4keH0WnUGE2KZdWMCSobjNQZTEQE6vivj4+wbGseYJmO+ehAMfdfHhGxrxlpuWeK/dDpkhnDWf5A99XtXO+89PUGCnedYhLCXXTWVEpxlZ7Za3exaFq807/RnKJKWcIrWpEg4kZWZuVzZ3w/gn08KKttQq2A3mimwWhCc3nprnUeLWXKYC7VNWFSzIBjTcif5yTyyoM3OdSMtOTKeonrnZe+3kAhG9cJ0b7Oej3IyC7kz3MSWbW9wOE2IoO8WTQtnqffy5G+PaIVqRFxI9YT8VfHL6JSQXs/GbUKVAqE+ut4/LZBpG07xtIZw1nwvaEUV+nJyC4kJtiHZzcf4tkZw6kxNNtOuO5QL+EOxyCE6NoaEfn77rukRqSHsr6zj/D3bDeEAJgV8PLUMCE2iBVZx3hyymC+PllBTlGlbVThjvh+JMUFsyLrmO1diLvUS8gSPyFcr6teD+TvW1wLCSJuJjLImy1P3o6XVtXu9TRqWD17HBX1Rm4aGEhtk4lF0+KZv34v4f46SmsaSd2YgxkVXy/5ge0FYM3Ok8SF+bpFvYQs8RPCtbqyfkr+vkVHSRBxU9Ehvm1+TQXEhfryxtenWXr3CHQeGg6crSLts3zWPTaRE+V1PPzmHgwms8OS3OIqPacv1rNwyhCnt9vd9RLS10MI1+rK+in5+xYdJUHEDf3h48OcKK9r8+tqFVyobqRWb2RLznmWzBjOgXPVGIwmzIoCCtQZmh2+x93maGWJnxC9l/x9i2shQcTN5BRVsuNoGYoCnq234wG4/DU1FfVNHDhbxYptx8hMmQwqSE7fBSrITJmMTqMmdVOu283RukudihCi88nft7hWEkTcSHGVnqffy8HHQ0OQtwdNpitfa1kxYjSbCfT2oKS6kRH9/RkQ6A0K+HlpQYEBgd6sfigRg9HkdnO00tdDiN5L/r7FtZIg4gasDb4ysgvZtGAyMSHeVOmNeGlVaFQQ5uvJ5iduJT7cUjfi4aGif4AXhRX13D4sjIVThpC6MQedh4Z3fzYJnYeG1I05lNY0ovPQMDYqkBXbjrnNOxHp6yFE7yV/3+JaSR8RN9ByKHPQs58wPMKPyoYmQMXahyeQGBMMwDMbc8jMLWZ4hB8JAwPx1Wk5fbGeuDBfFk4ZYluz/+rnx/giv4x1j00kIsCLNTtPcvpivdtMzwghhOi9pI9ID2Mdspy9dhc5RZWceeVekgaHMijMjw+evI3EmGBb6/NXH0rk22d/wNRREbz603H4e2lZkTyG5XZ7zgCU1RpY99hEdhwtJTLIm+UPJDjchzPXu+23EEIIcb1kRMSN5BRVMn/9XlY8OIYNdvOs17riZWVWPnOTYp1eN6eokqffy2HTgskOX3e3VTVCCCF6rms5f0sQcTNZeSUs2LCPtfMmMD2hf5cEBGm/LIQQoivJ1EwPYy1WLa7SsyG7kLXzJrBk80GyDl8gdVOu006oN8Kd2i8np3/T5lRRTlElyenfdPMRCSGE6E4SRNzA3KRYUjfmkLoplxXJYzh4voolM4az4N391OqNDp1QO6uOw13aLz937yjmr9/bKoxYp6meu3eUS45LCCFE95Ag4gbW7DzJmYoGDM2WxiFTR0aQ9lk+Q/v5craygZc+PmIbNbFuaHej3KX9cmJMMOsem+gQRqwhZN1jE22rhYQQQvROEkTcSFOzmZSMfSx4dx8DA73x9dIyINCb7UdLOFVex5LMgyyaFn/DDYHcrf2yfRjJyC6UECKEEH2IBBE34O+l5eUfJXD2kp6CklrKag2gAmOzwvlqPX46D745WcEjt8SyanvBDY2IuGv75cSYYBZPH87vt+SxePpwCSFCCNFHSBBxA3OTYnlndyE3xwbR1GwmNtSHoooGTlXUoZihssHI0hkjWP7xkRsuKnVl+2Xr9JIz2w+X8IePjvDyrARWZh1rs4BVCCFE7yJBxA1kZBfir9Oyv6iK39w1nMKKBuqbTBiazKhUkBQXzIsfHeHPcxJvuKi0u9sv24ePuUmxDqMu1sLb7YdLSMnYz2tzEpmbFNuqZkQIIUTvJUHEDZRUN/Ll8XIiArz4044CfC9vu+upVdFsMpF9upIX7x9F2qdHeeb93B7V/dQ+fNhPAeUUVbIk8yAxwT6kZOwnfe54po3uDzgvYHVH7Y3wSJdaIYToGAkibuDTQxdQASfK6zCaFOqbTHhqwNCs0GQCrQr+J6uAg+eq+fhAMVNHRvSYHhst608ig7xZNC2e+ev3Mi8plpc+PeoQQqysYeSlT4646MivruUIj1Vnrm4SoiUJwKK3kSDiBjzUKmoNJhQFFCw/FKPpytebFaiob6Kx2czyBxLYsLuwR/XYaNlAbdX2AlY8OIYFG/axfv7EViHEKjEmmMyU27r5aDvOWZGvdKkVXc0dArCEIdGZJIi42MqsfFA5XmbGEkhaWjpjOKu2F7D9SKlteWtPeUGwb6A275YYNmQXttvDpCc9LnfpUiv6BncIwO4QhkTvIUHExeYmxRIV7HvV66mAf+w/T31TM9NGRdiWt/aUFwRrA7W1D49nSeYhFk2Lb7eHSU95XOA+XWpF3+HqAOwOYUj0HhJEXCwyyJubBwUzLNyvzev4eGrw9FBzoqyOpfeM5NWfjnP4fnd/QbAez6Jp8WzYXcS6xyayantBqwJW+9DREx6Xlbt0qRV9i6sDsKvDkOg9JIi4gYVThqDVqFCrnH9d32TCeLm/yIrP8lutJHHnFwT78LDjaCkrkseQGBPcqoDVWQ8Td35cVu7WpVb0He4QgF0dhkTvIEHEDazZeZKS6kbMzgpDsNSL+Ou0BPt4EhXsw6Nv7XEaRtzxBcG+gZp9D5OW4aOtHibu+rjAfbvUit7PXQKwO4Qh0fNJEHEDF2sNVDYYHS7TtLiOwWTCYDQxcoA/AwK9WoURWw3GvAmkbsx1mxeEG22g5s4vdK7sUiv6LncJwO4ShkTPJ0HEDXx1vNz2f41axdIZw/Hx0jpcp9lk+cPvH+jFuvmTGNzPj1+88x3FVXrHGozsQv48J5HUjTks25rX3Q+lU7n7C113d6kVAtwjALtLGBK9gwQRFyuu0qM3mlEDwT4exIf78cWxct752STGRQeh01rCSaC3Fh+dlrlJsUQGefOXueOJC/PliYz9pG7MYdG0eEt/juQxRAR4gUpF/oWaHvuCIC90QjjnDgHYHcJQb9RT2hZ0NgkiLpaRXcjIAf7MGj+QT56+g4lxIaAoRAR48Ze54/nhiP54eai5d0wkmSm3OtRUrH4okdIaPXWGZtI+PcqK5DEALMk8yOrZ41j9UGKPfUGQFzoh3FdXhaG+eiK26kltCzqTSlGUNkokXa+mpobAwECqq6sJCAhw9eF0G+svXerUYazecbzdlSL7Ci+RnL6LsdGBLLtv1FWvL4QQ7qqtJfruunS/K7R8rD31sV/L+VtGRNxQR1eK2BdyouCWK0uEEKKjelL/oK7SE9oWdDYJIi62MiufZVvzHIbi7LuQPvX3/U6LTu3/OAcEeqPz0DA2KpAV245J/YQQosfqiyfilty5bUFXkCDiYnOTYsm/UEPqptxWK2De/Oo0EQHerYpO7UMIXKkJSZ83ARTFdltCCNET9bUTcUvu3LagK0iNiBsortKTujGHMxUNDAj04sX7R5P26VFQqVg9exwAqRtzGDEggOUPJLAyK99WtNTy3UJxlZ41O09y+mJ9n3sXIYToHa6lTq63kRoR0W3sq8OtK2AATpTV8dt/HuRspZ7Vs8dd+cVTXen/bq1Yd7ayJDLIm+UPJMjKEiFEj+Tu/YO6Ul9tWyAjIi5i/wuXkV3I3KRYLlTrSU7fBUDarATmJMWSU1TJ0+/lsGnB5B6VhoUQ4lr19VUz1tFuZ4+xuEpPRnZhj2mUeC3n7y4NIunp6aSnp3PmzBkARo8ezbJly7j77rs79P29OYgkp3/DwjuH8M7uQhZNiyft06PUGUwUXWogyMeDC1V6XpqVwIptx1j32EQSY4JdfchCCNGletOJuK9zmyDy0UcfodFoGDZsGIqi8Pbbb7Ny5UpycnIYPXr0Vb+/NweRnKJK5q/fy8rkMbzx1SkqG5o4dbGB394Vz+v/PkmAl5ZzVY2kPZjAnEm9s4mNEEKI3sltakRmzpzJPffcw7Bhw4iPj+fll1/Gz8+P3bt3d+Xdui37upDEmGDWPTaRZ/5xgLJaA8fL6hkbFcDKzwtImTKEmsZmhoX7kvZpfquddoUQrtfXu4AK0Vm6rVjVZDKxceNG6uvrmTx5stPrGAwGampqHD56k7lJscxeu8sWLBJjgrl9aBhnKhroH6gjt6iaMVGBrMg6xsofj+XtnyUxbVQET7+X02uLlIToqW60HbcEGSEsujyIHDp0CD8/P3Q6HQsXLmTLli2MGjXK6XXT0tIIDAy0fURHR3f14XWryCBv/jwnkfnr95JTVElxlZ6LdQYGhfpQUm0gIsCLA2erSJuVwIbLK15e/ek4Ni2Y3KsrpoXoiW60C2hf3VdEiJa6fNVMU1MTRUVFVFdX889//pO//e1v7Ny502kYMRgMGAwG2+c1NTVER0f3uhqRnKJKHl23h4GB3qhUKoou1RPmp6OwooFnZwzn65MVDrvpWteSS6GWEO7nRnpe9JaeEUK05DbFqs5MnTqVIUOGsHbt2qtetzcXqy7c8B3b8krx9lATF+aHn07Dz+8YzOLMg6xMHkPuuSrbOyZ5URLCvVk3n8xMmcyE2JBr+t6+3LxL9F5uU6zqjNlsdhj16IuKq/SUVBvw9lCjN5oJ9/dk9UOJTBvdn3WPTWRx5kGmjowAIC7MVxqTCeHGbrQdd19vZy6EtitvfOnSpdx9993ExMRQW1vL3//+d/7973+TlZXVlXfr1qzvfvRNzQzvH8Dsm6N4futhDp+vJjLI27aaZtnWPIJ8POXdkRBurOVUirVm5FqnZ+yDjPzNi76mS0dEysrKeOSRRxg+fDg//OEP2bt3L1lZWUybNq0r79Zt2W9oFx7gxV/mjmdOUizpc8eTkrGf7YdLAIgI8JIQIoSb64x23H25nbkQVtLivRutzMpn6sgIhyJUq+2HS/jlezn875xxbNhdJCFECDd3o11A+3o7c9G7uXWNSF+2ePoIdhwtJS7Mt9XXpo3uz//OGceCd/fTz18nL0BCuDnr5pPOWPePcsbaI8TZppVwZVRFasNEXyFBpButzMqntrGZWYkDWZJ5kGVb82xDsDlFlTz/wWHC/XWcq9STU1QpDY2E6KE60iOkvSATGeQtS/VFnyFBpBt9WVDO3jOXSPv0KIumxZN/oYbUTblsP1zCnL/t5mKdgbUPT2Dp3SOYv36vbeWMEKJnudFmZ0L0JRJEutHyBxI4X6Wn1tBM2qdHWXrPSGr1Rn7+7j4am8ykzbqJiAAvVm0vYN1jE1m1vUCK1oTooezDyL7CSxJChGiDBJFulBgTzNvzJ3GyvJ5LDU28+OFhzlU2ABAb6s3QCD9SN+YQF+ZLYkywzBML0cNJjxAhrk6CSDdLjAnm9TmJnL7YwIFz1dQaTMydFE2QtyfJ6bswmMwsnDKkU+5LNtUSouu193eWU1RJ6sbc6252JkRfIEHEBUYPDMRLa3nqQ3w82HrwAvVNzfjpLP3lSmsaWbY1r92NrzoSMmRTLSG6Xlt/ZzlFlcxfv5c/z0mUHiFCtEOCSDcrrtLz6Ft7qG8y0T9AR7XeSJOxmdMXG3j+3pGgwMNvZnPwXHW788kdCRlSMCdE13P2d2YNIesem0hiTHCb1xNCSBDpVsVVeh5/ey8nyutY8eBNpM+bgIdWTZMJ+gfo+MMnR2loMuHjqcVTo2r3tjoaMjq7YE6me4RoreXf2dPv5TiEkJbXk9ovIa6QINKNXv38GCfL6/nrvAnMnhTDlpzzjOgfwNB+vpRUN6JC4XhZHenzxrP6ocSrvlh1NGR0ZsGcTPcI4Zz939nqh8a1CiH215MeIUJcIUGkG52pqGfmmAGMHhhIcZWe0xfr+cvc8az48RhMCtQ3mZiREMHqHccBOvRi1ZGQcaO7g7a8P5nuEaK1zvw7E6IvkSDSjTJTbuOZu4aTujGH1E25rEgeA8CKz/IZNcAfH08NJdUGFk2Lv6ZNs9p78euKTbWkP4IQjmTzOiGunwSRbhYZ5M2IAQGgKJTWNJK6MYcmk4KfTsu7jyfhqVGxJed8h17Irvbi1xm7g7b3OKQ/ghBd+3cmRF8gQcQFlj+QwNJ7RjJ//V5C/XR4alSsfiiRxJhgVj+UyOmL9QDtFrV15MWvKzfVkmFoISxk8zohboxKURTF1QfRlmvZRrinWZmVz5iBQSzYsI/MlMlMiA2xfa0jW4jf6BbkN6JlCJIaESGEEPau5fwtQcRFrCfv1KnDWL3jeI85ibcVOiSMCCGEsLqW87dMzbhATy5sk2FoIYQQnUlGRLqZjCgIIYTo7WRExI3JiIIQQghxhYyICCGEEKJTyYiIEEIIIXoECSJCCCGEcBkJIkIIIYRwGQki3WBlVn6bS3OLq/SszMrv5iMSQggh3IMEkW5Q29hM6qZcpxvSpW7Kpbax2UVHJoQQQriWBJFusHDKEFAUhzBiDSEoiuXrQgghRB8kQaQbZGQXsvSekbYwsq/wki2ELL1npPQOEUII0WdJEOkGc5NiWbW9gKX3jMRgNJGcvguD0cTSe0ayansBc5NiXX2IQgghhEtIEOkG1q6paZ8exWgyA2A0mUn79Ki0dBdCCNGnSRDpRk0mhXOVel6elcC5Sj1NJrdtaiuEEEJ0Cwki3cBamOqpUbHyx2P5/ZY8Vv54LJ4aldPVNEIIIURfIUGkG6zZedJWmLohu5DMlMlssCtgXbPzpKsPUQghhHAJCSLdwN9LaytMXZE8hgmxIaxIHmMrYPX30rr6EIUQQgiXkCDSDayrZuwLU60FrLJqRgghRF8mQaQbZGQXOl0dYw0j0kdECCFEX6VSFMVtl27U1NQQGBhIdXU1AQEBrj4cIYQQQnTAtZy/ZURECCGEEC4jQUQIIYQQLiNBRAghhBAuI0FECCGEEC4jQUQIIYQQLtOlQSQtLY2JEyfi7+9PeHg4P/rRjzh27FhX3qUQQgghepAuDSI7d+7kySefZPfu3Wzfvh2j0chdd91FfX19V96tEEIIIXqIbu0jUl5eTnh4ODt37uTOO++86vWlj4gQQgjR81zL+btbNzmprq4GICQkxOnXDQYDBoPB9nlNTU23HJcQQgghXKPbilXNZjOpqancdtttJCQkOL1OWloagYGBto/o6OjuOjwhhBBCuEC3Tc2kpKTw2Wef8fXXXxMVFeX0Os5GRKKjo2VqRgghhOhB3G5q5qmnnuLjjz/myy+/bDOEAOh0OnQ6XXcckhBCCCHcQJdOzSiKwlNPPcWWLVv4v//7P+Li4rry7lxuZVY+xVV6p18rrtKzMiu/m49ICCGEcG9dGkSefPJJNmzYwN///nf8/f0pKSmhpKQEvd75ybqnm5sUy5LMg63CSHGVniWZB5mbFOuiIxNCCCHcU5fWiKhUKqeXr1u3jscee+yq398Tl+9aQ8eK5DFEBnm3+lwIIYTo7dymRqQbW5S4jcggb1Ykj2FJ5kFSpw5j9Y7jEkKEEEKINsheM10gMsib1KnDSE7fRerUYRJChBBCiDZIEOkCxVV6Vu84TmbKZFbvON5mAasQQgjR10kQ6WT2NSETYkNs0zQSRoQQQojWJIh0ImeFqfY1IxJGhBBCCEcSRDpRRnah08JUaxjJyC500ZEJIYQQ7qlbd9+9Vj1x+a4QQgjR113L+VtGRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMl0aRL788ktmzpxJZGQkKpWKDz74oCvvTgghhBA9TJcGkfr6esaOHcvrr7/elXcjhBBCiB5K25U3fvfdd3P33Xd35V0IIYQQogfr0iByrQwGAwaDwfZ5TU2NC49GCCGEEF3NrYpV09LSCAwMtH1ER0e7+pCEEEII0YXcKogsXbqU6upq28fZs2ddfUhCCCGE6EJuNTWj0+nQ6XSuPgwhhBBCdBO3GhERQgghRN/SpSMidXV1nDhxwvb56dOnyc3NJSQkhJiYmK68ayGEEEL0AF0aRL777ju+//3v2z5ftGgRAI8++ijr16/vyrsWQgghRA/QpUHke9/7HoqidOVdCCGEEKIHkxoRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEkTaszMqnuErv9GvLtuaxbGue068VV+lZmZXflYcmhBBC9BoSRNowNymWJZkHW4WR4io9+SW15F+ocfq1JZkHmZsU2+5ttxdyJMgIIYToS/psEGkZBuw/L67Sk5FdyIrkMSzJPEhOUaXt60syD7J69jhWP5ToEFSsX1uRPIbIIO9277u9kNORICOEEEL0Fn0yiKzMymfqyAiHMFDb2Ezqxhy2Hy5h9tpdzE2KJTLIm0XT4nn0rT2cKq93CBqRQd62oLKv8FKHQwjg8L1XCzIyeiKEEKI365NBZG5SLKu2F7BoWrwtDCycMoQ6g4lfvLuPp74/lMggb4qr9KR9epSoYB8+yyth3i0xZGQX2m4nMsib1KnDSE7fRerUYR0KIfbfax9kpq/+kkduiW11G3OTYnkiYz8zX/vK4XIZPRFCCNEb9MkgYg0B9mHk0LlqzlU24KlV89zWPN7bU0jqxhyaTAp+Og1rHx7PM+8f5LNDF2yjEMVVelbvOE5mymRW7zje5sgFOB/ZsA8yE2KDWXx5GsheaU0jp8rr0KjVDqMnbQUXgJyiSpLTv7nRp0kIIYTocn0yiKzMymfNzpMsmhbPqu0FzEjoz4IN+1j547Esv380RpPC0s15lNUa8NSoWHrPSNL/fQoFhbJaAyXVjeQUVbIk8yBxYb4MCPRuVU8CjlMn9nUhyenfkFNUSXGVnhWf5TM2OpAGQzMrk8fw8JvZTP/TTsASKOav38vbP5vEX+aOd5gGWvWTsU6Di/V7nrt3VPc+qd3M+hw6I0FMCCF6DpWiKIqrD6ItNTU1BAYGUl1dTUBAQKfdriUMfEt5rYE5E6N4N/ssL89KYGvOeeoMJk6U1dBkslx3yfThZB0p5URZLUP6+fHi/aP5/ZY8TpTVsmnBZCICvJi9dheT4kKYMbo/izMPsu6xiWzJOU/+hRqW3jOSLTnnOVxczXP3juLp93IYGOTF/qIqhvTzw9/bg6V3j2DZ1jzOVeqp0RtRq+DFBxJYmXWMdY9NJCLAizU7T1JnaGbz/vNkpkxmQmyILXSse2wiiTHBrT7vaVZm5dtqc1qyFhAvnj4CoM3H2tOfAyGE6A2u5fzdJ0dEMrILSYwOQq1SeDf7LAmR/ry/9yyVDUaOXKjBaLpy3RVZxzh0zhIa0udNICLAi4t1BqKCvVm1vYA1O0+y7L5RbMk5z9ObcliZPIZlW/M4cK6KOoOJBe/u47szlZwuryfts3zmJcWQfbqSZpNCfkktPxzejycy9uOpUVHZYEQB7rspkt9vyWNsVBBgefd/4FwV5yr1DtNAiTHBrHtsIvPX7yUju9AtT8DXUmxb29hM6qZcp6uJUjflUtvYbLvM/rFbR0YkhAghRM/TLSMir7/+OitXrqSkpISxY8fy2muvMWnSpKt+X1eMiIxeto2xUYHsK6rE0Nz2Qw/29qBSb7R97uWhwl/nQVw/P1AUVj+UCFhCQrCPjoamZs5UNDAo1Ad9kwl/Ly3nq/QYTQoqFG4f1o+zlQ2cKKsnUKel2nDlpKpRgUkBbw8VjUYFBdBpVRiaFbw81Ph6aokO8WFMVCDLH0hotcImI7uQ32/J4+VZCddUvHotIxDXq63VQM4uL67Sk7oxB1QqVs8eZysYTt2Ua3vOWx6rNXwsnj7cNoIkIUQIIVzLrUZENm3axKJFi3jhhRfYv38/Y8eOZfr06ZSVlXX1XTu17L5RfHvqki2EqFTOr2cfQgAajQp6o4mzlxpYes9IfvneftbsPImHWs2RCzV4aNT4eKo5U9GA0WTmZHk9eqOZZrNCoI8nRy5Uc6KsHq0Kqg3NBHprbbdtUkAF6C+HEABDs4IKaDSaCfDScuUrFnFhvizJPMj2wyWszDrGlPgwVmzLd1o30dYy3+vtZ9LRUQ7rv86WKqduzCEuzNchWEQGeVsCnqKQuimXfYWX2g0hYBkZWTx9OL/fksfi6cMlhAghRA/T5UFk1apV/PznP2f+/PmMGjWKNWvW4OPjw1tvvdXVd+3U7EkxaOzCx7WMB9UZTDw4LpInMvZR19jMJwcvUFprQAUcL6tD32QG4NLlKRYATw3U6I2U1jQB1pEPNdX6ZofbdnYYCjA8wo/CSw00Gc0snDLEFhJmJQ7ET6fhl+/lsO6xiaQ9OIa4UF8eXbfHIYws25pH6qZc5ibFOg0Q1kDTsmnbiuQxZGQXOg0cc5NiSd2U26rNfcsAYw06gMNSZeuox8IpQ1rdtjWMGIwmktN3YTCa2gwhwOXjPsbLsyw1NW0VsAohhHBPXRpEmpqa2LdvH1OnTr1yh2o1U6dOZdeuXa2ubzAYqKmpcfjobMnp35AUF3Ld3//6zlNcqDZQpTdSVd+Eodls+5qzMNFkgiaT4nAdvdHs5JqOVCroH6DjWGkdAwJ0aC+npyWZB1k0LZ4XPzzMNycreO8Xt5AYE0xkkDfp8yY4hBFrO3pr2rKvwbCGhoVThliatq3bw8nyOofpkrZGTCwPxFLjsmxrnsPttZyCsQYdwLZU2WAy26Ze2n4CWvzrhH1NyNyk2FY1I0IIIdxflwaRixcvYjKZiIiIcLg8IiKCkpKSVtdPS0sjMDDQ9hEdHd3px/TcvZapmRtVWmPAWtPaFUU2nhoVJTUGdFo1JTUG/mNSrC2ErNpegFlReHv+JHYcLbUFBfsw8mzmQVI35jCiv7+tHf2sxIGgKDyRsZ/UjTmsSB4DQNpn+QwM9GZbXqlDY7b2OsCufiiR1bPHkV9SS8qGfaRuym1V72ENOiuSx5C6KZflHx1hbHQgOk3bv3bWmhCdRk1mymR0GrXTAlZnhanOCliFEEK4N7daNbN06VKqq6ttH2fPnu30+0j+y7edfpudTaMCjdryo4kJ8SE6xIelWw5xy+AQVm0vYEXyGD765R0kxgS3GrWIDPJm2cxRHCutI6+4hjuGhjk0cHv89sGcKq+jssHI/HV7SN2Ui8Fowt/bw2ljtvZa2UcGebNkxnAOnKvGYLfUyNnoiMFo4sC5apbdN6rVPj3232dfEzIhNsShZsT++i99csRpYao1jLz0yZFO/ZkIIYToGl0aRMLCwtBoNJSWljpcXlpaSv/+/VtdX6fTERAQ4PDR2RJjAjv9NjubSYFmk5lh4b6cr2qgvNbAuOhAVmYVMO+WmFYFnvajFvZN0mJDfXjmHwfIKaq0dXFdsGEfKVMGc6KsjmOlddTqjeg8NKyePY4JsSGW0YuNOQ71H221srfvLKvz0JC6MadVWLEWpuo8NLagA60LWAHW7DzZqjDVvoB1wbvf2a6fmXKbQwixL5JNjAkmM+W2LvrpCCGE6ExdGkQ8PT2ZMGECX3zxhe0ys9nMF198weTJk7vyrtvk0c60gKupuLKKp8mkEOanQ6NW46FRc+piPWkPJrAk81CraQdrGEndlEvKhn2gUpE+dwJvPjoRLw8ND7+5h+2HS1i94zhT4sP4n88LGBjkBUB+SS0/vz3OFhrW7DwJKhX5F2ocClhbtrIvrtIze+0uFk2Lt4xczB6HwWS2hRW40sHWuhzXGnTsC1jt9+7x99I6LUy1hpHEmGDZtVgIIXqZLj8rL1q0iDfeeIO3336bo0ePkpKSQn19PfPnz+/qu3Zq92n3rB3QqMBPp0GlgK+nBn+dhu8KK1n1k7F4e2qIC/Xl6IVa1j02kYf+upvth1vX2FinP6KCvW1TJ3cMC8NkNrNwwz6MJhPfnanEU6umpKYRFaBWwa/fz2XTniKS078hv6SWpXePwFenZfbaXbZdilckj3EYMUndlMuf5ySyanuBLRjoNGrGRgey/KMjpG7KZerICE5frHcoTLUfwQFYPH2ELewsnj6iVQixjnREBnmz/IGEDu9aLIQQomfo8iAye/Zs/ud//odly5Yxbtw4cnNz2bZtW6sC1u7yxsMTXHK/bVEB/joNarUKo1lhxIAA7owPw8tDw/1jI3lndyHTRkXw4v2j8ffSkhgTzP+bk8gv38uxrYyxjjyoVCrC/XWcLK+znajn3RKLRq3Cx1PD7lOVNBpNDAzyto0MadRgNJlZsvkQfjoPooO9+e0/D7LzWDm3DQ2z1aRYR0wysgsZMSAAFIWIAC9WJI8hOf1bUjbsY/VDibw4czRltQZQFLbknGfRtHiHUQ+4Ekasl19LPxP7ILPo/Vxb0a01hFhDjbPeKS17nHS046sQQoiu0y3zFE899RSFhYUYDAays7NJSkrqjrt1Ksxf57L7bkudwYQKBR9PDecqG/Dx1LL3uWn46rTEhfmycMoQVm0vsPUCGT0wkPd+cQtPZOwjOf1bxkUFkV9Si4LCHcPCKKsx8KPXv2H74RLm/S2bR2+JpaHJhEYFZgVOX6zHx1NLZKCOJhO25m7ltQb2FVZyvKyOmFAfvDzUVDU02ULIksyDTB0ZYZtCsQaJcH8dpyvqOXS+ilXbC8hMuZXVDyWSf6GGtM/ynU6ZRAZ527q2trc6x9lIh7VmZfP+8xhMjkuhrT1OUjfmONxvWz1OZJpHCCFcy30LJrqIu62mUKwfCmjVaqaNiqB/oNeVHiBcOVHPXruLcVFBtmkNUOHrqeHX/zjA8dJaCkprmXdLLH+ZO57yOgM/f3cfPp4aXt95Cl9PDVqNCgVLMWxtYxMXqg2249CqodZgaVMfFeTFucoGtuWVEh3sY2mKtjGHRdPiefq9HE6X1zN/3R5qG5tZs/MkY6ODWPnjMSx4dz9qlaXo1FobMqK/f4emTNpbndOSQ5Gss+W9iuLQMrdlqGm34+um3FYdX4UQQnSdPhdEbhl0/c3MupLRDKF+njxz13CmjowgOf1bWyOy4io9kUHeTIoL4am/7+fWwSHMX7+X4RF+nCyvp66xmaoGIzqNmicy9nGx1mDrGFteZ+noWmMwOdxfkwmsYwmeGhXNZjCZFTQqUFBoNlsC0oGzlew9c4kzFQ2kfXqUZfeN4svj5RRd0vPJwQvkX6hhVuJA3vzqNEPDffn6eAXfnrxIfkktq2ePY/kDCR1+DlquznHW2dUaKhZNi+f/8statYS373HSVqhps+Pr5aXDzjq+CiGE6Bp9Loj8Zecpl9xvOw1CAUvbd0+tmkPnq3h03R5MZoWl94xk4ZQhLMk8yMzXvmLG6P7EhPqyIquAwWE+fFdYaWumpgB1jc1oVCoWbthHkN1eNlaGZgW1kwOx7/xqUuB8lWWkJC7Uh/PVBk6V1RHq50mTSeH//fsE0cE+qFQKjUaTbYfhWkMzfjoP4vr5cKKsnsdvu7ISp6N1Gcu25jmszrEWyracrrE2dbNu2GffEt66vLitJcfQumDW1vH1Ku3khRBCdL4+F0Rcpb3uq94eauLC/KhvbGbBu/sZGOjN2ocnsGp7AWB5165Rq3kyYx9GkxkvDzX7i6rx0zmGDZMC56oa0aihqsVeNlbmDraB9VBBeZ0Bb081BpPC0Qu11BmMHDpXTUlNI4NC/fDQqCkoraXW0ExRRQP1hmZq9M3MSIjglxtzeOytbKd1GTlFlSSnf9NqSiT/Qg0rkscwINCbuDBfVm0vYNG0eFsYycgutIWQltM2Og8ND44f6LC8uOWSY3u2Jc8bc1j+8RHGRgWi89B07MkRQgjRafpUEFmZld8l7dhvlKdWTT9/T05XNFg+91DbVqRY37UP6eeLWmPZ3TfM1xONCkpqLBvueVz+KVofW5PdLIzH1YZi2mBUoMFgotFuX5wTZfWE+npS2WCkWt9Eo9FEs1lB32SivsnEyfI6Ary0+HpqiQnxZufxi9w+JJRfvrcfsJz8F02LZ/76vSy8c4gtlFinREYMCKC0ptGhNbw1jKRuzOFCdWOrEGJrOT97HKt+Os5hebH9kuO29swxmMwcOFvNspmjbNM5bY3aCCGE6Hx9Koi460qIan0zOwsuEhfqw9joQFAgdVMupTWNxIX5kroxh6MXalFfLsA8V9Voq8VUsNSXOMsbKiyB4nqZab07cXldEzNGRXC+qhGT2Uyg3RSQVgXFNY0cuVBDiK+OZ6cP55Vtx/jpBMueQcVVelZtL2DdYxN5Z3ch85Jimb9+L9HBlumVWYkDmb9+L4umxdumV1Ykj2FLjmV1zOb951t1dnVa1KpSORx4W6ty7Pe0aa/jqxBCiK7Tp4JIy34Wruavc5wKqNIbeXHmaMZGB2Ewmnj4zWzqDc1Ehfhw5EINsaE+DA33BaDZDAF23++sYWxnjv5o7ZLOtiOl9PPzxGgGfZPJFkaazGAymTl6oZbvx/fjTzuO88qDN7H6iwIWvZ9rCw2JMcG2dvMrHhxDWa2BC9VXQop9kzSwdH+1DwzWr2VkF7YKIRnZhayePc5heTG07l1i307+ah1fhRBCdB2VorR8z+s+ampqCAwMpLq6ulP2nZn40nYq6powX/2qXc5TDRqNhkajyRYYgr09GNbfn5/fHsfvP8ijocnEHcNCqag3WjaVU8GZ8nqqG5vRqi1hxMtD7TB9otNAiwUyHabTqDCY2v51uHNoKF+eqAAsHVl1WjX6y/c9MEhnK3IN99NRVmfglrhg4vsHcOBcFQfOVpOZMpkJsSG2kYzUqcNYveM4826JYcG7+51+fcW2Yw77z3RGJ9W2bkO6tAohROe4lvN3nxoR+eOsm1CprrNoopNpNGoMzSY8NCq8PNTotCo8tGq8PdQszNhPqJ8nf/rpWHLPVoOikD5vAiHenlQ3NtM/QEezGXx1GhqNZjT2D0mlQqe9vh9reyFEBXx5ooJBYT6osMx8mOwqX8trDQwK9UGjgrI6A1o1HCqu4bszlQ6jGdaakLgwXybEhlwuRj3E2ofH20Y77Fe8WFeyWJfytjXNci2dUJ2NpEDrURMh+gLpMixcrU8FkWmj+7Nm3niX3b+KK0+43mjGU6OmyaRwS1wI//rN9wn187TUioT5MGpAAG98fZowP09WP5RIaU0j356qYHxMIOV1TUyJD8PXU4uXVo1JgaRBlp1oDc0Khma7ERK7OZXr3e/P11ONgmU/nDMXGwj19UDBsuw3NtQbT42lQLau0WjpJQaYzWAyKRRdqmfpPSP5v/wy/HVaHn4zG4PRxMIpQ2w1IyuTx/Dih4cJ99fZClhX7zjOg+MH2lay2HdCtQ8M19MJ1dmeNlb2HV/dmZw8er/k9G9abXBpZV151hmky7BwtT4VRABy2/jD7g4KlgJQ6whGY7MZL62K3HPV7DxWxtlLlheC8loDnxy6AIrC2odvJjLIm5c+OcLy+0dz6mIDa+aOJzbUlwGBXhhNZpLHRbK3sJJAby1qwNtDw9jowFb76piuc06qvslMiI8H1gGTi/VGwNKNVd9kZubYgYT5eXCx3ogKy3JkM9BkMtM/wIstOef5sqCcXScrLDWkKmyrYxZNi+dPO45TrW/GV6e1rapZNC2eVT8dZ1vJsmbnSYelvJFB3rYXUGf72fR2cvK4oreGsufuHcX89XtbhZGcokrmr9/Lc/eO6pT7udYtFoTobH0qiBRX6XndRQ3N7NnPgBiaFWbeNIClW/II8/MkOtiban0zKhUsvWek7UXgtTnj+X//OsG6xyYybXR/Fk4Zgs5DQ3yEP1lHSxnSzw+dVsPw/v6WjeyaFVbtKMCsWIJBVPCVFxONyvLhcZWfvv2Mz6UGI74tvqHZDIPDfJgxuj+VDZa+JSagwWgmMTqQEF9PSmoaOXC2ihBfTyr1Rvr561ChYtnWPBZNi+f3W/I4VlLDpMsjOtaC1R1HS4ErL5LWfWusYcTaMdW+uVlf4o4nj44Egq4IDb01lCXGBLPusYkOYcQaQtY9NpHEmOBOu69r2WJBiM7Wp4KIO75rVoB3s4voH6ij6FIDZyv1+Os0eGrU/HTtLocVIpsWTLa9+GRkF7L07hF4eqhRFLhU38SgUB/efGwi7z6exMU6A8dL62g2KQwK8yXYx4OYEG9L3xGtirHRgVhrXAO9tPh4XtmNF0CrVtmmY6zqL3+Dl4ca1eWv7T1TyVPv7ScqyAvPy4t4NMCymaP58KnbGRruT22jkZ0FF4kN9aawooGqhibUahW//edBjpXUEBfmS4PRTL2hmbgwXxJjgh2mR6zdU0f092fV9gLm3RJDcvou7kkY4LS5WU9+F3wt3O3k0ZFA0BmhoWWYsX8ecooqbV939fPRGezDSEZ2Ybsh5EZDXnvdiIXoSn0qiLjzO6PSaoOt66mfzoPaxmYGBnmz4N3vgNZ1DYunj2DH0VLGRgVxR3wYFfVNLLnbcp3EmGAmxAahUsHgfr5MHBTCR7+8gy9/+wM2P3ErGpWlM2uwjweDw3yIDPJGq1HzvfgwfHVaHkmKodmsMDzCD5PSukdJXJgvf314AjpPDYpiGdUprtLTZIKlM4aTEBXIo+v2UFrTyAszR3Gpwcjah8cT4e9FTKg3ZyoaOHC2muNldUQF+xDip2Pp3SMoqzW0uc9LZJA3yx9IsBW3+uk0/P4Dy7/2iqv0PL5+L+/vPdsnAokrTh5tnfCsDetm2wXoloGgM0ZynIUZ+2Z5Y6ICe0QI6WhwsATz4fx+Sx6Lpw9vcyTkRkPe1boRC9FV+lQQ+cPHh23/9/JQEebn6cKjcWS/hPdCTSPzJsVgaDa1O/y6ePoIFk4ZQr3B1OrFo6zWwPsLJvPFM99z2HguIsCLmFAffDw1DIvwZ+ndIzlfrSfcT8f+s1W8PX8Sh0tqWDpjOCfK64nw97Qdm/byRjVnLjYwemAgU4aF2W7XaIb/mBTDgu8NJX3eBOJCfVnw7j7SPj3KuscmcvBcNasfSiTI29M2JaRRw6WGJn5+e5zTkY2WrMWtKx4cQ93lNcr/OlbGExn7bW3dH1+/l4KyOn5z1/AePSzfUV198nBWMGk94W0/XOJQMGn9+UyKCyF1Y47TURprq/4bGclpK8xYfoduYsG7+3vEO/qOBgfLKM8xXp6VwMqsY20WsN5IyLO/3tW6EQvR2fpUEMkpquLuhAhuGhjAez+fTHWDsdNue1xU4DV/zy2DWoeMSr2R5HGRvJtdxPL7E1j+QEKb7+zbe/HITLmtVYixdhP112nZsWgKq2ePY9E/DvDqj8fi7akhLtSXiAAvXpszni+OlRMV5EVZbRMqFYT5eqIoimVaRqXwRMZ+fDy1+F4ekRgY7MWm786y/XAJkUHevHj/aDRqGDEgwGGqpaHJZJsSMl1egvzkezk8ckus0xdK62O33/BuQ3YhmSmTGdE/gKZmhfySav7z7b088mY2BWV1/PFHCXx86ILbvyO+Ud1x8nBWMBkZ5M0jt8SSkrGfhXcOITn9G7YfLrEdi69OS62h2TJ9dlN/W4v/4io9qRtzqG1svq6RHPsRhJbTUqkbcwj317FhdxGZKZNJ3Zjb5gnbXUbKOhIc7GtC5ibFtqoZae82OxrynIUVZ8cmRFfpUw3NnFmZlc/7e89ysa6JJ6YMvq5i1hmjIth2pLTTjsnHU83/zk7knd1tb/J2rU25rJfHhfmycMoQp23S4UodTUl1IzuPldPYbOZPPx3LO7sLeeSWWH79/gE8NCrUKssGejWNzayZO57RAwN5/O29FJTWsWbueHLPVdl2x7Xez3++/R3HSmqICvFG32TGQ6PifFUj/QN0NBhNvD1/kkN4sj+2jOxCpo6McHgurLeZf6HG1qRuyfR4vj11qU+FkPZ+/iuz8h1+Di1vIyO78KrLlVsWSFo/X5k8hnd2F3LPTf