This commit is contained in:
Jakub Pokrywka 2022-03-22 12:32:31 +01:00
parent 67e6944843
commit 11ce1b1e10
3 changed files with 1713 additions and 0 deletions

1109
cw/03a_tfidf.ipynb Normal file

File diff suppressed because it is too large Load Diff

View File

@ -0,0 +1,89 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Logo 1](https://git.wmi.amu.edu.pl/AITech/Szablon/raw/branch/master/Logotyp_AITech1.jpg)\n",
"<div class=\"alert alert-block alert-info\">\n",
"<h1> Ekstrakcja informacji </h1>\n",
"<h2> 3. <i>tfidf (1)</i> [ćwiczenia]</h2> \n",
"<h3> Jakub Pokrywka (2021)</h3>\n",
"</div>\n",
"\n",
"![Logo 2](https://git.wmi.amu.edu.pl/AITech/Szablon/raw/branch/master/Logotyp_AITech2.jpg)"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"def word_to_index(word):\n",
" vec = np.zeros(len(vocabulary))\n",
" if word in vocabulary:\n",
" idx = vocabulary.index(word)\n",
" vec[idx] = 1\n",
" else:\n",
" vec[-1] = 1\n",
" return vec"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def tf(document):\n",
" document_vector = None\n",
" for word in document:\n",
" if document_vector is None:\n",
" document_vector = word_to_index(word)\n",
" else:\n",
" document_vector += word_to_index(word)\n",
" return document_vector"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def similarity(query, document):\n",
" numerator = np.sum(query * document)\n",
" denominator = np.sqrt(np.sum(query*query)) * np.sqrt(np.sum(document*document)) \n",
" return numerator / denominator"
]
}
],
"metadata": {
"author": "Jakub Pokrywka",
"email": "kubapok@wmi.amu.edu.pl",
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"lang": "pl",
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.3"
},
"subtitle": "3.tfidf (1)[ćwiczenia]",
"title": "Ekstrakcja informacji",
"year": "2021"
},
"nbformat": 4,
"nbformat_minor": 4
}

