File size: 10,024 Bytes
fc07932 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 |
{
"cells": [
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"from gensim import corpora\n",
"from gensim.similarities import SparseMatrixSimilarity\n",
"from gensim.models import TfidfModel\n",
"import pandas as pd\n",
"import gensim\n",
"import pprint\n",
"from gensim import corpora\n",
"from gensim.utils import simple_preprocess\n",
"from gensim.models import TfidfModel\n",
"from gensim.parsing import strip_tags, strip_numeric, \\\n",
" strip_multiple_whitespaces, stem_text, strip_punctuation, \\\n",
" remove_stopwords, preprocess_string\n",
"import re\n",
"import os\n",
"from typing import List"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"ename": "NameError",
"evalue": "name 'strip_tags' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
"Cell \u001b[0;32mIn[6], line 5\u001b[0m\n\u001b[1;32m 1\u001b[0m transform_to_lower \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mlambda\u001b[39;00m s: s\u001b[38;5;241m.\u001b[39mlower()\n\u001b[1;32m 2\u001b[0m remove_single_char \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mlambda\u001b[39;00m s: re\u001b[38;5;241m.\u001b[39msub(\u001b[38;5;124mr\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m\\\u001b[39m\u001b[38;5;124ms+\u001b[39m\u001b[38;5;124m\\\u001b[39m\u001b[38;5;124mw\u001b[39m\u001b[38;5;132;01m{1}\u001b[39;00m\u001b[38;5;124m\\\u001b[39m\u001b[38;5;124ms+\u001b[39m\u001b[38;5;124m'\u001b[39m, \u001b[38;5;124m'\u001b[39m\u001b[38;5;124m'\u001b[39m, s)\n\u001b[1;32m 4\u001b[0m cleaning_filters \u001b[38;5;241m=\u001b[39m [\n\u001b[0;32m----> 5\u001b[0m \u001b[43mstrip_tags\u001b[49m,\n\u001b[1;32m 6\u001b[0m strip_numeric,\n\u001b[1;32m 7\u001b[0m strip_punctuation, \n\u001b[1;32m 8\u001b[0m strip_multiple_whitespaces, \n\u001b[1;32m 9\u001b[0m transform_to_lower,\n\u001b[1;32m 10\u001b[0m remove_stopwords,\n\u001b[1;32m 11\u001b[0m remove_single_char\n\u001b[1;32m 12\u001b[0m ]\n\u001b[1;32m 14\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mgensim_tokenizer\u001b[39m(docs: List[\u001b[38;5;28mstr\u001b[39m]):\n\u001b[1;32m 15\u001b[0m \u001b[38;5;250m \u001b[39m\u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[1;32m 16\u001b[0m \u001b[38;5;124;03m Tokenizes a list of strings using a series of cleaning filters.\u001b[39;00m\n\u001b[1;32m 17\u001b[0m \n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 22\u001b[0m \u001b[38;5;124;03m List[List[str]]: A list of tokenized documents, where each document is represented as a list of tokens.\u001b[39;00m\n\u001b[1;32m 23\u001b[0m \u001b[38;5;124;03m \"\"\"\u001b[39;00m\n",
"\u001b[0;31mNameError\u001b[0m: name 'strip_tags' is not defined"
]
}
],
"source": [
"\n",
"transform_to_lower = lambda s: s.lower()\n",
"remove_single_char = lambda s: re.sub(r'\\s+\\w{1}\\s+', '', s)\n",
"\n",
"cleaning_filters = [\n",
" strip_tags,\n",
" strip_numeric,\n",
" strip_punctuation, \n",
" strip_multiple_whitespaces, \n",
" transform_to_lower,\n",
" remove_stopwords,\n",
" remove_single_char\n",
"]\n",
"\n",
"def gensim_tokenizer(docs: List[str]):\n",
" \"\"\"\n",
" Tokenizes a list of strings using a series of cleaning filters.\n",
"\n",
" Args:\n",
" docs (List[str]): A list of strings to be tokenized.\n",
"\n",
" Returns:\n",
" List[List[str]]: A list of tokenized documents, where each document is represented as a list of tokens.\n",
" \"\"\"\n",
" tokenized_docs = list()\n",
" for doc in docs:\n",
" processed_words = preprocess_string(doc, cleaning_filters)\n",
" tokenized_docs.append(processed_words)\n",
" \n",
" return tokenized_docs\n",
"\n",
"\n",
"def cleaning_pipe(document):\n",
" \"\"\"\n",
" Applies a series of cleaning steps to a document.\n",
"\n",
" Args:\n",
" document (str): The document to be cleaned.\n",
"\n",
" Returns:\n",
" list: A list of processed words after applying the cleaning filters.\n",
" \"\"\"\n",
" # Invoking gensim.parsing.preprocess_string method with set of filters\n",
" processed_words = preprocess_string(document, cleaning_filters)\n",
" return processed_words\n",
"\n",
"\n",