15/oPDpM8dz7TR/ckpquTRt/YQ7OvBucpG1swdz1cnLnLgbBU6Dw2rZ48DuNK07rN8RgwIcBi56+hj3ld4ieT0XYwaEICfTmNrfNdWUeeyrXnkl9Syeva4Vs9JW89HZzyHHXlc1gZ/zkKI/WPoSMGq9XmxNglsT1c/PtE3SUOza7B4+gj2PjeN06/cy9zJg1CrIDHacXRjQkwQzvZl9ffS8r34MLLsQsjVVqI4o2rxb0OT5bS6aFo8T7+X4/Skeq1NuazXX/5AgsP32F/f2kPD+qLTZLoSQlYkj2Ha6P68+/gkjCaFpmYzlQ1G/vijBKaN7k9kkDdvPjqR+Ag/FmbsZ1xUkMP9vPr5MY6X1RIX5ktEgGV34YHBPsSGeFNSY2BImC9Pv5fT5jvDuUmxrQJZZJA3L89KwONyA7cwP09WZBX0iGH5Gy0s7OjP/1rrBpwdl7Vg8tF1e3j0rWzbSXDa6P6kTh3G0s15/OFHo1l8ebpm1fYCfn/PSM5VNhIX5sP/+9cJ9p65xOmKepbebfndsv5sBwR6g0pF/oWaq77zbvkuvbhKz4rP8hk1IIBzlQ0Oq8ycrTgprtKTX1LrsA+R9fE6ez6sP4euXpXT1ujQS58ccRo2rI/tpU+OOL29a52u6w19dUTP1udHROxZX3QeeyubgrJ6NMAfH7yJjw9dsBRJ/vMABWX1hPt5UNHQjEpRCPHTUdtoxGgyEx3szekKyx99i33XrmrswAAOnK9Bq1bRbFbw8VAzYVCIy97ZJ6d/w3P3jmLH0dJW75aeeT+XrbnneflHN1FU2eDwQlVcpecPHx8mp6iK3b+bart85mtf0WxS8Pf2YOndI2yhAuCRN7M5fbGeNfMm8M7uwlbvDMH5uzZrTcix0lpG9vfn8IVaYkO8iQj0dvqO1510Zpv5q72jXbPzJKcv1juMJLV1H+0d1yNvZnOivJ6XZyUwNym21Tv5+24awO8/sCzL/utXp1iZPIbff5BHWa0BP52WP/10LG98fZqzl+r5y1xLj5un38th04LJALaptx1HS9s9+VmnGA1GEzoPDSP6+zMrcaDTkcOcokqefi+H1Q+Ns/1OWe/L+v/UjTmgUjGiv79ttNDZFIn1WDtrq4GWz7mz3/vrva2O/KyF6ErXcv6WIOLEoGc/IXlcJK8+lNjqRf6ZjTlk5hbz7bM/YM3Ok+QUVZIYE8zCKUOYtmonKhTUajUoUGNott2mp0ZFUxst1AeFeHPmkp6HJ8Ww5cB5zIqlliLlzsEsuWdktzzmrrYyK5/axmanJ4ziKj2vfn6MPacv8dy9o1iwYd9Vh5TtC1OtNSH33TSA332QR1yoZSWOu4eRltMEzqbJOvJutCOhBujwyc7ZySxlwz5OV9SzZMYIVmYds03L2F9n9tpd3HvTANZ8eYqXZyXw/eHhPL5+L0dLahka7sutQyzFzdvySgjw0lJWZ3CYjsspquSJjP1MGxVhm6ax/t4ALJwyxDZF9+KHhzlwrpoZCRGE+3vh76Vl6sgIh7Bg5Wyawv7kb93PaOk9I1m1vcA2HWqdxrQ+d9bLr2U66Vp/dm39LDsyfWIduZE9lIQ7kCDiYs+8n8uHuecxmkGnhZEDgpgzMZrffZDnsD8LWHpumIDvDQvj65MVLLgjjje+Pk18hD+Hi2vY/MStndq4yNXae0F19u61rRfOlA3f8fmRslaFqdsPl7AwYz93DA3FpODWL77W4k1UKpbMGO70Hbv1BHO1+fuWNUBrdp5sVQuRlVfCgg37WDtvAgfPV7UZclZm5dvqcVKnDmP5R0c4XVHP2/MnERHgxR8+PkzW4VLSZt3E7Ekxtu/bfriElIz9/Oftcbz17RmGhPni7+3B47cPYvE/DxIX6kv6vAnsPFbG0i15DO3nyzuPJ9lOwNbnwv6YrZc3mRQ8NSqW3jOSJzL2EeyjQ1EUzlXpGRzmy4v3j7aFCPsRlfZGG+wDyoBAb5ZkHmReUixLNh9k3WMTiQjwInVTrsOmi9bvCffXsfbhCU7/NjtaV3Eto2IduW5Hf1c6SmpHxI2QIOJiM1/7ikPna7gp0p/E2BBmJQ7k6fdyuH/MAIdi2MlxIWSfuUSoryfldU0kj4vkgwPFtimKR26JZfnHR1q9w+uNrnVI+ZY/7uAPDyS0KooFywnx+a15bH7iNrd/sbSONhw4V+1wQrza89DWyco6ZVFS3Uj/QC/S501wKHxcMn04aZ/lM21UBK/+dFybU15LMg9yT8IAlm45hIdGxV/+w1KQbA07owcE8PsP8nj5RwkUVTYwdWQE89fvZemMEaz+ogBfnZbTl7cjsBWwrttj25l56YwRpG3LJy7Ul2UzR7His/xWIcT+eKxhRFEUjJf3MIoJ8cVDo0KlUuGpUdnCQlvPkX1Ys47M2QeU//n8GJv3n2dwmA9h/l5EBXtzsqwOnYeG6Mudib86fhGTohDur6OqwUhmyq0AthDQchqsPdd6ou/uaZfOnD4UfY8EERdr+QJj/fyB//c1Pp4aLtY3seHxJHYcLWXqyAi25Jxn57EyCi/peePhCbaTa3vDrb1JX37Bsw8PqECnUXf4hOrsebG+Yx/Z3x+dhwZPjYqf3zGYxZkHWZk8hje+OoXBZEandVy9Yh1Jycgu5P29Z/nphCjWfnUaH08Nwb4e6JvMDAj0IsTHk6MlNez+3VTW/vsEr2w7xisP3kTatnzbdE24v46zlXp+fnscizMP2gou38suYumWQ0yJD+PtnyWRU1TJw2/toa6xmbHRgaTPndDudFHKhn2cKKtDQWFgkA+lNY2s/PFY3vz6lNORFGfP0bKteXx3ppJzlQ0MCPQiYaClMP1EeR1FFQ3UNRoZFhFAvcFI4SU9ob6eJMWFsLOgnMZmE15aDbGhlgCEClSobCM1aZ8ebTNMXau2Qoo1lEWF+FBea+jyv42Wz+P1rDoSfZOsmnExZ11QI4O8+enEaKYMD2f7r6fYemskxgSz/IEEpgwPJ3n8QHLPVdmub/23t29Nf60rgHoL64v86tnjWDZzFAfOVmNwsjNhR3tD2K+W8Pf2QFEUKhuaWJixnwAvLat2FIBKRfrcCbbNBAEeuSWWzP3nSE7/hnFRQfjqtLy+8xQL7ogjrp8vF2ubqNYbyDtfzZfHy0mMCWL74RLSvzzFXx+ewEufHmXJjOG25eZnKupZPXsc00b354cjwlm2NY+cokpWZOWTNusmTpXXs2xrHhdrDegu7ynQZDQ7PI6Wq4Yig7wZEu5HfZOJhiYzx8vqWHL3CBZs2MeSu0ewevY4MrILbatgnP1OFVfpqTc0c66ygduHhuHloSHrSCkHz1VzvLSOqgYjg8L8aGhqpvCSHk+Nimq9kS/ySzGZFUxmGBDkhYdWhc5Dw39MjOFocTV1BhMvfnjYFkKsu0Lbs1+d0/KxObustrGZ1I05Dn10bF8zNLN5/3n6+essYczue6/1fq6m5e9ey1VH9rfdFQ0EV2bls2xrntOVP8VVepZtzXOLnjDixsiIiBAu0FYhaXtTFO31hnBaYJqxjwNnq4kK8uJcVSMalYo18yxTJWCpyXk28yCltQZWJo/h9X+fpKC0FhXQz1/H2UsNRAV7U3hJjwpL999BoT4suHMwz2+90jfEelwzEiI4fL6GP89JtK22WrPzJNvySlABL/3IMpU2dWQEP3/nOy7VNeGhVaPTqPDUahjcz9dWMGp9Xl79/BhnKup5bc54UjfmUGdopuhSAwMCvSmtaeTWoaFcqjfaVrxYn0v7gl/7KRP7aZlF0+L53eZDHCutRVFgaLgfob4e7D5diacGNBoNZrMZQ7OCWgW/uWs4a748SVyoLw9NjOH5rXk8M20Y6749Q1ltEw+OH8hv7hpuOwb7nj2WQtx9RIf42kairlZgmropl7OXGlh+/2jbSNPRC7Wcq2xg5U/G8LvNh6htbGbTgslEBHjZbs86jTUmKtBWTHujo4vO6mm6Y4roarVDnTUCJTqfjIgI4eas79jhyolzQmwIqx9KBEVhzc6TDtdvrzdEWycCnUbNsHA/iqsaAYgK8eKZfxwgp6iS4io9L354mAs1jaxMHkPuuSqev28kDU0mFGDBlMF4atUUXtLjobFsgBgV5EW13sjSLXn84QHLyTGnqJIV244xcoA/356oYNl9o1i1vYCpIyNYknmQO4aG0WA0caneYLv8xQ8PU29otkxtqBSC/TypaTRS2WDk0bf2sGhaPAApGfvYfqSUhXcOsZ1cLcuAx1Fa00hUsA8XqhsxGE0cPFdN6sYc1uw8yaJp8aRuyiV1Yw5TR0aQuimX/As1LJoWj7+XlsSYYFYkj+Hp93II9PHArICXpwa90cTu05VEBXmh1WjQN1mmYgA8tGr+O+sYK388BqNJYemWQ/zn7XGsyCpgQJA3mSmTOVdpOTmW1jQSF+ZLfkktqRtzyCmqJO3To0QEeIOiUFrTyJLMg7bnqK3RwNWzxxEeoOOZfxxArYJPD13gWEkNr/5kLOH+XjSZFLQay+aRyenf2EYjmkwKp8rrmJU4sM3fj2vpY9Pydw+4oRb91yIyyNv2N5G6Kdc22tPTQ8joZdvYtKfI6dc27Sli9LJt3XxEriUjIkK4SEdrY672ect6AuvXF02L58UPD3O+Sk9DkwkVlqmFkhoD/f0tRaOv/nisQwff1KnDeP6DPI5cqGVgsBfnKy0h5j8mRZNTVGVbihviqyM5cSDPb81jSLgf/jptq+Wv1oZ8T31/KP/18REMRhNxYb5U6Y3ojSZign2IC/Pl65MXCfbx4NwlPYP7+TEmKpAT5XWcLq/n1Z+M5Y2vTjmsmnn6vRxuGhjI+So9KpWK4io91Y1NxAT7cqm+CZNZwVOrZkCQF03NZspqDIT5eRLo7eGw5Pa97EKWbsnjt9Pj2bz/PCfK620/A28PNXqjGY0KfjM9ntf/dQqdVk1EgGWX7DA/HUUVDQwJ9yPYx8NysgSeyNjPqfI63v6ZZYXRExn7OVFWy9Bwf/4ydzylNY3MX7+XFQ+OYUMbU5Itf0cef3svRy9YRqqGhvuhVas4X63n7fmTKCipZcnmQ8SF+pAYG8yp8nrb87RqewHzkmJ56ZMjTIoLwVendTpCYg291u0kOto/5UK13jZK8n/5ZW0W3i7bmme7fWePr6MrjFI35limLhVsHXp7YggBS9h4dvMhXnnQceVZW5f3RFKsKkQP0BW9IexDiH3xZGlNIw+/mY1ZgWaTmSaTwm+nx7Pr1CWHbQQAUjbs41hpLY1GMyosowHNJjM6rRofTy3hATqMJjOnLzYQFexFYYXeNmRvnQKx3qZ1OeySGcNZ/tFR9EYTgV5aBtktuZ13SwwL3t3P0HBfLlQ14qlVYzIrvP2zSew4WurQR+SX71n2t7GGkyH9fAHYWVDOxbomtCpoViA21JsafTOVDUa8PdREBftQZ2gmM+VWMrILiQn24YUPD/PCzFG8/OlRzAoYjCbMimUKysdTg7eHBrOiUGdo5jfT4ln9xXH0RjPDwn3RatSUVDdS32hk6qj+XKwzEB3iw9lLDbbQtOTuESz/6AgHzlUzNiqQZTNHsXrHceYlxXaoV47152ldVeXtqSHcX0dhRQMj+/uTOm0YSzIPkXLnYF7dXnB5xEjDu48nkRgTbFuunTYrgS0550GlQt/UzPIHEkiMCb5SKN1sQqdRkzw+ipc/Pcq21DsdVlq17O5q/T4UxVKfs+N4q94rzqZQTpTVkhQXQvq8m1v9Dh8+X83zW/McmiC2ZJ0eAjr03Lm7lqGjN4UQkKkZIXqEjrTWvt5W/tYTj/VdY2JMMK/+dCwNTSaaTAqjBvizarulG6p9CFmSeZDIIC+s7090Hmp+ftsgzAo0Gs384o44KuqaKLxYj8msUFFnJO3BBJ76+35SN+WycMoQViSPYcfRUlKnDmPBhn0snj6cpZvzGHh5CWx1YzN3jY6wjZpYN6rz1WmpbzJR2WBkyd0jbAXdyx9IsG1N8Nqc8byzu5DoEB8UReFspZ5n7hpO2qybAEsIuXNYKIUVeqoajKgAs6JwvlrPrUNCiQzyJibYh2c3H+K/7h/NiAEBKAromyxdWq3vyhqbTJgVhTA/HVq1ihVZxzAYzei0Kk6U1XO6vJ5Go4nYMD++PF5OZYOR7UdKWXrPSP4ydzwhfp4kp+/iQnUja+dNABUkp+9i3i0xtk0bV3yWbxstsLKfMrGexHUeGjJTJhNxOYTEhvpgNCsseHc/K5JvYua4gQwN9wNgQKA3aZ8etWxCuPkgabMSWJF1jKX3jGT17HGo1SoefWvPlU3zFAWdVsP3h4fz7OZD/P5ym/ziKj1Pv5fDuscmsuNoaeuiWUVhxIAABgR6ExfmS9qnRwn315F/ocZhCmVJ5kFGDAjg57fHYWg28/nhUrYfLnEIzC99fISUjP38oZ3mcMVVlnb+Y6MDGRsVyIptx9qcWuopZk+K4ZUHb+LZzYd4+r2cXhVCrpWMiAjRCzmbrknZsI+TF+sJ9fGg8JKetFk38V8fHea1OYm2HiGP3BJL6vu5NDaZ+OOsm9hZUM5neSU8nBRDxp4izArEhlqWjlqX0VbrjUQGeTv08rCeaOYlxbL4nweICPDiQrWe6BAfGppMFFY0OGxQCFg2TSypIyrEm8r6Jt7+2aQ2G4a17CtypqIeRVEwmRVL7cTlrRLA8m5Lc7kXSpi/jjl/3c1/3T+av+8t4mRZPQMCvSirbaRa34ynRoXRpKBg2fvJy1NDqI8H56oa8fHUYGgyYbp8HME+HsSG+nK6vI7qxmbWzpvANycvUm9oZvuRUlvwqqhrYkCgF0aTmXOVetu0TctmadbH5mxE6/D5ahZm7OfytkoYmhWGhvvh66mx9VH5/vBw/ufzY7aRmZdmJfDpoZLWI14Z+zhZVs+QcF/S515pMLdk+nC+PVXRqvmbwzF9lm87ZqDV5dZlzAaT2bYU3TodtTJ5DH/acZxjJTUM7x/Ay7MSLEXL9U2snTfBVkRt/f2dOjLCVvRsXxNSWtPoUPjbFdMzo5dtY9l9o5yGgk17ilj+8REOL5/RKff19Hs5fHigmPvHRvLnOYmdcpvuQEZEhOjj7Edb7Fu0/+knY4kI9GZsVCBbcs/z2pxEFmce5NXPj7FoWjzLPz6CTqtmzbwJzJ4UQ1w/X/768AT+VVDOb+6KRwHOXmpgaLgfg0L9uFTfRKPRxAszR7H6oUSWZB4kp6jSdoJ68+tTl4tK9cSE+OKpURPk48GgUB9WZBVw300DAMvJ8XylnjXzxhMR4MXAYG/Hd+52rAWMnhoV9U0mjlyooaHJxNM/GIbm8nLgZrNCkLcWAE8PNYNCffn1+wdY8O4+3vvFLcyeFIPZrKDTqrlYZ8BkVvje8H4E+XgwckAAussFuoYmE+eqGkkeF0lDkwkzloCiVkGoryf1hmbqDCaGhvvy/NY8PjpQzLa8Emobm/lBfD88tWoampppaDJRWmPAy0PDL/++n9tX/B+rZ49jxIAAktO/tT1O60jXExn7qDOYiA725g8fHyYlYz9r5o5n0dR4DM0K/joNlfUGjpfVcaKsluTxUaz96hS/uCOOMxUNhPt7sXRzHv38dUQEeBEX5msJPsCy+0ZRZ2jmRGkdv96Uw3NbD/O9+DC+PVXBvFtiWm2+Zz2mVdsLLI3dVCpbwa015KyePY7VD1lWSy25e4RtKfqFaj2rthew7rGJvLO7kF9NHYpJgfwLNfzyvRwu1Tcxon8Aowc6bjRqbZA3LirIIYQArNpeQICXB7WNRtvoi72cokqS07+5ob+fZfeN4tnNh1oVlFqnT5bdN+qGbt/+9j66HEI+OlDcZgFrbydBRIhezPpu1qwotsLU1bPHkT5vAigKb3x9mpXJYzhTUc+Oo6VsWjCZ/c/fZXt3unj6CKaN7s+f5yTyzu4iRvb3R6NR89DEaDw0Kirqm1j547Gs2l4AXNkx2nqCig7xwUOrYki4n60Lqk6jZkJsMMPC/Xhuax6PvJXN6fJ63v7ZJKaN7s/q2eOYOCiEuH6+7YaRx++I40RZHWoVRAV7s/LzY3hqVOg81KiAKn0z4ZenVpTLtR4AEQFeAKx9+GYiAryobDDy6k/GsX7+JD548nb8dBpiQi21J2YgzN+TDw4Uo7VsIYUC/OLOwZwor+fMxTrWzBuPn86DGr1l5U+z2czDSTG8vvMUp8vqef7eUZy5WEej0cTtQ0I5V9VITIilIRtAeW0jP1u/l5mvfUVOUSWRQd7cOiSMokv1fHPiIp/llbL4rnj+5/NjpH95iqUzhmMwmWk2KTQ0mYgL9eX3H+SxdMYIPj5UQsqdg7lQ08j46CB2nawgdWOOZQWNovDq58eYv87SYbfZrJB9upKoYC9iQn155JZYnvnHQQaH+TiszLLvy1JWa+Dx2wcxf/1e5rXYETsyyLJLtnWFDQq2UJMYE2zZODTzkG3H7HOVev7zjsH87dGbHXY3Lq66El6Wf3yE6BAfh+Z7K5LH8EryGM5X6gn19XCYnrTWtTx375WgMOx3n7L23yec/n2s/fcJhv3uU4fLVmblc0d8P9u0iTUcbNpTxJLNloZ8nTF9Yl8T8uc5ia3ury+RqRkhejHrFA3Qqri1uKr1zrzOtNzH5tC5ahZs2MeoAQH84UejbcWKO46WAjh0Bc7ILrRtdpj2Wb6t34f16/PX7eHUxXr+8h/jHYbmrff76ufH+ObERTY/cZvD8eUUVfLoW3vQqFUoClTpjei0aozNZmJCfSipaaTxcpO0J6cM5o1vLHvfeHqo0Wk1LJkxnD98fJRT5XWs/PFYhxUsOUWVPPzmHuoMza02q4wK9qaizoCh2YyHRoVWreaO+DAuVDVy4Fw1Oq0KtUqNRg3x4X7sP1uNTgt+Ok/C/Dw4VlpPypTB3DW6v201yjObctl1+hJRQV7UNDZbVgp9fZriygbOVTUSH+FHSXUjXh4a1j48gR1HSxkXFcST7+1nUKgPBaX1PJIUw78Kynnq+0NZ9uFhFk0dxp92HOe1OYm88fVpUBR+fsdgfvleDoHeHpTWGlBhGS3yUKt5/t6RLN1yCEUBnVZNbKhlCm3ZfaNs202s+fIkv7tnJMnpu/jt9Hj+O6ug1UaCs9fu4s9zEm1TT9YdkpfePYK0T48S4ufJtrxS1CoYEORNSXUja+Ze2T6grWkhZ3sFtSymbau4du2/T5C27RhLZwxnwfeGXvVy+yLarwrKeXbzIWaOjeTDA8UO+yPdCFk140iCiBB9wPVuYOZs6bClsZiJc5UNtnqHq/WSuJEN1JwtKX30rT1EBVtGW5qazZypaMDQZMJXp6G+ycSI/gH8aupQfr3pAA1NJpbOGM4Xx8qJDvbmRFkdB85VE+7vydqHb7atILGvdzA0mwj29uDLgotYe756aFQMDfej2aRwvKwOLw/Lct7CCj3Dwv3w8dRgNCkcuVBDqK8nUSHenCqro9ZgIsBLQ02jiYeTYjhWVmerp9iSc578CzWU1Ro4U9FA/0Ad5TUG+vnrLMusA3WU1zZx16hwnr9vtMPJefLgEP47q4BZ4wbwQe4FFk6J442vzvDy5Y0g7ZdRv/jRYU6X15MyZQj/nXXMVpT75JTB/KvgIvklNVj34/TUqIgL8+U3dw0nJWM/i++KJ/3LUyhmhcggbx65NZbnPzjMHx4Yzad5JbYT9vNb89i0YHKrOpKUDfs4WV5HTIgvRpOZU+V1thqR32/J41hpLWvsCnxbrohpr5GfNXwsnj6clVnHWoUQq5aho60Q4ux37pXP8vnwQDFhfp58+NTtnVKT0p01KK4iQUQI0SnsA0TLRlKlNY22d/XQesSlM9kHBVsfkWo9Oo2aEQMCyD5VgVmBMxX1+HpquX9cJAunDCFlwz6KLjXQZDKz4fEktuSc5+C5aofeHvbvsB9dt4e4UF+e+v5Qfv2PAzQZm2mybAOEAnhqVXhq1Dz5/SH897YCVEBksDcXqvT8cdZNZO4/h8FkpqnZzNELtQC2EGJdtmzdo0ZBsRV0AvzHG7s5U9Fge8z9/D0xGM28+hPLlJr9c2vd6XjRtGH8+YsT/GR8FO/uKWLhnXEcvlDrEByto1JltY1syyvFy0NtGy0CiI/wpaDU0kPFQw1GM7ZCYpPJzK5Tl3jlwZtY/+0ZjpXW4n25h8cbX50iOsSHoxdqOXqhhlcevIk74vu1ajb2zPu5/PtYOXqjiUajibXzJthGQBZNi+evX57k8yNlDI/wt42wdWRExCoju5Dfb8nj5VkJ7baYt4aP24eG8fWJi22GEPvfuUfezOZEeT1hfp5crGtiRS8YqeguUqwqhOgU1qJX+6WY9kuCNy2YbNuzpiv3BbIWTD79Xg7P3TuKi/VNtpP48gcSyPr1FMZEBRLs40m83fRP+rwJfPL0HYyODGTD7kJOX6xnTFQg7z6exJioQIfahC0554kL9WVgkDe//scBmk0mQMX3h/djxAB/PNQqmpoVfL20vP6vk8SG+qBWqzhXqec3dw3nhQ8Ps/Sekbw4czRnL+nx0WkI8/WwhZDqxmYOnKumyWjG0GzidHk9Sy8vlwUI8vHkcq0taqC8tokld1tqdKydTIur9OQUVbI48yDpc8ez69QlfvXDoWzYU8S8STGs+fI08+xGnqzLwBdOGUK9wcTgfr5o1SrSZiWg1agAKCitR6sGLw81HloNS6bHsyKrgNrGZvKKaxga7sfLnx7lkcmD8PawdKC9VN8EKhUHzlaRX2IZAYrv709GdiGrH0q07WW0r/ASJ8rq8NSq8dKqmT46gmmj+zsUwP7iziEMj/CnvLaRCbEhtsdqLXq2dh22fw6scooqWZl1jJdnJbAy65jTeiKrBd8bagshtw8NazeEAHxVUG5rcrf24Qms6MM1HF1NRkSEEFd1I1Mrnck6TG/d18X+eOzrYVoeT8sOofbHbj3ZWetWFrz7HcVVjWjUKtY+PMFWf/BExn7UwPnqRoJ9PAj31+HlqaHJaOZ8tZ5Xf2yp7bDupKxVqdhXVEWEvychfjpOlddhaFbw1FoCzdqHx3PwXLVleeqmXCrrmy7vLmxhHRGxLmO21sx8kV9mm4KwNi1LuXMwb3x9mmemDWPtV6cdpijsR5PsO98OCvHm3eyztudiYJAX/+8/xpP26VHOVzdyvlLPb6fHU1Jj4LszlRy5UMPah8dTVW/k2c2HmDMphk3fneWuUeH84s4hrZ5f688q3F9nex5byimq5JG39vBOi6Xa1ikX6/YD1p+l/c/LuizYvkbkx+nf8sdZzkctlm05xDvZRR0aEbEWpg7t58uKH4+xjcZYa0Z6Qw1HV5OpGSFEr9ORYfq2XEuQWpmVT21js0OHUGtfi7TP8rlYZ6Cu0bJT7+rZ41iz8yT1hmbOVuqp1Rup0ht5bHIsr2w7xrMzhlNQVsf2I6WEB+g4UWZ5hx0R4ElsiGWTv7TP8qmsb+JUeR2RgV6U1hoI99dxrqrRVsBqDSPJ6d/w3L2jHIozl0wfbmlaNmMEf9pRwB8eSGBx5kHWPTbRVr9jrX0Z0d8ffy8tATotaduO2TrRWkUFedFoNNNkMrNwymBWbT/Oyz9KsE05WdurD+3nx9/3FPGLOwfzu3tGOvx8rP1KrC3ZrSNXzp576/Nqv/rGevm4qCCWf3zEIdxYr/9s5kFOlNU5hI7iKj2/33KIfx0rJzE6ELVaRWbKbQDc+79fcvhCLY8kxbB81k0ONSIzxw1sFVxHPf8ZkUHetsLUlgWsvaGGo6tJEBFC9CpX22+nO+8fcNqMbNnWPLYfKeUvc8cz72/ZLLtvFHfE97M1ivv9B3mAgqdGzfmqRsZEB1JWY2BQiA97zlwiMtCL4upG1lyuoXj0rT2cKKvjlsEhHC2pbTXKMXvtLtuqlpajHY/cEsvyj4/w/RHhLJwyhDU7T5J/oYbVDyXyVUE5SzYfwlMDCioiA704V9mIydpNV6ti4y8mW