Spaces:
Sleeping
Sleeping
logreg and toxic bert
Browse files- Hello.py +30 -0
- images/pipeline_logreg.png +0 -0
- images/toxity_metrics.png +0 -0
- models/model1/logistic_regression_pipeline.pkl +3 -0
- models/model1/model_weights.pth +3 -0
- models/sds +0 -0
- notebooks/first_ml.ipynb +1539 -0
- pages/policlinic.py +15 -0
Hello.py
ADDED
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
|
3 |
+
|
4 |
+
st.set_page_config(
|
5 |
+
page_title="Hello",
|
6 |
+
page_icon="👋",
|
7 |
+
)
|
8 |
+
|
9 |
+
st.write("# Добро пожаловать на страничку нашего проекта! 👋")
|
10 |
+
|
11 |
+
st.sidebar.success("Выберите интересующую вас задачу.")
|
12 |
+
|
13 |
+
st.markdown(
|
14 |
+
"""
|
15 |
+
|
16 |
+
**👈 Выберите интересующую вас задачу и наши модели постараются вам помочь!**
|
17 |
+
### Что можно найти в этом сервисе?
|
18 |
+
- Страницу, позволяющую выполнить классификацию отзыва на поликлиники (при помощи трех различных моделей)
|
19 |
+
- Страницу, позволяющую выполнить оценку степени токсичности пользовательского сообщения с помощью модели rubert-tiny-toxicity
|
20 |
+
- Страницу, позволяющую выполнить генерацию текста GPT-моделью по пользовательскому prompt
|
21 |
+
- Страницу с информацией о:
|
22 |
+
- - процессе обучения модели: кривые обучения и метрик
|
23 |
+
- - времени обучения
|
24 |
+
- - значениях метрик
|
25 |
+
|
26 |
+
### Над проектом трудились:
|
27 |
+
- [Даша](https://github.com/Dasha0203)
|
28 |
+
- [Вера](https://github.com/VerVelVel)
|
29 |
+
"""
|
30 |
+
)
|
images/pipeline_logreg.png
ADDED
images/toxity_metrics.png
ADDED
models/model1/logistic_regression_pipeline.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e522e0db3ea799a291336149ab421d2ec56a6ea03e402bd438bec16b92a49dfb
|
3 |
+
size 5705593
|
models/model1/model_weights.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:da7fd2151d6a5446fc178462ff93ee61c24f98cb0aa41343e2e8c36802e2170b
|
3 |
+
size 47712485
|
models/sds
ADDED
File without changes
|
notebooks/first_ml.ipynb
ADDED
@@ -0,0 +1,1539 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cells": [
|
3 |
+
{
|
4 |
+
"cell_type": "markdown",
|
5 |
+
"metadata": {},
|
6 |
+
"source": [
|
7 |
+
"# TF-IDF"
|
8 |
+
]
|
9 |
+
},
|
10 |
+
{
|
11 |
+
"cell_type": "code",
|
12 |
+
"execution_count": 50,
|
13 |
+
"metadata": {},
|
14 |
+
"outputs": [],
|
15 |
+
"source": [
|
16 |
+
"import numpy as np\n",
|
17 |
+
"import pandas as pd\n",
|
18 |
+
"import re\n",
|
19 |
+
"import string\n",
|
20 |
+
"from collections import defaultdict\n",
|
21 |
+
"from sklearn import metrics\n",
|
22 |
+
"from time import time\n",
|
23 |
+
"from nltk.corpus import stopwords\n",
|
24 |
+
"from nltk.stem import WordNetLemmatizer\n",
|
25 |
+
"from nltk.tokenize import RegexpTokenizer\n",
|
26 |
+
"from sklearn.feature_extraction.text import TfidfVectorizer\n",
|
27 |
+
"from sklearn.cluster import KMeans\n",
|
28 |
+
"from sklearn.datasets import fetch_20newsgroups\n",
|
29 |
+
"from sklearn.decomposition import TruncatedSVD\n",
|
30 |
+
"from sklearn.pipeline import make_pipeline\n",
|
31 |
+
"from sklearn.preprocessing import Normalizer\n",
|
32 |
+
"import pymorphy2\n",
|
33 |
+
"from sklearn.model_selection import train_test_split\n",
|
34 |
+
"from sklearn.linear_model import LogisticRegression\n",
|
35 |
+
"from sklearn.metrics import classification_report, accuracy_score, f1_score"
|
36 |
+
]
|
37 |
+
},
|
38 |
+
{
|
39 |
+
"cell_type": "markdown",
|
40 |
+
"metadata": {},
|
41 |
+
"source": [
|
42 |
+
"## Загрузка данных"
|
43 |
+
]
|
44 |
+
},
|
45 |
+
{
|
46 |
+
"cell_type": "code",
|
47 |
+
"execution_count": 7,
|
48 |
+
"metadata": {},
|
49 |
+
"outputs": [],
|
50 |
+
"source": [
|
51 |
+
"df = pd.read_json('data/healthcare_facilities_reviews.jsonl', lines=True)"
|
52 |
+
]
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"cell_type": "code",
|
56 |
+
"execution_count": 8,
|
57 |
+
"metadata": {},
|
58 |
+
"outputs": [
|
59 |
+
{
|
60 |
+
"data": {
|
61 |
+
"text/html": [
|
62 |
+
"<div>\n",
|
63 |
+
"<style scoped>\n",
|
64 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
65 |
+
" vertical-align: middle;\n",
|
66 |
+
" }\n",
|
67 |
+
"\n",
|
68 |
+
" .dataframe tbody tr th {\n",
|
69 |
+
" vertical-align: top;\n",
|
70 |
+
" }\n",
|
71 |
+
"\n",
|
72 |
+
" .dataframe thead th {\n",
|
73 |
+
" text-align: right;\n",
|
74 |
+
" }\n",
|
75 |
+
"</style>\n",
|
76 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
77 |
+
" <thead>\n",
|
78 |
+
" <tr style=\"text-align: right;\">\n",
|
79 |
+
" <th></th>\n",
|
80 |
+
" <th>review_id</th>\n",
|
81 |
+
" <th>category</th>\n",
|
82 |
+
" <th>title</th>\n",
|
83 |
+
" <th>content</th>\n",
|
84 |
+
" <th>sentiment</th>\n",
|
85 |
+
" <th>source_url</th>\n",
|
86 |
+
" </tr>\n",
|
87 |
+
" </thead>\n",
|
88 |
+
" <tbody>\n",
|
89 |
+
" <tr>\n",
|
90 |
+
" <th>0</th>\n",
|
91 |
+
" <td>0</td>\n",
|
92 |
+
" <td>Поликлиники стоматологические</td>\n",
|
93 |
+
" <td>Классный мастер</td>\n",
|
94 |
+
" <td>Огромное спасибо за чудесное удаление двух зуб...</td>\n",
|
95 |
+
" <td>positive</td>\n",
|
96 |
+
" <td>http://www.spr.ru/forum_vyvod.php?id_tema=2727539</td>\n",
|
97 |
+
" </tr>\n",
|
98 |
+
" <tr>\n",
|
99 |
+
" <th>1</th>\n",
|
100 |
+
" <td>1</td>\n",
|
101 |
+
" <td>Поликлиники стоматологические</td>\n",
|
102 |
+
" <td>Замечательный врач</td>\n",
|
103 |
+
" <td>Хочу выразить особую благодарность замечательн...</td>\n",
|
104 |
+
" <td>positive</td>\n",
|
105 |
+
" <td>http://www.spr.ru/forum_vyvod.php?id_tema=2302877</td>\n",
|
106 |
+
" </tr>\n",
|
107 |
+
" <tr>\n",
|
108 |
+
" <th>2</th>\n",
|
109 |
+
" <td>2</td>\n",
|
110 |
+
" <td>Поликлиники стоматологические</td>\n",
|
111 |
+
" <td>Благодарность работникам рентгена</td>\n",
|
112 |
+
" <td>Добрый вечер! Хотелось бы поблагодарить сотруд...</td>\n",
|
113 |
+
" <td>positive</td>\n",
|
114 |
+
" <td>http://www.spr.ru/forum_vyvod.php?id_tema=2815031</td>\n",
|
115 |
+
" </tr>\n",
|
116 |
+
" <tr>\n",
|
117 |
+
" <th>3</th>\n",
|
118 |
+
" <td>3</td>\n",
|
119 |
+
" <td>Поликлиники стоматологические</td>\n",
|
120 |
+
" <td>Доктор Рабинович</td>\n",
|
121 |
+
" <td>Женщины советского образца в регистратуре не и...</td>\n",
|
122 |
+
" <td>negative</td>\n",
|
123 |
+
" <td>http://www.spr.ru/forum_vyvod.php?id_tema=3443161</td>\n",
|
124 |
+
" </tr>\n",
|
125 |
+
" <tr>\n",
|
126 |
+
" <th>4</th>\n",
|
127 |
+
" <td>4</td>\n",
|
128 |
+
" <td>Поликлиники стоматологические</td>\n",
|
129 |
+
" <td>Есть кому сказать спасибо</td>\n",
|
130 |
+
" <td>У меня с детства очень плохие зубы (тонкая и х...</td>\n",
|
131 |