View File

@ -0,0 +1,515 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Logo 1](https://git.wmi.amu.edu.pl/AITech/Szablon/raw/branch/master/Logotyp_AITech1.jpg)\n",
"<div class=\"alert alert-block alert-info\">\n",
"<h1> Ekstrakcja informacji </h1>\n",
"<h2> 3. <i>tfidf (2)</i> [ćwiczenia]</h2> \n",
"<h3> Jakub Pokrywka (2021)</h3>\n",
"</div>\n",
"\n",
"![Logo 2](https://git.wmi.amu.edu.pl/AITech/Szablon/raw/branch/master/Logotyp_AITech2.jpg)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Zajecia 3\n",
"\n",
"Przydatne materiały:\n",
"\n",
"https://scikit-learn.org/0.19/datasets/twenty_newsgroups.html\n",
"\n",
"https://scikit-learn.org/stable/modules/generated/sklearn.feature_extraction.text.TfidfVectorizer.html\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Importy"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"import sklearn.metrics\n",
"\n",
"from sklearn.datasets import fetch_20newsgroups\n",
"\n",
"from sklearn.feature_extraction.text import TfidfVectorizer"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Zbiór danych"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"newsgroups = fetch_20newsgroups()['data']"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"11314"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"len(newsgroups)"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"From: lerxst@wam.umd.edu (where's my thing)\n",
"Subject: WHAT car is this!?\n",
"Nntp-Posting-Host: rac3.wam.umd.edu\n",
"Organization: University of Maryland, College Park\n",
"Lines: 15\n",
"\n",
" I was wondering if anyone out there could enlighten me on this car I saw\n",
"the other day. It was a 2-door sports car, looked to be from the late 60s/\n",
"early 70s. It was called a Bricklin. The doors were really small. In addition,\n",
"the front bumper was separate from the rest of the body. This is \n",
"all I know. If anyone can tellme a model name, engine specs, years\n",
"of production, where this car is made, history, or whatever info you\n",
"have on this funky looking car, please e-mail.\n",
"\n",
"Thanks,\n",
"- IL\n",
" ---- brought to you by your neighborhood Lerxst ----\n",
"\n",
"\n",
"\n",
"\n",
"\n"
]
}
],
"source": [
"print(newsgroups[0])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Naiwne przeszukiwanie"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"all_documents = list() \n",
"for document in newsgroups:\n",
" if 'car' in document:\n",
" all_documents.append(document)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"From: lerxst@wam.umd.edu (where's my thing)\n",
"Subject: WHAT car is this!?\n",
"Nntp-Posting-Host: rac3.wam.umd.edu\n",
"Organization: University of Maryland, College Park\n",
"Lines: 15\n",
"\n",
" I was wondering if anyone out there could enlighten me on this car I saw\n",
"the other day. It was a 2-door sports car, looked to be from the late 60s/\n",
"early 70s. It was called a Bricklin. The doors were really small. In addition,\n",
"the front bumper was separate from the rest of the body. This is \n",
"all I know. If anyone can tellme a model name, engine specs, years\n",
"of production, where this car is made, history, or whatever info you\n",
"have on this funky looking car, please e-mail.\n",
"\n",
"Thanks,\n",
"- IL\n",
" ---- brought to you by your neighborhood Lerxst ----\n",
"\n",
"\n",
"\n",
"\n",
"\n"
]
}
],
"source": [
"print(all_documents[0])"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"From: guykuo@carson.u.washington.edu (Guy Kuo)\n",
"Subject: SI Clock Poll - Final Call\n",
"Summary: Final call for SI clock reports\n",
"Keywords: SI,acceleration,clock,upgrade\n",
"Article-I.D.: shelley.1qvfo9INNc3s\n",
"Organization: University of Washington\n",
"Lines: 11\n",
"NNTP-Posting-Host: carson.u.washington.edu\n",
"\n",
"A fair number of brave souls who upgraded their SI clock oscillator have\n",
"shared their experiences for this poll. Please send a brief message detailing\n",
"your experiences with the procedure. Top speed attained, CPU rated speed,\n",
"add on cards and adapters, heat sinks, hour of usage per day, floppy disk\n",
"functionality with 800 and 1.4 m floppies are especially requested.\n",
"\n",
"I will be summarizing in the next two days, so please add to the network\n",
"knowledge base if you have done the clock upgrade and haven't answered this\n",
"poll. Thanks.\n",
"\n",
"Guy Kuo <guykuo@u.washington.edu>\n",
"\n"
]
}
],
"source": [
"print(all_documents[1])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### jakie są problemy z takim podejściem?\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## TFIDF i odległość cosinusowa- gotowe biblioteki"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"vectorizer = TfidfVectorizer()\n",
"#vectorizer = TfidfVectorizer(use_idf = False, ngram_range=(1,2))"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"document_vectors = vectorizer.fit_transform(newsgroups)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<11314x130107 sparse matrix of type '<class 'numpy.float64'>'\n",
"\twith 1787565 stored elements in Compressed Sparse Row format>"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"document_vectors"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<1x130107 sparse matrix of type '<class 'numpy.float64'>'\n",
"\twith 89 stored elements in Compressed Sparse Row format>"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"document_vectors[0]"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"matrix([[0., 0., 0., ..., 0., 0., 0.]])"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"document_vectors[0].todense()"
]
},
{
"cell_type": "code",
"execution_count": 13,
"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",
" [0., 0., 0., ..., 0., 0., 0.]])"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"document_vectors[0:4].todense()"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [],
"source": [
"query_str = 'speed'\n",
"#query_str = 'speed car'\n",
"#query_str = 'spider man'"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [],
"source": [
"query_vector = vectorizer.transform([query_str])"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<11314x130107 sparse matrix of type '<class 'numpy.float64'>'\n",
"\twith 1787565 stored elements in Compressed Sparse Row format>"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"document_vectors"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<1x130107 sparse matrix of type '<class 'numpy.float64'>'\n",
"\twith 1 stored elements in Compressed Sparse Row format>"
]
},
"execution_count": 17,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"query_vector"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [],
"source": [
"similarities = sklearn.metrics.pairwise.cosine_similarity(query_vector,document_vectors)"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([0.26949927, 0.3491801 , 0.44292083, 0.47784165])"
]
},
"execution_count": 19,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"np.sort(similarities)[0][-4:]"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([4517, 5509, 2116, 9921])"
]
},
"execution_count": 20,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"similarities.argsort()[0][-4:]"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"scrolled": true
},
"outputs": [],
"source": [
"for i in range (1,5):\n",
" print(newsgroups[similarities.argsort()[0][-i]])\n",
" print(np.sort(similarities)[0,-i])\n",
" print('-'*100)\n",
" print('-'*100)\n",
" print('-'*100)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Zadanie domowe\n",
"\n",
"\n",
"- Wybrać zbiór tekstowy, który ma conajmniej 10000 dokumentów (inny niż w tym przykładzie).\n",
"- Na jego podstawie stworzyć wyszukiwarkę bazującą na OKAPI BM25, tzn. system który dla podanej frazy podaje kilka (5-10) posortowanych najbardziej pasujących dokumentów razem ze scorami. Należy wypisywać też ilość zwracanych dokumentów, czyli takich z niezerowym scorem. Można korzystać z gotowych bibliotek do wektoryzacji dokumentów, należy jednak samemu zaimplementować OKAPI BM25. \n",
"- Znaleźć frazę (query), dla której wynik nie jest satysfakcjonujący.\n",
"- Poprawić wyszukiwarkę (np. poprzez zmianę preprocessingu tekstu, wektoryzer, zmianę parametrów algorytmu rankującego lub sam algorytm) tak, żeby zwracała satysfakcjonujące wyniki dla poprzedniej frazy. Należy zrobić inną zmianę niż w tym przykładzie, tylko wymyślić coś własnego.\n",
"- prezentować pracę na zajęciach (06.04) odpowiadając na pytania:\n",
" - jak wygląda zbiór i system wyszukiwania przed zmianami\n",
" - dla jakiej frazy wyniki są niesatysfakcjonujące (pokazać wyniki)\n",
" - jakie zmiany zostały naniesione\n",
" - jak wyglądają wyniki wyszukiwania po zmianach\n",
" - jak zmiany wpłynęły na wyniki (1-2 zdania)\n",
" \n",
"Prezentacja powinna być maksymalnie prosta i trwać maksymalnie 2-3 minuty.\n",
"punktów do zdobycia: 70\n"
]
}
],
"metadata": {
"author": "Jakub Pokrywka",
"email": "kubapok@wmi.amu.edu.pl",
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"lang": "pl",
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.8.3"
},
"subtitle": "3.tfidf (2)[ćwiczenia]",
"title": "Ekstrakcja informacji",
"year": "2021"
},
"nbformat": 4,
"nbformat_minor": 4
}