6be7FZxDI3wIzl9F7GhPpyr1PP47XG89fVp2y7M1uOyXy3Tctnt1VZmXe3na728rMZyvGW1Bt6eP8khcD31932crzLY6jmsIyERAZ5seeJKsak1jMSGevPez6+EnY7uASXaJzUiQohew9kJwFpj0LJmoDvuPyO70NbAy/7+lz+QQGbKrew4Wsrh5TOYPSmGjGxLSHjj69MMCvXhgydv5/2FtzIxzrL649YhoXxXVEl0iA+X9EbWXO4wGhnkzds/m8TIAQGcKKtj3WMTeemTIw7Hs2nBZHLPVbEieYxtR2FrGHlndyF/npOIv5elqdvpi/W2otgXPz6MTqvCZIah/fx47xeTuX1YqO3xNl0eIimu0vPxoQu88uBNLNuaxx8+PsqS6cMprGggLsyHx24dxF2jIli4YR/bD5eQnP4Nh89XYzCaOHCumlnjIsnILrSstLkllumrv7S1frff3bflz7O9HX2t1wsP8GJsVBADA715+M1snsjYb9vksdZgWSX1woeHycgu5J3sIh5JimHLE7c7/LxmjhtIbKg35yv1DvdzrdsqiBsnIyJCCLfm6vqUzlh67HQzuMujKqF+OnLPVpKZclur+3B2+9eyWWLLaayVWfl8WVBOQ5MJT40af28PHr8tjiWbDxLgpaW6wUijyUR8uD93xvez1dw8kbGfynoD56saLVM1OedtPWH+8+3vOFZSw2+nD+fVHccZ0s/SHO35rYdJn2tpq2+t9/jqxMV2+9a0t1S35eNM3ZRLbaORoooGFBVEBng57Cg9eXAo/325kNX6OG5kek9cG7eYmnn55Zf55JNPyM3NxdPTk6qqqmu+DQkiQoie7Gqh4X8+P8bm/ee7ZDfZ9k7q1uOyNqdbO28C0xP6207wI/r7s/yBBIfRoAf/8g1/eCCBaaP7t5qq+s+3v6Pk8l5Cpy7WMzjs8hLo93NRqVS804F+M9caEqyPb2i4r6325rfTh7PrVAWP3BLL4syDTvuLdDTsiBvjFlMzTU1N/OQnPyElJaWr7kIIIdxaezssA5TXGshMmezQUr0zFFfpba3Wnd324ukjLPvFbD7I2nkT2JBdSHGVZYpi9exxnL5Y79DaPTLIm92/m2qrBbGfpogM8ublWQmYgSe+P4TBYb6crqinrM6ASmVpr19W29ihENLeUl1nj2/tvAlcqGpErbKs+nl1ewG3Dwm1FevOTYpl3WMTmb9+LzlFlVd9XoSLKF1s3bp1SmBg4HV9b3V1tQIo1dXVnXtQQgjhQucrG5R5f9utnK9scPp5V9/2+coG5fZXvlD2F15q83v+e9vRDt/nf287qmzLK1Zil3ysfHemQtmw+4wSu+RjZcPuMw6Xd+R4O3r5/sJLyv2vfaWMXvaZMvXVfyuxSz5Wbv3jdiV2ycfKxuxCh+/ZX3hJuenFbcr9r33VJc95TzPq+c9aPUdWG7MLlVHPf3bD93Et52+3KlY1GAzU1NQ4fAghRG/SlcW3Hblt+2JX63RFy+tYG6F11NykWDbsLiIzZTLLPzrCim35vDwrgRXb8vnLv062OwJxLcWh9j1RXvzwsGVH6Z+OI9jHg7HRgVzSG4kN8eblz446NDeLCPAiLtQXs6K0+bwkp3/TZkO0ztjR15101+7CHXbDsecqrmVE5IUXXrBubunwISMiQoje4r+3HW3zXfi1jkRcz2139v3bjyzYjzxYRyxuenGbsr/wUqeMQFiP/b+3HVXu+/OXyud5F2y3eb6yQXn+g0PKT9Z8qyx8d6/y4F++bvV97T3m/YWXlLH/lWUbJbJq6/KebmN2oTLIbvSo5ec36lpGRK4piCxZssRpULD/OHrU8Zf4WoJIY2OjUl1dbfs4e/asBBEhhHBTLUOI9YTd1v+7cgrqapd3RMvQ0VtDiJU1fPzy7/s7NYQoyrUFkWtaNVNeXk5FRUW71xk8eDCenp62z9evX09qaqqsmhFCiF7GflWQfddXawfUHUdLWTx9BDlFlbz0yREyU27rtCXXXbWsu6M7+vYWT7+Xw4cHirl/bCR/npPYabfrFst3rSSICCGE6Ek6uqNvT2etCZk5NpKPDhR36h4613L+1nbKPTpRVFTEpUuXKCoqwmQykZubC8DQoUPx8/PrqrsVQgghrlvLHX1HDQjolSMi9u37Z0+K4bYhoTy7+RBAt2/o12VBZNmyZbz99tu2zxMTLUM+//rXv/je977XVXcrhBBCXBfrtIx1OmbUgACHz3uLliEEroQPV4QRafEuhBCiz2sZQq52eU82etk2lt03ymnY2LSnqFN2F3arGpEbIUFECCFEd7Avtm3JvthWdIwEESGEEEK4jFvsNSOEEEIIcTUSRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TISRIQQQgjhMhJEhBBCCOEyEkSEEEII4TJdtvtuZ7B2n6+pqXHxkQghhBCio6zn7Y7sIuPWQaS2thaA6OhoFx+JEEIIIa5VbW0tgYGB7V7HrTe9M5vNFBcX4+/vj0qlcvXhdImamhqio6M5e/Zsn9vYTx67PHZ57H2HPPa+9dgVRaG2tpbIyEjU6varQNx6REStVhMVFeXqw+gWAQEBfeYXtCV57PLY+xp57PLY+4KrjYRYSbGqEEIIIVxGgogQQgghXEaCiIvpdDpeeOEFdDqdqw+l28ljl8fe18hjl8cuWnPrYlUhhBBC9G4yIiKEEEIIl5EgIoQQQgiXkSAihBBCCJeRICKEEEIIl5Eg4mKvv/46gwYNwsvLi6SkJPbs2ePqQ+pyX375JTNnziQyMhKVSsUHH3zg6kPqNmlpaUycOBF/f3/Cw8P50Y9+xLFjx1x9WN0iPT2dMWPG2Jo6TZ48mc8++8zVh9XtXnnlFVQqFampqa4+lG7x4osvolKpHD5GjBjh6sPqNufPn2fevHmEhobi7e3NTTfdxHfffefqw3IrEkRcaNOmTSxatIgXXniB/fv3M3bsWKZPn05ZWZmrD61L1dfXM3bsWF5//XVXH0q327lzJ08++SS7d+9m+/btGI1G7rrrLurr6119aF0uKiqKV155hX379vHdd9/xgx/8gAceeIDDhw+7+tC6zd69e1m7di1jxoxx9aF0q9GjR3PhwgXbx9dff+3qQ+oWlZWV3HbbbXh4ePDZZ59x5MgRXn31VYKDg119aO5FES4zadIk5cknn7R9bjKZlMjISCUtLc2FR9W9AGXLli2uPgyXKSsrUwBl586drj4UlwgODlb+9re/ufowukVtba0ybNgwZfv27cqUKVOUX/3qV64+pG7xwgsvKGPHjnX1YbjEkiVLlNtvv93Vh+H2ZETERZqamti3bx9Tp061XaZWq5k6dSq7du1y4ZGJ7lRdXQ1ASEiIi4+ke5lMJjZu3Eh9fT2TJ0929eF0iyeffJJ7773X4W++rzh+/DiRkZEMHjyYuXPnUlRU5OpD6hYffvghN998Mz/5yU8IDw8nMTGRN954w9WH5XYkiLjIxYsXMZlMREREOFweERFBSUmJi45KdCez2Uxqaiq33XYbCQkJrj6cbnHo0CH8/PzQ6XQsXLiQLVu2MGrUKFcfVpfbuHEj+/fvJy0tzdWH0u2SkpJYv34927ZtIz09ndOnT3PHHXdQW1vr6kPrcqdOnSI9PZ1hw4aRlZVFSkoKTz/9NG+//barD82tuPXuu0L0Zk8++SR5eXl9Zr4cYPjw4eTm5lJdXc0///lPHn30UXbu3Nmrw8jZs2f51a9+xfbt2/Hy8nL14XS7u+++2/b/MWPGkJSURGxsLO+//z6PP/64C4+s65nNZm6++Wb++Mc/ApCYmEheXh5r1qzh0UcfdfHRuQ8ZEXGRsLAwNBoNpaWlDpeXlpbSv39/Fx2V6C5PPfUUH3/8Mf/617+Iiopy9eF0G09PT4YOHcqECRNIS0tj7Nix/O///q+rD6tL7du3j7KyMsaPH49Wq0Wr1bJz507+/Oc/o9VqMZlMrj7EbhUUFER8fDwnTpxw9aF0uQEDBrQK2SNHjuwzU1MdJUHERTw9PZkwYQJffPGF7TKz2cwXX3zRZ+bM+yJFUXjqqafYsmUL//d//0dcXJyrD8mlzGYzBoPB1YfRpX74wx9y6NAhcnNzbR8333wzc+fOJTc3F41G4+pD7FZ1dXWcPHmSAQMGuPpQutxtt93Wanl+QUEBsbGxLjoi9yRTMy60aNEiHn30UW6++WYmTZrE6tWrqa+vZ/78+a4+tC5VV1fn8G7o9OnT5ObmEhISQkxMjAuPrOs9+eST/P3vf2fr1q34+/vb6oECAwPx9vZ28dF1raVLl3L33XcTExNDbW0tf//73/n3v/9NVlaWqw+tS/n7+7eqAfL19SU0NLRP1Ab95je/YebMmcTGxlJcXMwLL7yARqNhzpw5rj60LvfrX/+aW2+9lT/+8Y/89Kc/Zc+ePfz1r3/lr3/9q6sPzb24etlOX/faa68pMTExiqenpzJp0iRl9+7drj6kLvevf/1LAVp9PProo64+tC7n7HEDyrp161x9aF3uZz/7mRIbG6t4enoq/fr1U374wx8qn3/+uasPyyX60vLd2bNnKwMGDFA8PT2VgQMHKrNnz1ZOnDjh6sPqNh999JGSkJCg6HQ6ZcSIEcpf//pXVx+S21EpiqK4KAMJIYQQoo+TGhEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQriMBBEhhBBCuIwEESGEEEK4jAQRIYQQQrjM/wdlDlkfC6+T1gAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Perform PCA to reduce the dimensionality to 2\n",
    "pca = PCA(n_components=2)\n",
    "reduced_matrix = pca.fit_transform(cosine_sim_mat)\n",
    "\n",
    "# Create a scatter plot of the reduced matrix\n",
    "sns.scatterplot(x=reduced_matrix[:, 0], y=reduced_matrix[:, 1], marker='x')\n",
    "\n",
    "# Label the points with the index of the corresponding item or document\n",
    "# for i, (x, y) in enumerate(reduced_matrix):\n",
    "#     plt.text(x, y, str(i))\n",
    "    \n",
    "# Show the plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "id": "255c3ddb",
   "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>course_name</th>\n",
       "      <th>university</th>\n",
       "      <th>difficulty_level</th>\n",
       "      <th>course_rating</th>\n",
       "      <th>course_url</th>\n",
       "      <th>course_description</th>\n",
       "      <th>course_skills</th>\n",
       "      <th>clean_course_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Write A Feature Length Screenplay For Film Or ...</td>\n",
       "      <td>Michigan State University</td>\n",
       "      <td>Beginner</td>\n",