+
" <td>positive</td>\n",
|
132 |
+
" <td>http://www.spr.ru/forum_vyvod.php?id_tema=2592430</td>\n",
|
133 |
+
" </tr>\n",
|
134 |
+
" <tr>\n",
|
135 |
+
" <th>...</th>\n",
|
136 |
+
" <td>...</td>\n",
|
137 |
+
" <td>...</td>\n",
|
138 |
+
" <td>...</td>\n",
|
139 |
+
" <td>...</td>\n",
|
140 |
+
" <td>...</td>\n",
|
141 |
+
" <td>...</td>\n",
|
142 |
+
" </tr>\n",
|
143 |
+
" <tr>\n",
|
144 |
+
" <th>70592</th>\n",
|
145 |
+
" <td>70592</td>\n",
|
146 |
+
" <td>Водительские комиссии</td>\n",
|
147 |
+
" <td>Хуже районной поликлиники</td>\n",
|
148 |
+
" <td>Заведение ужасное. Врачи делят 1 кабинет на 2х...</td>\n",
|
149 |
+
" <td>negative</td>\n",
|
150 |
+
" <td>http://www.spr.ru/forum_vyvod.php?id_tema=273326</td>\n",
|
151 |
+
" </tr>\n",
|
152 |
+
" <tr>\n",
|
153 |
+
" <th>70593</th>\n",
|
154 |
+
" <td>70593</td>\n",
|
155 |
+
" <td>Водительские комиссии</td>\n",
|
156 |
+
" <td>Справки</td>\n",
|
157 |
+
" <td>Люди, не обращайтесь в эту фирму! Муж проходил...</td>\n",
|
158 |
+
" <td>negative</td>\n",
|
159 |
+
" <td>http://www.spr.ru/forum_vyvod.php?id_tema=3401583</td>\n",
|
160 |
+
" </tr>\n",
|
161 |
+
" <tr>\n",
|
162 |
+
" <th>70594</th>\n",
|
163 |
+
" <td>70594</td>\n",
|
164 |
+
" <td>Водительские комиссии</td>\n",
|
165 |
+
" <td>Мед-Альфа - это наше будущее</td>\n",
|
166 |
+
" <td>Дорогие посетители медицинского центра ООО \"Ме...</td>\n",
|
167 |
+
" <td>positive</td>\n",
|
168 |
+
" <td>http://www.spr.ru/forum_vyvod.php?id_tema=326078</td>\n",
|
169 |
+
" </tr>\n",
|
170 |
+
" <tr>\n",
|
171 |
+
" <th>70595</th>\n",
|
172 |
+
" <td>70595</td>\n",
|
173 |
+
" <td>Водительские комиссии</td>\n",
|
174 |
+
" <td>Хамское поведение</td>\n",
|
175 |
+
" <td>В регистратуре сидит хамка, такое отношение и ...</td>\n",
|
176 |
+
" <td>negative</td>\n",
|
177 |
+
" <td>http://www.spr.ru/forum_vyvod.php?id_tema=3171911</td>\n",
|
178 |
+
" </tr>\n",
|
179 |
+
" <tr>\n",
|
180 |
+
" <th>70596</th>\n",
|
181 |
+
" <td>70596</td>\n",
|
182 |
+
" <td>Водительские комиссии</td>\n",
|
183 |
+
" <td>Только хорошие впечатления</td>\n",
|
184 |
+
" <td>Хочу поблагодарить весь персонал \"МедАльфаПроф...</td>\n",
|
185 |
+
" <td>positive</td>\n",
|
186 |
+
" <td>http://www.spr.ru/forum_vyvod.php?id_tema=3391562</td>\n",
|
187 |
+
" </tr>\n",
|
188 |
+
" </tbody>\n",
|
189 |
+
"</table>\n",
|
190 |
+
"<p>70597 rows × 6 columns</p>\n",
|
191 |
+
"</div>"
|
192 |
+
],
|
193 |
+
"text/plain": [
|
194 |
+
" review_id category \\\n",
|
195 |
+
"0 0 Поликлиники стоматологические \n",
|
196 |
+
"1 1 Поликлиники стоматологические \n",
|
197 |
+
"2 2 Поликлиники стоматологические \n",
|
198 |
+
"3 3 Поликлиники стоматологические \n",
|
199 |
+
"4 4 Поликлиники стоматологические \n",
|
200 |
+
"... ... ... \n",
|
201 |
+
"70592 70592 Водительские комиссии \n",
|
202 |
+
"70593 70593 Водительские комиссии \n",
|
203 |
+
"70594 70594 Водительские комиссии \n",
|
204 |
+
"70595 70595 Водительские комиссии \n",
|
205 |
+
"70596 70596 Водительские комиссии \n",
|
206 |
+
"\n",
|
207 |
+
" title \\\n",
|
208 |
+
"0 Классный мастер \n",
|
209 |
+
"1 Замечательный врач \n",
|
210 |
+
"2 Благодарность работникам рентгена \n",
|
211 |
+
"3 Доктор Рабинович \n",
|
212 |
+
"4 Есть кому сказать спасибо \n",
|
213 |
+
"... ... \n",
|
214 |
+
"70592 Хуже районной поликлиники \n",
|
215 |
+
"70593 Справки \n",
|
216 |
+
"70594 Мед-Альфа - это наше будущее \n",
|
217 |
+
"70595 Хамское поведение \n",
|
218 |
+
"70596 Только хорошие впечатления \n",
|
219 |
+
"\n",
|
220 |
+
" content sentiment \\\n",
|
221 |
+
"0 Огромное спасибо за чудесное удаление двух зуб... positive \n",
|
222 |
+
"1 Хочу выразить особую благодарность замечательн... positive \n",
|
223 |
+
"2 Добрый вечер! Хотелось бы поблагодарить сотруд... positive \n",
|
224 |
+
"3 Женщины советского об��азца в регистратуре не и... negative \n",
|
225 |
+
"4 У меня с детства очень плохие зубы (тонкая и х... positive \n",
|
226 |
+
"... ... ... \n",
|
227 |
+
"70592 Заведение ужасное. Врачи делят 1 кабинет на 2х... negative \n",
|
228 |
+
"70593 Люди, не обращайтесь в эту фирму! Муж проходил... negative \n",
|
229 |
+
"70594 Дорогие посетители медицинского центра ООО \"Ме... positive \n",
|
230 |
+
"70595 В регистратуре сидит хамка, такое отношение и ... negative \n",
|
231 |
+
"70596 Хочу поблагодарить весь персонал \"МедАльфаПроф... positive \n",
|
232 |
+
"\n",
|
233 |
+
" source_url \n",
|
234 |
+
"0 http://www.spr.ru/forum_vyvod.php?id_tema=2727539 \n",
|
235 |
+
"1 http://www.spr.ru/forum_vyvod.php?id_tema=2302877 \n",
|
236 |
+
"2 http://www.spr.ru/forum_vyvod.php?id_tema=2815031 \n",
|
237 |
+
"3 http://www.spr.ru/forum_vyvod.php?id_tema=3443161 \n",
|
238 |
+
"4 http://www.spr.ru/forum_vyvod.php?id_tema=2592430 \n",
|
239 |
+
"... ... \n",
|
240 |
+
"70592 http://www.spr.ru/forum_vyvod.php?id_tema=273326 \n",
|
241 |
+
"70593 http://www.spr.ru/forum_vyvod.php?id_tema=3401583 \n",
|
242 |
+
"70594 http://www.spr.ru/forum_vyvod.php?id_tema=326078 \n",
|
243 |
+
"70595 http://www.spr.ru/forum_vyvod.php?id_tema=3171911 \n",
|
244 |
+
"70596 http://www.spr.ru/forum_vyvod.php?id_tema=3391562 \n",
|
245 |
+
"\n",
|
246 |
+
"[70597 rows x 6 columns]"
|
247 |
+
]
|
248 |
+
},
|
249 |
+
"execution_count": 8,
|
250 |
+
"metadata": {},
|
251 |
+
"output_type": "execute_result"
|
252 |
+
}
|
253 |
+
],
|
254 |
+
"source": [
|
255 |
+
"df"
|
256 |
+
]
|
257 |
+
},
|
258 |
+
{
|
259 |
+
"cell_type": "code",
|
260 |
+
"execution_count": 9,
|
261 |
+
"metadata": {},
|
262 |
+
"outputs": [],
|
263 |
+
"source": [
|
264 |
+
"df = df[['sentiment', 'content']]"
|
265 |
+
]
|
266 |
+
},
|
267 |
+
{
|
268 |
+
"cell_type": "code",
|
269 |
+
"execution_count": 10,
|
270 |
+
"metadata": {},
|
271 |
+
"outputs": [
|
272 |
+
{
|
273 |
+
"data": {
|
274 |
+
"text/html": [
|
275 |
+
"<div>\n",
|
276 |
+
"<style scoped>\n",
|
277 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
278 |
+
" vertical-align: middle;\n",
|
279 |
+
" }\n",
|
280 |
+
"\n",
|
281 |
+
" .dataframe tbody tr th {\n",
|
282 |
+
" vertical-align: top;\n",
|
283 |
+
" }\n",
|
284 |
+
"\n",
|
285 |
+
" .dataframe thead th {\n",
|
286 |
+
" text-align: right;\n",
|
287 |
+
" }\n",
|
288 |
+
"</style>\n",
|
289 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
290 |
+
" <thead>\n",
|
291 |
+
" <tr style=\"text-align: right;\">\n",
|
292 |
+
" <th></th>\n",
|
293 |
+
" <th>sentiment</th>\n",
|
294 |
+
" <th>content</th>\n",
|
295 |
+
" </tr>\n",
|
296 |
+
" </thead>\n",
|
297 |
+
" <tbody>\n",
|
298 |
+
" <tr>\n",
|
299 |
+
" <th>0</th>\n",
|
300 |
+
" <td>positive</td>\n",