"def get_gensim_dictionary(tokenized_docs: List[str], dict_name: str = \"corpus\", save_dict: bool = False):\n",
" \"\"\"\n",
" Create dictionary of words in preprocessed corpus and saves the dict object\n",
" \"\"\"\n",
" dictionary = corpora.Dictionary(tokenized_docs)\n",
" if save_dict: \n",
" parent_folder = \"/Users/luis.morales/Desktop/arxiv-paper-recommender/models/nlp_dictionaries\"\n",
" dictionary.save(f'{parent_folder}/{dict_name}.dict')\n",
" return dictionary\n",
"\n",
"\n",
"def get_closest_n(index_matrix: SparseMatrixSimilarity, query: str, n: int):\n",
" '''\n",
" Retrieves the top matching documents as per cosine similarity\n",
" between the TF-IDF vector of the query and all documents.\n",
"\n",
" Args:\n",
" query (str): The query string to find matching documents.\n",
" n (int): The number of closest documents to retrieve.\n",
"\n",
" Returns:\n",
" numpy.ndarray: An array of indices representing the top matching documents.\n",
" '''\n",
" # Clean the query document using cleaning_pipe function\n",
" query_document = cleaning_pipe(query)\n",
"\n",
" # Convert the query document to bag-of-words representation\n",
" query_bow = dictionary.doc2bow(query_document)\n",
"\n",
" # Calculate similarity scores between the query and all documents using TF-IDF model\n",
" sims = index_matrix[index_matrix[query_bow]]\n",
"\n",
" # Get the indices of the top n closest documents based on similarity scores\n",
" top_idx = sims.argsort()[-1 * n:][::-1]\n",
"\n",
" return top_idx\n",
"\n",
"\n",
"def get_recomendations_metadata(query: str, df: pd.DataFrame, n: int):\n",
" '''\n",
" Retrieves metadata recommendations based on a query using cosine similarity.\n",
"\n",
" Args:\n",
" query (str): The query string for which recommendations are sought.\n",
" n (int): The number of recommendations to retrieve.\n",
" df (pd.DataFrame): The DataFrame containing metadata information.\n",
"\n",
" Returns:\n",
" pd.DataFrame: A DataFrame containing the recommended metadata, reset with a new index.\n",
" '''\n",
" # Get the indices of the closest matching documents based on the query\n",
" recommendations_idxs = get_closest_n(query, n)\n",
" \n",
" # Retrieve the recommended metadata rows from the DataFrame based on the indices\n",
" recommendations_metadata = df.iloc[recommendations_idxs]\n",
" \n",
" # Reset the index of the recommended metadata DataFrame\n",
" recommendations_metadata = recommendations_metadata.reset_index(drop=True)\n",
" \n",
" return recommendations_metadata"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"corpus = corpora.Dictionary.load()"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"td_idf_model = TfidfModel.load(\"/Users/luis.morales/Desktop/arxiv-paper-recommender/models/tfidf/SemanticSherlock.model\")"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"similarities = SparseMatrixSimilarity.load(\"/Users/luis.morales/Desktop/arxiv-paper-recommender/models/similarities_matrix/LanguageLiberatorSimilarities/LanguageLiberator\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
" query = args.query\n",
" \n",
" df = pd.read_parquet(\"/Users/luis.morales/Desktop/arxiv-paper-recommender/data/processed/reduced_arxiv_papers.parquet.gzip\")\n",
" \n",
" dict_corpus = Dictionary.load(\"/Users/luis.morales/Desktop/arxiv-paper-recommender/models/dictionaries/LanguageLiberator.dict\")\n",
" \n",
" td_idf_model = TfidfModel.load(\"/Users/luis.morales/Desktop/arxiv-paper-recommender/models/tfidf/SemanticSherlock.model\")\n",
" \n",
" similarities = SparseMatrixSimilarity.load(\"/Users/luis.morales/Desktop/arxiv-paper-recommender/models/similarities_matrix/LanguageLiberatorSimilarities/LanguageLiberator\")\n",
" \n",
" results_df = get_recomendations_metadata(query=query, df=df, n=3)\n",
" print(results_df.head())"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3.11.4 ('arxiv-env': venv)",
"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": "3.11.4"
},
"orig_nbformat": 4,
"vscode": {
"interpreter": {
"hash": "aae17c2ae2f38cc6f211be9b71a2aa280701d8462782cbc1f67caa83a1603363"
}
}
},
"nbformat": 4,
"nbformat_minor": 2
}
|