       "      <td>4.8</td>\n",
       "      <td>https://www.coursera.org/learn/write-a-feature...</td>\n",
       "      <td>Write a Full Length Feature Film Script  In th...</td>\n",
       "      <td>Drama  Comedy  peering  screenwriting  film  D...</td>\n",
       "      <td>Write Feature Length Screenplay Film Television</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Business Strategy: Business Model Canvas Analy...</td>\n",
       "      <td>Coursera Project Network</td>\n",
       "      <td>Beginner</td>\n",
       "      <td>4.8</td>\n",
       "      <td>https://www.coursera.org/learn/canvas-analysis...</td>\n",
       "      <td>By the end of this guided project, you will be...</td>\n",
       "      <td>Finance  business plan  persona (user experien...</td>\n",
       "      <td>Business Strategy Business Model Canvas Analys...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Silicon Thin Film Solar Cells</td>\n",
       "      <td>�cole Polytechnique</td>\n",
       "      <td>Advanced</td>\n",
       "      <td>4.1</td>\n",
       "      <td>https://www.coursera.org/learn/silicon-thin-fi...</td>\n",
       "      <td>This course consists of a general presentation...</td>\n",
       "      <td>chemistry  physics  Solar Energy  film  lambda...</td>\n",
       "      <td>Silicon Thin Film Solar Cells</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Finance for Managers</td>\n",
       "      <td>IESE Business School</td>\n",
       "      <td>Intermediate</td>\n",
       "      <td>4.8</td>\n",
       "      <td>https://www.coursera.org/learn/operational-fin...</td>\n",
       "      <td>When it comes to numbers, there is always more...</td>\n",
       "      <td>accounts receivable  dupont analysis  analysis...</td>\n",
       "      <td>Finance Managers</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Retrieve Data using Single-Table SQL Queries</td>\n",
       "      <td>Coursera Project Network</td>\n",
       "      <td>Beginner</td>\n",
       "      <td>4.6</td>\n",
       "      <td>https://www.coursera.org/learn/single-table-sq...</td>\n",
       "      <td>In this course you�ll learn how to effectively...</td>\n",
       "      <td>Data Analysis  select (sql)  database manageme...</td>\n",
       "      <td>Retrieve Data SingleTable SQL Queries</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         course_name   \n",
       "0  Write A Feature Length Screenplay For Film Or ...  \\\n",
       "1  Business Strategy: Business Model Canvas Analy...   \n",
       "2                      Silicon Thin Film Solar Cells   \n",
       "3                               Finance for Managers   \n",
       "4       Retrieve Data using Single-Table SQL Queries   \n",
       "\n",
       "                  university difficulty_level course_rating   \n",
       "0  Michigan State University         Beginner           4.8  \\\n",
       "1   Coursera Project Network         Beginner           4.8   \n",
       "2        �cole Polytechnique         Advanced           4.1   \n",
       "3       IESE Business School     Intermediate           4.8   \n",
       "4   Coursera Project Network         Beginner           4.6   \n",
       "\n",
       "                                          course_url   \n",
       "0  https://www.coursera.org/learn/write-a-feature...  \\\n",
       "1  https://www.coursera.org/learn/canvas-analysis...   \n",
       "2  https://www.coursera.org/learn/silicon-thin-fi...   \n",
       "3  https://www.coursera.org/learn/operational-fin...   \n",
       "4  https://www.coursera.org/learn/single-table-sq...   \n",
       "\n",
       "                                  course_description   \n",
       "0  Write a Full Length Feature Film Script  In th...  \\\n",
       "1  By the end of this guided project, you will be...   \n",
       "2  This course consists of a general presentation...   \n",
       "3  When it comes to numbers, there is always more...   \n",
       "4  In this course you�ll learn how to effectively...   \n",
       "\n",
       "                                       course_skills   \n",
       "0  Drama  Comedy  peering  screenwriting  film  D...  \\\n",
       "1  Finance  business plan  persona (user experien...   \n",
       "2  chemistry  physics  Solar Energy  film  lambda...   \n",
       "3  accounts receivable  dupont analysis  analysis...   \n",
       "4  Data Analysis  select (sql)  database manageme...   \n",
       "\n",
       "                                   clean_course_name  \n",
       "0    Write Feature Length Screenplay Film Television  \n",
       "1  Business Strategy Business Model Canvas Analys...  \n",
       "2                      Silicon Thin Film Solar Cells  \n",
       "3                                   Finance Managers  \n",
       "4              Retrieve Data SingleTable SQL Queries  "
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "id": "be88edf1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get Course Index\n",
    "course_indices = pd.Series(df.index,index=df['course_name']).drop_duplicates()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "id": "54757a1f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "course_name\n",
       "Write A Feature Length Screenplay For Film Or Television                 0\n",
       "Business Strategy: Business Model Canvas Analysis with Miro              1\n",
       "Silicon Thin Film Solar Cells                                            2\n",
       "Finance for Managers                                                     3\n",
       "Retrieve Data using Single-Table SQL Queries                             4\n",
       "                                                                      ... \n",
       "Capstone: Retrieving, Processing, and Visualizing Data with Python    3517\n",
       "Patrick Henry: Forgotten Founder                                      3518\n",