|
301 |
+
" <td>Огромное спасибо за чудесное удаление двух зуб...</td>\n",
|
302 |
+
" </tr>\n",
|
303 |
+
" <tr>\n",
|
304 |
+
" <th>1</th>\n",
|
305 |
+
" <td>positive</td>\n",
|
306 |
+
" <td>Хочу выразить особую благодарность замечательн...</td>\n",
|
307 |
+
" </tr>\n",
|
308 |
+
" <tr>\n",
|
309 |
+
" <th>2</th>\n",
|
310 |
+
" <td>positive</td>\n",
|
311 |
+
" <td>Добрый вечер! Хотелось бы поблагодарить сотруд...</td>\n",
|
312 |
+
" </tr>\n",
|
313 |
+
" <tr>\n",
|
314 |
+
" <th>3</th>\n",
|
315 |
+
" <td>negative</td>\n",
|
316 |
+
" <td>Женщины советского образца в регистратуре не и...</td>\n",
|
317 |
+
" </tr>\n",
|
318 |
+
" <tr>\n",
|
319 |
+
" <th>4</th>\n",
|
320 |
+
" <td>positive</td>\n",
|
321 |
+
" <td>У меня с детства очень плохие зубы (тонкая и х...</td>\n",
|
322 |
+
" </tr>\n",
|
323 |
+
" <tr>\n",
|
324 |
+
" <th>...</th>\n",
|
325 |
+
" <td>...</td>\n",
|
326 |
+
" <td>...</td>\n",
|
327 |
+
" </tr>\n",
|
328 |
+
" <tr>\n",
|
329 |
+
" <th>70592</th>\n",
|
330 |
+
" <td>negative</td>\n",
|
331 |
+
" <td>Заведение ужасное. Врачи делят 1 кабинет на 2х...</td>\n",
|
332 |
+
" </tr>\n",
|
333 |
+
" <tr>\n",
|
334 |
+
" <th>70593</th>\n",
|
335 |
+
" <td>negative</td>\n",
|
336 |
+
" <td>Люди, не обращайтесь в эту фирму! Муж проходил...</td>\n",
|
337 |
+
" </tr>\n",
|
338 |
+
" <tr>\n",
|
339 |
+
" <th>70594</th>\n",
|
340 |
+
" <td>positive</td>\n",
|
341 |
+
" <td>Дорогие посетители медицинского центра ООО \"Ме...</td>\n",
|
342 |
+
" </tr>\n",
|
343 |
+
" <tr>\n",
|
344 |
+
" <th>70595</th>\n",
|
345 |
+
" <td>negative</td>\n",
|
346 |
+
" <td>В регистратуре сидит ��амка, такое отношение и ...</td>\n",
|
347 |
+
" </tr>\n",
|
348 |
+
" <tr>\n",
|
349 |
+
" <th>70596</th>\n",
|
350 |
+
" <td>positive</td>\n",
|
351 |
+
" <td>Хочу поблагодарить весь персонал \"МедАльфаПроф...</td>\n",
|
352 |
+
" </tr>\n",
|
353 |
+
" </tbody>\n",
|
354 |
+
"</table>\n",
|
355 |
+
"<p>70597 rows × 2 columns</p>\n",
|
356 |
+
"</div>"
|
357 |
+
],
|
358 |
+
"text/plain": [
|
359 |
+
" sentiment content\n",
|
360 |
+
"0 positive Огромное спасибо за чудесное удаление двух зуб...\n",
|
361 |
+
"1 positive Хочу выразить особую благодарность замечательн...\n",
|
362 |
+
"2 positive Добрый вечер! Хотелось бы поблагодарить сотруд...\n",
|
363 |
+
"3 negative Женщины советского образца в регистратуре не и...\n",
|
364 |
+
"4 positive У меня с детства очень плохие зубы (тонкая и х...\n",
|
365 |
+
"... ... ...\n",
|
366 |
+
"70592 negative Заведение ужасное. Врачи делят 1 кабинет на 2х...\n",
|
367 |
+
"70593 negative Люди, не обращайтесь в эту фирму! Муж проходил...\n",
|
368 |
+
"70594 positive Дорогие посетители медицинского центра ООО \"Ме...\n",
|
369 |
+
"70595 negative В регистратуре сидит хамка, такое отношение и ...\n",
|
370 |
+
"70596 positive Хочу поблагодарить весь персонал \"МедАльфаПроф...\n",
|
371 |
+
"\n",
|
372 |
+
"[70597 rows x 2 columns]"
|
373 |
+
]
|
374 |
+
},
|
375 |
+
"execution_count": 10,
|
376 |
+
"metadata": {},
|
377 |
+
"output_type": "execute_result"
|
378 |
+
}
|
379 |
+
],
|
380 |
+
"source": [
|
381 |
+
"df"
|
382 |
+
]
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"cell_type": "code",
|
386 |
+
"execution_count": 12,
|
387 |
+
"metadata": {},
|
388 |
+
"outputs": [
|
389 |
+
{
|
390 |
+
"data": {
|
391 |
+
"text/plain": [
|
392 |
+
"'Добрый вечер! Хотелось бы поблагодарить сотрудников рентгена! Протезируюсь, отношусь к поликлинике № 189. Там меня отфутболили! Подходила к Кочину, зам. гл. врачу, заведующей просто сделать 3 снимка (пол-ка рядом с домом)- мне грубо отказали! А сотрудник рентгена просто сидела кроссворд разгадывала! Они видите ли, не принимают с протезирования! Сказали, где протезируетесь, там и делайте, а я говорю, мне у Вас удобно. Побоялись они! Первый раз попала к молодой девушке, она меня выслушала и сделала 1 снимок, а потом записала на другие дни, мне это удобно. Конечно, народу полно было! Бедные сотрудники. Все, кто читает отзыв (особенно жители Люблино 189 пол-ки), давайте жаловаться в департамент! Спасибо еще раз, за рентген (слышала в очереди, что народу у Вас было много и вы уже перебрали с нормой). Спасибо.'"
|
393 |
+
]
|
394 |
+
},
|
395 |
+
"execution_count": 12,
|
396 |
+
"metadata": {},
|
397 |
+
"output_type": "execute_result"
|
398 |
+
}
|
399 |
+
],
|
400 |
+
"source": [
|
401 |
+
"df['content'][2]"
|
402 |
+
]
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"cell_type": "markdown",
|
406 |
+
"metadata": {},
|
407 |
+
"source": [
|
408 |
+
"## Очистка текста"
|
409 |
+
]
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"cell_type": "code",
|
413 |
+
"execution_count": 25,
|
414 |
+
"metadata": {},
|
415 |
+
"outputs": [],
|
416 |
+
"source": [
|
417 |
+
"morph = pymorphy2.MorphAnalyzer()\n",
|
418 |
+
"russian_stopwords = stopwords.words(\"russian\")"
|
419 |
+
]
|
420 |
+
},
|
421 |
+
{
|
422 |
+
"cell_type": "code",
|
423 |
+
"execution_count": 26,
|
424 |
+
"metadata": {},
|
425 |
+
"outputs": [
|
426 |
+
{
|
427 |
+
"name": "stderr",
|
428 |
+
"output_type": "stream",
|
429 |
+
"text": [
|
430 |
+
"/tmp/ipykernel_75887/650983554.py:12: SettingWithCopyWarning: \n",
|
431 |
+
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
|
432 |
+
"Try using .loc[row_indexer,col_indexer] = value instead\n",
|
433 |
+
"\n",
|
434 |
+
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
|
435 |
+
" df['cleaned_text'] = df['content'].apply(clean_text)\n"
|
436 |
+
]
|
437 |
+
}
|
438 |
+
],
|
439 |
+
"source": [
|
440 |
+
"def clean_text(text):\n",
|
441 |
+
" # Удаление всего, что не является буквами или знаками препинания\n",
|
442 |
+
" clean_pattern = re.compile(r'[^a-zA-Zа-яА-ЯёЁ0-9.,!?;:\\s]')\n",
|
443 |
+
" text = clean_pattern.sub('', text)\n",
|
444 |
+
" url_pattern = re.compile(r'http\\S+|www\\S+|https\\S+')\n",
|
445 |
+
" text = url_pattern.sub(r'', text)\n",
|
446 |
+
" text = text.translate(str.maketrans('', '', string.punctuation))\n",
|
447 |
+
" text = text.lower()\n",
|
448 |
+
" lemmatized_text = ' '.join([morph.parse(word)[0].normal_form for word in text.split() if word not in russian_stopwords])\n",
|
449 |
+
" return lemmatized_text\n",
|
450 |
+
"\n",
|
451 |
+
"df['cleaned_text'] = df['content'].apply(clean_text)"
|
452 |
+
]
|
453 |
+
},
|
454 |
+
{
|
455 |
+
"cell_type": "code",
|
456 |
+
"execution_count": 28,
|
457 |
+
"metadata": {},
|
458 |
+
"outputs": [
|
459 |
+
{
|
460 |
+
"data": {
|
461 |
+
"text/html": [
|
462 |
+
"<div>\n",
|
463 |
+
"<style scoped>\n",
|
464 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
465 |
+
" vertical-align: middle;\n",
|
466 |
+
" }\n",
|
467 |
+
"\n",
|
468 |
+
" .dataframe tbody tr th {\n",
|
469 |
+
" vertical-align: top;\n",
|
470 |
+
" }\n",
|
471 |
+
"\n",
|
472 |
+
" .dataframe thead th {\n",
|
473 |
+
" text-align: right;\n",
|
474 |
+
" }\n",
|
475 |
+
"</style>\n",
|
476 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
477 |
+
" <thead>\n",
|
478 |
+
" <tr style=\"text-align: right;\">\n",
|
479 |
+
" <th></th>\n",
|
480 |
+
" <th>sentiment</th>\n",
|
481 |
+
" <th>content</th>\n",
|
482 |
+
" <th>cleaned_text</th>\n",
|
483 |
+
" </tr>\n",
|
484 |
+
" </thead>\n",
|
485 |
+
" <tbody>\n",
|
486 |
+
" <tr>\n",
|
487 |
+
" <th>0</th>\n",
|
488 |
+
" <td>positive</td>\n",
|
489 |
+
" <td>Огромное спасибо за чудесное удаление двух зуб...</td>\n",
|
490 |
+
" <td>огромный спасибо чудесный удаление два зуб муд...</td>\n",
|
491 |
+
" </tr>\n",
|
492 |
+
" <tr>\n",
|
493 |
+
" <th>1</th>\n",
|
494 |
+
" <td>positive</td>\n",
|
495 |
+
" <td>Хочу выразить особую благодарность замечательн...</td>\n",
|
496 |
+
" <td>хотеть выразить особый благодарность замечател...</td>\n",
|
497 |
+
" </tr>\n",
|
498 |
+
" <tr>\n",
|
499 |
+
" <th>2</th>\n",
|
500 |
+
" <td>positive</td>\n",
|
501 |
+
" <td>Добрый вечер! Хотелось бы поблагодарить сотруд...</td>\n",
|
502 |
+
" <td>добрый вечер хотеться поблагодарить сотрудник ...</td>\n",
|
503 |
+
" </tr>\n",
|
504 |
+
" <tr>\n",
|
505 |
+
" <th>3</th>\n",
|
506 |
+
" <td>negative</td>\n",
|
507 |
+
" <td>Женщины советского образца в регистратуре не и...</td>\n",
|
508 |
+
" <td>женщина советский образец регистратура иметь п...</td>\n",
|
509 |
+
" </tr>\n",
|
510 |
+
" <tr>\n",
|
511 |
+
" <th>4</th>\n",
|
512 |
+
" <td>positive</td>\n",
|
513 |
+
" <td>У меня с детства очень плохие зубы (тонкая и х...</td>\n",
|
514 |
+
" <td>детство очень плохой зуб тонкий хрупкий эмаль ...</td>\n",
|
515 |
+
" </tr>\n",
|
516 |
+
" <tr>\n",
|
517 |
+
" <th>...</th>\n",
|
518 |
+
" <td>...</td>\n",
|
519 |
+
" <td>...</td>\n",
|
520 |
+
" <td>...</td>\n",
|
521 |
+
" </tr>\n",
|
522 |
+
" <tr>\n",
|
523 |
+
" <th>70592</th>\n",
|
524 |
+
" <td>negative</td>\n",
|
525 |
+
" <td>Заведение ужасное. Врачи делят 1 кабинет на 2х...</td>\n",
|
526 |
+
" <td>заведение ужасный врач делить 1 кабинет 2х спе...</td>\n",
|
527 |
+
" </tr>\n",
|
528 |
+
" <tr>\n",
|
529 |
+
" <th>70593</th>\n",
|
530 |
+
" <td>negative</td>\n",
|
531 |
+
" <td>Люди, не обращайтесь в эту фирму! Муж проходил...</td>\n",
|
532 |
+
" <td>человек обращаться фирма муж проходить анализ ...</td>\n",
|
533 |
+
" </tr>\n",
|
534 |
+
" <tr>\n",
|
535 |
+
" <th>70594</th>\n",
|
536 |
+
" <td>positive</td>\n",
|
537 |
+
" <td>Дорогие посетители медицинского центра ООО \"Ме...</td>\n",
|
538 |
+
" <td>дорогой посетитель медицинский центр ооо медал...</td>\n",
|
539 |
+
" </tr>\n",
|
540 |
+
" <tr>\n",
|
541 |
+
" <th>70595</th>\n",
|
542 |
+
" <td>negative</td>\n",
|
543 |
+
" <td>В регистратуре сидит хамка, такое отношение и ...</td>\n",
|
544 |
+
" <td>регистратура сидеть хамка такой отношение мане...</td>\n",
|
545 |
+
" </tr>\n",
|
546 |
+
" <tr>\n",
|
547 |
+
" <th>70596</th>\n",
|
548 |
+
" <td>positive</td>\n",
|
549 |
+
" <td>Хочу поблагодарить весь персонал \"МедАльфаПроф...</td>\n",
|
550 |
+
" <td>хотеть поблагодарить весь персонал медальфапро...</td>\n",
|
551 |
+
" </tr>\n",
|
552 |
+
" </tbody>\n",
|
553 |
+
"</table>\n",
|
554 |
+
"<p>70597 rows × 3 columns</p>\n",
|
555 |
+
"</div>"
|
556 |
+
],
|
557 |
+
"text/plain": [
|
558 |
+
" sentiment content \\\n",
|
559 |
+
"0 positive Огромное спасибо за чудесное удаление двух зуб... \n",
|
560 |
+
"1 positive Хочу выразить особую благодарность замечательн... \n",
|
561 |
+
"2 positive Добрый вечер! Хотелось бы поблагодарить сотруд... \n",
|
562 |
+
"3 negative Женщины советского образца в регистратуре не и... \n",
|
563 |
+
"4 positive У меня с детства очень плохие зубы (тонкая и х... \n",
|
564 |
+
"... ... ... \n",
|
565 |
+
"70592 negative Заведение ужасное. Врачи делят 1 кабинет на 2х... \n",
|
566 |
+
"70593 negative Люди, не обращайтесь в эту фирму! Муж проходил... \n",
|
567 |
+
"70594 positive Дорогие посетители медицинского центра ООО \"Ме... \n",
|
568 |
+
"70595 negative В регистратуре сидит хамка, такое отношение и ... \n",
|
569 |
+
"70596 positive Хочу поблагодарить весь персонал \"МедАльфаПроф... \n",
|
570 |
+
"\n",
|
571 |
+
" cleaned_text \n",
|
572 |
+
"0 огромный спасибо чудесный удаление два зуб муд... \n",
|
573 |
+
"1 хотеть выразить особый благодарность замечател... \n",
|
574 |
+
"2 добрый вечер хотеться поблагодарить сотрудник ... \n",
|
575 |
+
"3 женщина советский образец регистратура иметь п... \n",
|
576 |
+
"4 детство очень плохой зуб тонкий хрупкий эмаль ... \n",
|
577 |
+
"... ... \n",
|
578 |
+
"70592 заведение ужасный врач делить 1 кабинет 2х спе... \n",
|
579 |
+
"70593 человек обращаться фирма муж проходить анализ ... \n",
|
580 |
+
"70594 дорогой посетитель медицинский центр ооо медал... \n",
|
581 |
+
"70595 регистратура сидеть хамка такой отношение мане... \n",
|
582 |
+
"70596 хотеть поблагодарить весь персонал медальфапро... \n",
|
583 |
+
"\n",
|
584 |
+
"[70597 rows x 3 columns]"
|
585 |
+
]
|
586 |
+
},
|
587 |
+
"execution_count": 28,
|
588 |
+
"metadata": {},
|
589 |
+
"output_type": "execute_result"
|
590 |
+
}
|
591 |
+
],
|
592 |
+
"source": [
|
593 |
+
"df"
|
594 |
+
]
|
595 |
+
},
|
596 |
+
{
|
597 |
+
"cell_type": "code",
|
598 |
+
"execution_count": 29,
|
599 |
+
"metadata": {},
|
600 |
+
"outputs": [
|
601 |
+
{
|
602 |
+
"data": {
|
603 |
+
"text/plain": [
|
604 |
+
"'добрый вечер хотеться поблагодарить сотрудник рентген протезироваться относиться поликлиника 189 отфутболить подходить кочин зам гл врач заведовать просто сделать 3 снимок полка рядом дом грубо отказать сотрудник рентген просто сидеть кроссворд разгадывать видеть принимать протезирование сказать протезироваться делать говорить удобно побояться первый попасть молодой девушка выслушать сделать 1 снимка записать другой день это удобно народ полно бедный сотрудник читать отзыв особенно житель люблино 189 полка давать жаловаться департамент спасибо рентген слышать очередь народ перебрать норма спасибо'"
|
605 |
+
]
|
606 |
+
},
|
607 |
+
"execution_count": 29,
|
608 |
+
"metadata": {},
|
609 |
+
"output_type": "execute_result"
|
610 |
+
}
|
611 |
+
],
|
612 |
+
"source": [
|
613 |
+
"df['cleaned_text'][2]"
|
614 |
+
]
|
615 |
+
},
|
616 |
+
{
|
617 |
+
"cell_type": "code",
|
618 |
+
"execution_count": 34,
|
619 |
+
"metadata": {},
|
620 |
+
"outputs": [
|
621 |
+
{
|
622 |
+
"name": "stderr",
|