       "Business intelligence and data analytics: Generate insights           3519\n",
       "Rigid Body Dynamics                                                   3520\n",
       "Architecting with Google Kubernetes Engine: Production                3521\n",
       "Length: 3522, dtype: int64"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "course_indices\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "id": "e8c53a18",
   "metadata": {},
   "outputs": [],
   "source": [
    "def recommend_course(title,num_of_rec=10):\n",
    "    idx = course_indices[title]\n",
    "    scores = list(enumerate(cosine_sim_mat[idx]))\n",
    "    sorted_scores = sorted(scores,key=lambda x:x[1],reverse=True)\n",
    "    selected_course_indices = [i[0] for i in sorted_scores[1:]]\n",
    "    selected_course_scores = [i[1] for i in sorted_scores[1:]]\n",
    "    result = df['course_name'].iloc[selected_course_indices]\n",
    "    rec_df = pd.DataFrame(result)\n",
    "    rec_df['similarity_scores'] = selected_course_scores\n",
    "    return rec_df.head(num_of_rec)  \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "id": "8cd49447",
   "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>course_name</th>\n",
       "      <th>similarity_scores</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>419</th>\n",
       "      <td>Finance for Non-Financial Managers</td>\n",
       "      <td>0.816497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>181</th>\n",
       "      <td>Construction Finance</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>668</th>\n",
       "      <td>How to finance your venture ?</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1236</th>\n",
       "      <td>Behavioral Finance</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3431</th>\n",
       "      <td>Innovative Finance: Hacking finance to change ...</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3448</th>\n",
       "      <td>Finance for Startups</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>646</th>\n",
       "      <td>Finance for Non-Financial Professionals</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>832</th>\n",
       "      <td>Finance for Everyone: Decisions</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>959</th>\n",
       "      <td>Finance for Everyone Capstone Project</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1010</th>\n",
       "      <td>Reinforcement Learning in Finance</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1164</th>\n",
       "      <td>Corporate Finance Essentials</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1567</th>\n",
       "      <td>Finance for Everyone: Debt</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1891</th>\n",
       "      <td>Accounting and Finance for IT professionals</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1985</th>\n",
       "      <td>Introduction to Finance: The Basics</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2068</th>\n",
       "      <td>Principles of Sustainable Finance</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2082</th>\n",
       "      <td>Finance for Non-Finance Professionals</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2851</th>\n",
       "      <td>Finance for Everyone: Markets</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3110</th>\n",
       "      <td>Finance For Everyone: Value</td>\n",
       "      <td>0.408248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>306</th>\n",
       "      <td>Blockchain, Cryptoassets, and Decentralized Fi...</td>\n",
       "      <td>0.353553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>348</th>\n",
       "      <td>On Strategy : What Managers Can Learn from Phi...</td>\n",
       "      <td>0.353553</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                            course_name  similarity_scores\n",
       "419                  Finance for Non-Financial Managers           0.816497\n",
       "181                                Construction Finance           0.500000\n",
       "668                       How to finance your venture ?           0.500000\n",
       "1236                                 Behavioral Finance           0.500000\n",
       "3431  Innovative Finance: Hacking finance to change ...           0.500000\n",
       "3448                               Finance for Startups           0.500000\n",
       "646             Finance for Non-Financial Professionals           0.408248\n",
       "832                     Finance for Everyone: Decisions           0.408248\n",
       "959               Finance for Everyone Capstone Project           0.408248\n",
       "1010                  Reinforcement Learning in Finance           0.408248\n",
       "1164                       Corporate Finance Essentials           0.408248\n",
       "1567                         Finance for Everyone: Debt           0.408248\n",
       "1891        Accounting and Finance for IT professionals           0.408248\n",
       "1985                Introduction to Finance: The Basics           0.408248\n",
       "2068                  Principles of Sustainable Finance           0.408248\n",
       "2082              Finance for Non-Finance Professionals           0.408248\n",
       "2851                      Finance for Everyone: Markets           0.408248\n",
       "3110                        Finance For Everyone: Value           0.408248\n",
       "306   Blockchain, Cryptoassets, and Decentralized Fi...           0.353553\n",
       "348   On Strategy : What Managers Can Learn from Phi...           0.353553"
      ]
     },
     "execution_count": 151,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "recommend_course('Finance for Managers',20)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8a09f0b4",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "75672d8b",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}