623 |
+
"output_type": "stream",
|
624 |
+
"text": [
|
625 |
+
"/tmp/ipykernel_75887/3526150694.py:1: SettingWithCopyWarning: \n",
|
626 |
+
"A value is trying to be set on a copy of a slice from a DataFrame.\n",
|
627 |
+
"Try using .loc[row_indexer,col_indexer] = value instead\n",
|
628 |
+
"\n",
|
629 |
+
"See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
|
630 |
+
" df['sentiment'] = df['sentiment'].apply(lambda x: 1 if x == 'negative' else 0)\n"
|
631 |
+
]
|
632 |
+
}
|
633 |
+
],
|
634 |
+
"source": [
|
635 |
+
"df['sentiment'] = df['sentiment'].apply(lambda x: 1 if x == 'negative' else 0)"
|
636 |
+
]
|
637 |
+
},
|
638 |
+
{
|
639 |
+
"cell_type": "code",
|
640 |
+
"execution_count": 35,
|
641 |
+
"metadata": {},
|
642 |
+
"outputs": [
|
643 |
+
{
|
644 |
+
"data": {
|
645 |
+
"text/html": [
|
646 |
+
"<div>\n",
|
647 |
+
"<style scoped>\n",
|
648 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
649 |
+
" vertical-align: middle;\n",
|
650 |
+
" }\n",
|
651 |
+
"\n",
|
652 |
+
" .dataframe tbody tr th {\n",
|
653 |
+
" vertical-align: top;\n",
|
654 |
+
" }\n",
|
655 |
+
"\n",
|
656 |
+
" .dataframe thead th {\n",
|
657 |
+
" text-align: right;\n",
|
658 |
+
" }\n",
|
659 |
+
"</style>\n",
|
660 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
661 |
+
" <thead>\n",
|
662 |
+
" <tr style=\"text-align: right;\">\n",
|
663 |
+
" <th></th>\n",
|
664 |
+
" <th>sentiment</th>\n",
|
665 |
+
" <th>content</th>\n",
|
666 |
+
" <th>cleaned_text</th>\n",
|
667 |
+
" </tr>\n",
|
668 |
+
" </thead>\n",
|
669 |
+
" <tbody>\n",
|
670 |
+
" <tr>\n",
|
671 |
+
" <th>0</th>\n",
|
672 |
+
" <td>0</td>\n",
|
673 |
+
" <td>Огромное спасибо за чудесное удаление двух зуб...</td>\n",
|
674 |
+
" <td>огромный спасибо чудесный удаление два зуб муд...</td>\n",
|
675 |
+
" </tr>\n",
|
676 |
+
" <tr>\n",
|
677 |
+
" <th>1</th>\n",
|
678 |
+
" <td>0</td>\n",
|
679 |
+
" <td>Хочу выразить особую благодарность замечательн...</td>\n",
|
680 |
+
" <td>хотеть выразить особый благодарность замечател...</td>\n",
|
681 |
+
" </tr>\n",
|
682 |
+
" <tr>\n",
|
683 |
+
" <th>2</th>\n",
|
684 |
+
" <td>0</td>\n",
|
685 |
+
" <td>Добрый вечер! Хотелось бы поблагодарить сотруд...</td>\n",
|
686 |
+
" <td>добрый вечер хотеться поблагодарить сотрудник ...</td>\n",
|
687 |
+
" </tr>\n",
|
688 |
+
" <tr>\n",
|
689 |
+
" <th>3</th>\n",
|
690 |
+
" <td>1</td>\n",
|
691 |
+
" <td>Женщины советского образца в регистратуре не и...</td>\n",
|
692 |
+
" <td>женщина советский образец регистратура иметь п...</td>\n",
|
693 |
+
" </tr>\n",
|
694 |
+
" <tr>\n",
|
695 |
+
" <th>4</th>\n",
|
696 |
+
" <td>0</td>\n",
|
697 |
+
" <td>У меня с детства очень плохие зубы (тонкая и х...</td>\n",
|
698 |
+
" <td>детство очень плохой зуб тонкий хрупкий эмаль ...</td>\n",
|
699 |
+
" </tr>\n",
|
700 |
+
" <tr>\n",
|
701 |
+
" <th>...</th>\n",
|
702 |
+
" <td>...</td>\n",
|
703 |
+
" <td>...</td>\n",
|
704 |
+
" <td>...</td>\n",
|
705 |
+
" </tr>\n",
|
706 |
+
" <tr>\n",
|
707 |
+
" <th>70592</th>\n",
|
708 |
+
" <td>1</td>\n",
|
709 |
+
" <td>Заведение ужасное. Врачи делят 1 кабинет на 2х...</td>\n",
|
710 |
+
" <td>заведение ужасный врач делить 1 кабинет 2х спе...</td>\n",
|
711 |
+
" </tr>\n",
|
712 |
+
" <tr>\n",
|
713 |
+
" <th>70593</th>\n",
|
714 |
+
" <td>1</td>\n",
|
715 |
+
" <td>Люди, не обращайтесь в эту фирму! Муж проходил...</td>\n",
|
716 |
+
" <td>человек обращаться фирма муж проходить анализ ...</td>\n",
|
717 |
+
" </tr>\n",
|
718 |
+
" <tr>\n",
|
719 |
+
" <th>70594</th>\n",
|
720 |
+
" <td>0</td>\n",
|
721 |
+
" <td>Дорогие посетители медицинского центра ООО \"Ме...</td>\n",
|
722 |
+
" <td>дорогой посетитель медицинский центр ооо медал...</td>\n",
|
723 |
+
" </tr>\n",
|
724 |
+
" <tr>\n",
|
725 |
+
" <th>70595</th>\n",
|
726 |
+
" <td>1</td>\n",
|
727 |
+
" <td>В регистратуре сидит хамка, такое отношение и ...</td>\n",
|
728 |
+
" <td>регистратура сидеть хамка такой отношение мане...</td>\n",
|
729 |
+
" </tr>\n",
|
730 |
+
" <tr>\n",
|
731 |
+
" <th>70596</th>\n",
|
732 |
+
" <td>0</td>\n",
|
733 |
+
" <td>Хочу поблагодарить весь персонал \"МедАльфаПроф...</td>\n",
|
734 |
+
" <td>хотеть поблагодарить весь персонал медальфапро...</td>\n",
|
735 |
+
" </tr>\n",
|
736 |
+
" </tbody>\n",
|
737 |
+
"</table>\n",
|
738 |
+
"<p>70597 rows × 3 columns</p>\n",
|
739 |
+
"</div>"
|
740 |
+
],
|
741 |
+
"text/plain": [
|
742 |
+
" sentiment content \\\n",
|
743 |
+
"0 0 Огромное спасибо за чудесное удаление двух зуб... \n",
|
744 |
+
"1 0 Хочу выразить особую благодарность замечательн... \n",
|
745 |
+
"2 0 Добрый вечер! Хотелось бы поблагодарить сотруд... \n",
|
746 |
+
"3 1 Женщины советского образца в регистратуре не и... \n",
|
747 |
+
"4 0 У меня с детства очень плохие зубы (тонкая и х... \n",
|
748 |
+
"... ... ... \n",
|
749 |
+
"70592 1 Заведение ужасное. Врачи делят 1 кабинет на 2х... \n",
|
750 |
+
"70593 1 Люди, не обращайтесь в эту фирму! Муж проходил... \n",
|
751 |
+
"70594 0 Дорогие посетители медицинского центра ООО \"Ме... \n",
|
752 |
+
"70595 1 В регистратуре сидит хамка, такое отношение и ... \n",
|
753 |
+
"70596 0 Хочу поблагодарить весь персонал \"МедАльфаПроф... \n",
|
754 |
+
"\n",
|
755 |
+
" cleaned_text \n",
|
756 |
+
"0 огромный спасибо чудесный удаление два зуб муд... \n",
|
757 |
+
"1 хотеть выразить особый благодарность замечател... \n",
|
758 |
+
"2 добрый вечер хотеться поблагодарить сотрудник ... \n",
|
759 |
+
"3 женщина советский образец регистратура иметь п... \n",
|
760 |
+
"4 детство очень плохой зуб тонкий хрупкий эмаль ... \n",
|
761 |
+
"... ... \n",
|
762 |
+
"70592 заведение ужасный врач делить 1 кабинет 2х спе... \n",
|
763 |
+
"70593 человек обращаться фирма муж проходить анализ ... \n",
|
764 |
+
"70594 дорогой посетитель медицинский центр ооо медал... \n",
|
765 |
+
"70595 регистратура сидеть хамка такой отношение мане... \n",
|
766 |
+
"70596 хотеть поблагодарить весь персонал медальфапро... \n",
|
767 |
+
"\n",
|
768 |
+
"[70597 rows x 3 columns]"
|
769 |
+
]
|
770 |
+
},
|
771 |
+
"execution_count": 35,
|
772 |
+
"metadata": {},
|
773 |
+
"output_type": "execute_result"
|
774 |
+
}
|
775 |
+
],
|
776 |
+
"source": [
|
777 |
+
"df"
|
778 |
+
]
|
779 |
+
},
|
780 |
+
{
|
781 |
+
"cell_type": "code",
|
782 |
+
"execution_count": 46,
|
783 |
+
"metadata": {},
|
784 |
+
"outputs": [],
|
785 |
+
"source": [
|
786 |
+
"X_train, X_test, y_train, y_test = train_test_split(df['cleaned_text'], df['sentiment'], test_size=0.2, random_state=42)"
|
787 |
+
]
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"cell_type": "markdown",
|
791 |
+
"metadata": {},
|
792 |
+
"source": [
|
793 |
+
"## Векторизация и сжатие"
|
794 |
+
]
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"cell_type": "code",
|
798 |
+
"execution_count": 47,
|
799 |
+
"metadata": {},
|
800 |
+
"outputs": [
|
801 |
+
{
|
802 |
+
"name": "stdout",
|
803 |
+
"output_type": "stream",
|
804 |
+
"text": [
|
805 |
+
"vectorization done in 4.084 s\n",
|
806 |
+
"n_samples train: 56477, n_features: 1010\n",
|
807 |
+
"n_samples test: 14120, n_features: 1010\n"
|
808 |
+
]
|
809 |
+
}
|
810 |
+
],
|
811 |
+
"source": [
|
812 |
+
"vectorizer = TfidfVectorizer(\n",
|
813 |
+
" max_df=0.9,\n",
|
814 |
+
" min_df=500,\n",
|
815 |
+
" # ngram_range=(1, 2), # Использование униграмм и биграмм\n",
|
816 |
+
" # max_features=5000,\n",
|
817 |
+
" stop_words=stopwords.words('russian'),\n",
|
818 |
+
")\n",
|
819 |
+
"t0 = time()\n",
|
820 |
+
"X_train_tfidf = vectorizer.fit_transform(X_train)\n",
|
821 |
+
"X_test_tfidf = vectorizer.transform(X_test)\n",
|
822 |
+
"\n",
|
823 |
+
"print(f\"vectorization done in {time() - t0:.3f} s\")\n",
|
824 |
+
"print(f\"n_samples train: {X_train_tfidf.shape[0]}, n_features: {X_train_tfidf.shape[1]}\")\n",
|
825 |
+
"print(f\"n_samples test: {X_test_tfidf.shape[0]}, n_features: {X_test_tfidf.shape[1]}\")"
|
826 |
+
]
|
827 |
+
},
|
828 |
+
{
|
829 |
+
"cell_type": "code",
|
830 |
+
"execution_count": 48,
|
831 |
+
"metadata": {},
|
832 |
+
"outputs": [
|
833 |
+
{
|
834 |
+
"name": "stdout",
|
835 |
+
"output_type": "stream",
|
836 |
+
"text": [
|
837 |
+
"LSA done in 14.485 s\n",
|
838 |
+
"Explained variance of the SVD step: 74.3%\n"
|
839 |
+
]
|
840 |
+
}
|
841 |
+
],
|
842 |
+
"source": [
|
843 |
+
"lsa = make_pipeline(TruncatedSVD(n_components=500), Normalizer(copy=False))\n",
|
844 |
+
"t0 = time()\n",
|
845 |
+
"X_train_lsa = lsa.fit_transform(X_train_tfidf)\n",
|
846 |
+
"\n",
|
847 |
+
"# Применение обученной модели LSA к тестовым данным\n",
|
848 |
+
"X_test_lsa = lsa.transform(X_test_tfidf)\n",
|
849 |
+
"explained_variance = lsa[0].explained_variance_ratio_.sum()\n",
|
850 |
+
"\n",
|
851 |
+
"print(f\"LSA done in {time() - t0:.3f} s\")\n",
|
852 |
+
"print(f\"Explained variance of the SVD step: {explained_variance * 100:.1f}%\")"
|
853 |
+
]
|
854 |
+
},
|
855 |
+
{
|
856 |
+
"cell_type": "markdown",
|
857 |
+
"metadata": {},
|
858 |
+
"source": [
|
859 |
+
"## Логистическая регрессия"
|
860 |
+
]
|
861 |
+
},
|
862 |
+
{
|
863 |
+
"cell_type": "code",
|
864 |
+
"execution_count": 51,
|
865 |
+
"metadata": {},
|
866 |
+
"outputs": [
|
867 |
+
{
|
868 |
+
"name": "stdout",
|
869 |
+
"output_type": "stream",
|
870 |
+
"text": [
|
871 |
+
" precision recall f1-score support\n",
|
872 |
+
"\n",
|
873 |
+
" 0 0.94 0.94 0.94 8342\n",
|
874 |
+
" 1 0.91 0.92 0.91 5778\n",
|
875 |
+
"\n",
|
876 |
+
" accuracy 0.93 14120\n",
|
877 |
+
" macro avg 0.92 0.93 0.93 14120\n",
|
878 |
+
"weighted avg 0.93 0.93 0.93 14120\n",
|
879 |
+
"\n",
|
880 |
+
"Accuracy: 0.9277620396600567\n",
|
881 |
+
"F1 score: 0.9120689655172414\n"
|
882 |
+
]
|
883 |
+
}
|
884 |
+
],
|
885 |
+
"source": [
|
886 |
+
"model = LogisticRegression()\n",
|
887 |
+
"\n",
|
888 |
+
"# Обучение модели\n",
|
889 |
+
"model.fit(X_train_lsa, y_train)\n",
|
890 |
+
"\n",
|
891 |
+
"# Прогнозирование на тестовой выборке\n",
|
892 |
+
"y_pred = model.predict(X_test_lsa)\n",
|
893 |
+
"\n",
|
894 |
+
"# Вывод результатов\n",
|
895 |
+
"print(classification_report(y_test, y_pred))\n",
|
896 |
+
"print(f'Accuracy: {accuracy_score(y_test, y_pred)}')\n",
|
897 |
+
"print(f'F1 score: {f1_score(y_test, y_pred)}')"
|
898 |
+
]
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"cell_type": "markdown",
|
902 |
+
"metadata": {},
|
903 |
+
"source": [
|
904 |
+
"## Создание пайплайна"
|
905 |
+
]
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"cell_type": "code",
|
909 |
+
"execution_count": 54,
|
910 |
+
"metadata": {},
|
911 |
+
"outputs": [
|
912 |
+
{
|
913 |
+
"name": "stderr",
|
914 |
+
"output_type": "stream",
|
915 |
+
"text": [
|
916 |
+
"[nltk_data] Downloading package stopwords to /home/vera/nltk_data...\n",
|
917 |
+
"[nltk_data] Package stopwords is already up-to-date!\n",
|
918 |
+
"[nltk_data] Downloading package punkt to /home/vera/nltk_data...\n",
|
919 |
+
"[nltk_data] Package punkt is already up-to-date!\n"
|
920 |
+
]
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"data": {
|
924 |
+
"text/html": [
|
925 |
+
"<style>#sk-container-id-1 {\n",
|
926 |
+
" /* Definition of color scheme common for light and dark mode */\n",
|
927 |
+
" --sklearn-color-text: black;\n",
|
928 |
+
" --sklearn-color-line: gray;\n",
|
929 |
+
" /* Definition of color scheme for unfitted estimators */\n",
|
930 |
+
" --sklearn-color-unfitted-level-0: #fff5e6;\n",
|
931 |
+
" --sklearn-color-unfitted-level-1: #f6e4d2;\n",
|
932 |
+
" --sklearn-color-unfitted-level-2: #ffe0b3;\n",
|
933 |
+
" --sklearn-color-unfitted-level-3: chocolate;\n",
|
934 |
+
" /* Definition of color scheme for fitted estimators */\n",
|
935 |
+
" --sklearn-color-fitted-level-0: #f0f8ff;\n",
|
936 |
+
" --sklearn-color-fitted-level-1: #d4ebff;\n",
|
937 |
+
" --sklearn-color-fitted-level-2: #b3dbfd;\n",
|
938 |
+
" --sklearn-color-fitted-level-3: cornflowerblue;\n",
|
939 |
+
"\n",
|
940 |
+
" /* Specific color for light theme */\n",
|
941 |
+
" --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
|
942 |
+
" --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));\n",
|
943 |
+
" --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
|
944 |
+
" --sklearn-color-icon: #696969;\n",
|
945 |
+
"\n",
|
946 |
+
" @media (prefers-color-scheme: dark) {\n",
|
947 |
+
" /* Redefinition of color scheme for dark theme */\n",
|
948 |
+
" --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
|
949 |
+
" --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));\n",
|
950 |
+
" --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
|
951 |
+
" --sklearn-color-icon: #878787;\n",
|
952 |
+
" }\n",
|
953 |
+
"}\n",
|
954 |
+
"\n",
|
955 |
+
"#sk-container-id-1 {\n",
|
956 |
+
" color: var(--sklearn-color-text);\n",
|
957 |
+
"}\n",
|
958 |
+
"\n",
|
959 |
+
"#sk-container-id-1 pre {\n",
|
960 |
+
" padding: 0;\n",
|
961 |
+
"}\n",
|
962 |
+
"\n",
|
963 |
+
"#sk-container-id-1 input.sk-hidden--visually {\n",
|
964 |
+
" border: 0;\n",
|
965 |
+
" clip: rect(1px 1px 1px 1px);\n",
|
966 |
+
" clip: rect(1px, 1px, 1px, 1px);\n",
|
967 |
+
" height: 1px;\n",
|
968 |
+
" margin: -1px;\n",
|
969 |
+
" overflow: hidden;\n",
|
970 |
+
" padding: 0;\n",
|
971 |
+
" position: absolute;\n",
|
972 |
+
" width: 1px;\n",
|
973 |
+
"}\n",
|
974 |
+
"\n",
|
975 |
+
"#sk-container-id-1 div.sk-dashed-wrapped {\n",
|
976 |
+
" border: 1px dashed var(--sklearn-color-line);\n",
|
977 |
+
" margin: 0 0.4em 0.5em 0.4em;\n",
|
978 |
+
" box-sizing: border-box;\n",
|
979 |
+
" padding-bottom: 0.4em;\n",
|
980 |
+
" background-color: var(--sklearn-color-background);\n",
|
981 |
+
"}\n",
|
982 |
+
"\n",
|
983 |
+
"#sk-container-id-1 div.sk-container {\n",
|
984 |
+
" /* jupyter's `normalize.less` sets `[hidden] { display: none; }`\n",
|
985 |
+
" but bootstrap.min.css set `[hidden] { display: none !important; }`\n",
|
986 |
+
" so we also need the `!important` here to be able to override the\n",
|
987 |
+
" default hidden behavior on the sphinx rendered scikit-learn.org.\n",
|
988 |
+
" See: https://github.com/scikit-learn/scikit-learn/issues/21755 */\n",
|
989 |
+
" display: inline-block !important;\n",
|
990 |
+
" position: relative;\n",
|
991 |
+
"}\n",
|
992 |
+
"\n",
|
993 |
+
"#sk-container-id-1 div.sk-text-repr-fallback {\n",
|
994 |
+
" display: none;\n",
|
995 |
+
"}\n",
|
996 |
+
"\n",
|
997 |
+
"div.sk-parallel-item,\n",
|
998 |
+
"div.sk-serial,\n",
|
999 |
+
"div.sk-item {\n",
|
1000 |
+
" /* draw centered vertical line to link estimators */\n",
|
1001 |
+
" background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));\n",
|
1002 |
+
" background-size: 2px 100%;\n",
|
1003 |
+
" background-repeat: no-repeat;\n",
|
1004 |
+
" background-position: center center;\n",
|
1005 |
+
"}\n",
|
1006 |
+
"\n",
|
1007 |
+
"/* Parallel-specific style estimator block */\n",
|
1008 |
+
"\n",
|
1009 |
+
"#sk-container-id-1 div.sk-parallel-item::after {\n",
|
1010 |
+
" content: \"\";\n",
|
1011 |
+
" width: 100%;\n",
|
1012 |
+
" border-bottom: 2px solid var(--sklearn-color-text-on-default-background);\n",
|
1013 |
+
" flex-grow: 1;\n",
|
1014 |
+
"}\n",
|
1015 |
+
"\n",
|
1016 |
+
"#sk-container-id-1 div.sk-parallel {\n",
|
1017 |
+
" display: flex;\n",
|
1018 |
+
" align-items: stretch;\n",
|
1019 |
+
" justify-content: center;\n",
|
1020 |
+
" background-color: var(--sklearn-color-background);\n",
|
1021 |
+
" position: relative;\n",
|
1022 |
+
"}\n",
|
1023 |
+
"\n",
|
1024 |
+
"#sk-container-id-1 div.sk-parallel-item {\n",
|
1025 |
+
" display: flex;\n",
|
1026 |
+
" flex-direction: column;\n",
|
1027 |
+
"}\n",
|
1028 |
+
"\n",
|
1029 |
+
"#sk-container-id-1 div.sk-parallel-item:first-child::after {\n",
|
1030 |
+
" align-self: flex-end;\n",
|
1031 |
+
" width: 50%;\n",
|
1032 |
+
"}\n",
|
1033 |
+
"\n",
|
1034 |
+
"#sk-container-id-1 div.sk-parallel-item:last-child::after {\n",
|
1035 |
+
" align-self: flex-start;\n",
|
1036 |
+
" width: 50%;\n",
|
1037 |
+
"}\n",
|
1038 |
+
"\n",
|
1039 |
+
"#sk-container-id-1 div.sk-parallel-item:only-child::after {\n",
|
1040 |
+
" width: 0;\n",
|
1041 |
+
"}\n",
|
1042 |
+
"\n",
|
1043 |
+
"/* Serial-specific style estimator block */\n",
|
1044 |
+
"\n",
|
1045 |
+
"#sk-container-id-1 div.sk-serial {\n",
|
1046 |
+
" display: flex;\n",
|
1047 |
+
" flex-direction: column;\n",
|
1048 |
+
" align-items: center;\n",
|
1049 |
+
" background-color: var(--sklearn-color-background);\n",
|
1050 |
+
" padding-right: 1em;\n",
|
1051 |
+
" padding-left: 1em;\n",
|
1052 |
+
"}\n",
|
1053 |
+
"\n",
|
1054 |
+
"\n",
|
1055 |
+
"/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is\n",
|
1056 |
+
"clickable and can be expanded/collapsed.\n",
|
1057 |
+
"- Pipeline and ColumnTransformer use this feature and define the default style\n",
|
1058 |
+
"- Estimators will overwrite some part of the style using the `sk-estimator` class\n",
|
1059 |
+
"*/\n",
|
1060 |
+
"\n",
|
1061 |
+
"/* Pipeline and ColumnTransformer style (default) */\n",
|
1062 |
+
"\n",
|
1063 |
+
"#sk-container-id-1 div.sk-toggleable {\n",
|
1064 |
+
" /* Default theme specific background. It is overwritten whether we have a\n",
|
1065 |
+
" specific estimator or a Pipeline/ColumnTransformer */\n",
|
1066 |
+
" background-color: var(--sklearn-color-background);\n",
|
1067 |
+
"}\n",
|
1068 |
+
"\n",
|
1069 |
+
"/* Toggleable label */\n",
|
1070 |
+
"#sk-container-id-1 label.sk-toggleable__label {\n",
|
1071 |
+
" cursor: pointer;\n",
|
1072 |
+
" display: block;\n",
|
1073 |
+
" width: 100%;\n",
|
1074 |
+
" margin-bottom: 0;\n",
|
1075 |
+
" padding: 0.5em;\n",
|
1076 |
+
" box-sizing: border-box;\n",
|
1077 |
+
" text-align: center;\n",
|
1078 |
+
"}\n",
|
1079 |
+
"\n",
|
1080 |
+
"#sk-container-id-1 label.sk-toggleable__label-arrow:before {\n",
|
1081 |
+
" /* Arrow on the left of the label */\n",
|
1082 |
+
" content: \"▸\";\n",
|
1083 |
+
" float: left;\n",
|
1084 |
+
" margin-right: 0.25em;\n",
|
1085 |
+
" color: var(--sklearn-color-icon);\n",
|
1086 |
+
"}\n",
|
1087 |
+
"\n",
|
1088 |
+
"#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {\n",
|
1089 |
+
" color: var(--sklearn-color-text);\n",
|
1090 |
+
"}\n",
|
1091 |
+
"\n",
|
1092 |
+
"/* Toggleable content - dropdown */\n",
|
1093 |
+
"\n",
|
1094 |
+
"#sk-container-id-1 div.sk-toggleable__content {\n",
|
1095 |
+
" max-height: 0;\n",
|
1096 |
+
" max-width: 0;\n",
|
1097 |
+
" overflow: hidden;\n",
|
1098 |
+
" text-align: left;\n",
|
1099 |
+
" /* unfitted */\n",
|
1100 |
+
" background-color: var(--sklearn-color-unfitted-level-0);\n",
|
1101 |
+
"}\n",
|
1102 |
+
"\n",
|
1103 |
+
"#sk-container-id-1 div.sk-toggleable__content.fitted {\n",
|
1104 |
+
" /* fitted */\n",
|
1105 |
+
" background-color: var(--sklearn-color-fitted-level-0);\n",
|
1106 |
+
"}\n",
|
1107 |
+
"\n",
|
1108 |
+
"#sk-container-id-1 div.sk-toggleable__content pre {\n",
|
1109 |
+
" margin: 0.2em;\n",
|
1110 |
+
" border-radius: 0.25em;\n",
|
1111 |
+
" color: var(--sklearn-color-text);\n",
|
1112 |
+
" /* unfitted */\n",
|
1113 |
+
" background-color: var(--sklearn-color-unfitted-level-0);\n",
|
1114 |
+
"}\n",
|
1115 |
+
"\n",
|
1116 |
+
"#sk-container-id-1 div.sk-toggleable__content.fitted pre {\n",
|
1117 |
+
" /* unfitted */\n",
|
1118 |
+
" background-color: var(--sklearn-color-fitted-level-0);\n",
|
1119 |
+
"}\n",
|
1120 |
+
"\n",
|
1121 |
+
"#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {\n",
|
1122 |
+
" /* Expand drop-down */\n",
|
1123 |
+
" max-height: 200px;\n",
|
1124 |
+
" max-width: 100%;\n",
|
1125 |
+
" overflow: auto;\n",
|
1126 |
+
"}\n",
|
1127 |
+
"\n",
|
1128 |
+
"#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {\n",
|
1129 |
+
" content: \"▾\";\n",
|
1130 |
+
"}\n",
|
1131 |
+
"\n",
|
1132 |
+
"/* Pipeline/ColumnTransformer-specific style */\n",
|
1133 |
+
"\n",
|
1134 |
+
"#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
|
1135 |
+
" color: var(--sklearn-color-text);\n",
|
1136 |
+
" background-color: var(--sklearn-color-unfitted-level-2);\n",
|
1137 |
+
"}\n",
|
1138 |
+
"\n",
|
1139 |
+
"#sk-container-id-1 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
|
1140 |
+
" background-color: var(--sklearn-color-fitted-level-2);\n",
|
1141 |
+
"}\n",
|
1142 |
+
"\n",
|
1143 |
+
"/* Estimator-specific style */\n",
|
1144 |
+
"\n",
|
1145 |
+
"/* Colorize estimator box */\n",
|
1146 |
+
"#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
|
1147 |
+
" /* unfitted */\n",
|
1148 |
+
" background-color: var(--sklearn-color-unfitted-level-2);\n",
|
1149 |
+
"}\n",
|
1150 |
+
"\n",
|
1151 |
+
"#sk-container-id-1 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
|
1152 |
+
" /* fitted */\n",
|
1153 |
+
" background-color: var(--sklearn-color-fitted-level-2);\n",
|
1154 |
+
"}\n",
|
1155 |
+
"\n",
|
1156 |
+
"#sk-container-id-1 div.sk-label label.sk-toggleable__label,\n",
|
1157 |
+
"#sk-container-id-1 div.sk-label label {\n",
|
1158 |
+
" /* The background is the default theme color */\n",
|
1159 |
+
" color: var(--sklearn-color-text-on-default-background);\n",
|
1160 |
+
"}\n",
|
1161 |
+
"\n",
|
1162 |
+
"/* On hover, darken the color of the background */\n",
|
1163 |
+
"#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {\n",
|
1164 |
+
" color: var(--sklearn-color-text);\n",
|
1165 |
+
" background-color: var(--sklearn-color-unfitted-level-2);\n",
|
1166 |
+
"}\n",
|
1167 |
+
"\n",
|
1168 |
+
"/* Label box, darken color on hover, fitted */\n",
|
1169 |
+
"#sk-container-id-1 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {\n",
|
1170 |
+
" color: var(--sklearn-color-text);\n",
|
1171 |
+
" background-color: var(--sklearn-color-fitted-level-2);\n",
|
1172 |
+
"}\n",
|
1173 |
+
"\n",
|
1174 |
+
"/* Estimator label */\n",
|
1175 |
+
"\n",
|
1176 |
+
"#sk-container-id-1 div.sk-label label {\n",
|
1177 |
+
" font-family: monospace;\n",
|
1178 |
+
" font-weight: bold;\n",
|
1179 |
+
" display: inline-block;\n",
|
1180 |
+
" line-height: 1.2em;\n",
|
1181 |
+
"}\n",
|
1182 |
+
"\n",
|
1183 |
+
"#sk-container-id-1 div.sk-label-container {\n",
|
1184 |
+
" text-align: center;\n",
|
1185 |
+
"}\n",
|
1186 |
+
"\n",
|
1187 |
+
"/* Estimator-specific */\n",
|
1188 |
+
"#sk-container-id-1 div.sk-estimator {\n",
|
1189 |
+
" font-family: monospace;\n",
|
1190 |
+
" border: 1px dotted var(--sklearn-color-border-box);\n",
|
1191 |
+
" border-radius: 0.25em;\n",
|
1192 |
+
" box-sizing: border-box;\n",
|
1193 |
+
" margin-bottom: 0.5em;\n",
|
1194 |
+
" /* unfitted */\n",
|
1195 |
+
" background-color: var(--sklearn-color-unfitted-level-0);\n",
|
1196 |
+
"}\n",
|
1197 |
+
"\n",
|
1198 |
+
"#sk-container-id-1 div.sk-estimator.fitted {\n",
|
1199 |
+
" /* fitted */\n",
|
1200 |
+
" background-color: var(--sklearn-color-fitted-level-0);\n",
|
1201 |
+
"}\n",
|
1202 |
+
"\n",
|
1203 |
+
"/* on hover */\n",
|
1204 |
+
"#sk-container-id-1 div.sk-estimator:hover {\n",
|
1205 |
+
" /* unfitted */\n",
|
1206 |
+
" background-color: var(--sklearn-color-unfitted-level-2);\n",
|
1207 |
+
"}\n",
|
1208 |
+
"\n",
|
1209 |
+
"#sk-container-id-1 div.sk-estimator.fitted:hover {\n",
|
1210 |
+
" /* fitted */\n",
|
1211 |
+
" background-color: var(--sklearn-color-fitted-level-2);\n",
|
1212 |
+
"}\n",
|
1213 |
+
"\n",
|
1214 |
+
"/* Specification for estimator info (e.g. \"i\" and \"?\") */\n",
|
1215 |
+
"\n",
|
1216 |
+
"/* Common style for \"i\" and \"?\" */\n",
|
1217 |
+
"\n",
|
1218 |
+
".sk-estimator-doc-link,\n",
|
1219 |
+
"a:link.sk-estimator-doc-link,\n",
|
1220 |
+
"a:visited.sk-estimator-doc-link {\n",
|
1221 |
+
" float: right;\n",
|
1222 |
+
" font-size: smaller;\n",
|
1223 |
+
" line-height: 1em;\n",
|
1224 |
+
" font-family: monospace;\n",
|
1225 |
+
" background-color: var(--sklearn-color-background);\n",
|
1226 |
+
" border-radius: 1em;\n",
|
1227 |
+
" height: 1em;\n",
|
1228 |
+
" width: 1em;\n",
|
1229 |
+
" text-decoration: none !important;\n",
|
1230 |
+
" margin-left: 1ex;\n",
|
1231 |
+
" /* unfitted */\n",
|
1232 |
+
" border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
|
1233 |
+
" color: var(--sklearn-color-unfitted-level-1);\n",
|
1234 |
+
"}\n",
|
1235 |
+
"\n",
|
1236 |
+
".sk-estimator-doc-link.fitted,\n",
|
1237 |
+
"a:link.sk-estimator-doc-link.fitted,\n",
|
1238 |
+
"a:visited.sk-estimator-doc-link.fitted {\n",
|
1239 |
+
" /* fitted */\n",
|
1240 |
+
" border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
|
1241 |
+
" color: var(--sklearn-color-fitted-level-1);\n",
|
1242 |
+
"}\n",
|
1243 |
+
"\n",
|
1244 |
+
"/* On hover */\n",
|
1245 |
+
"div.sk-estimator:hover .sk-estimator-doc-link:hover,\n",
|
1246 |
+
".sk-estimator-doc-link:hover,\n",
|
1247 |
+
"div.sk-label-container:hover .sk-estimator-doc-link:hover,\n",
|
1248 |
+
".sk-estimator-doc-link:hover {\n",
|
1249 |
+
" /* unfitted */\n",
|
1250 |
+
" background-color: var(--sklearn-color-unfitted-level-3);\n",
|
1251 |
+
" color: var(--sklearn-color-background);\n",
|
1252 |
+
" text-decoration: none;\n",
|
1253 |
+
"}\n",
|
1254 |
+
"\n",
|
1255 |
+
"div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,\n",
|
1256 |
+
".sk-estimator-doc-link.fitted:hover,\n",
|
1257 |
+
"div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,\n",
|
1258 |
+
".sk-estimator-doc-link.fitted:hover {\n",
|
1259 |
+
" /* fitted */\n",
|
1260 |
+
" background-color: var(--sklearn-color-fitted-level-3);\n",
|
1261 |
+
" color: var(--sklearn-color-background);\n",
|
1262 |
+
" text-decoration: none;\n",
|
1263 |
+
"}\n",
|
1264 |
+
"\n",
|
1265 |
+
"/* Span, style for the box shown on hovering the info icon */\n",
|
1266 |
+
".sk-estimator-doc-link span {\n",
|
1267 |
+
" display: none;\n",
|
1268 |
+
" z-index: 9999;\n",
|
1269 |
+
" position: relative;\n",
|
1270 |
+
" font-weight: normal;\n",
|
1271 |
+
" right: .2ex;\n",
|
1272 |
+
" padding: .5ex;\n",
|
1273 |
+
" margin: .5ex;\n",
|
1274 |
+
" width: min-content;\n",
|
1275 |
+
" min-width: 20ex;\n",
|
1276 |
+
" max-width: 50ex;\n",
|
1277 |
+
" color: var(--sklearn-color-text);\n",
|
1278 |
+
" box-shadow: 2pt 2pt 4pt #999;\n",
|
1279 |
+
" /* unfitted */\n",
|
1280 |
+
" background: var(--sklearn-color-unfitted-level-0);\n",
|
1281 |
+
" border: .5pt solid var(--sklearn-color-unfitted-level-3);\n",
|
1282 |
+
"}\n",
|
1283 |
+
"\n",
|
1284 |
+
".sk-estimator-doc-link.fitted span {\n",
|
1285 |
+
" /* fitted */\n",
|
1286 |
+
" background: var(--sklearn-color-fitted-level-0);\n",
|
1287 |
+
" border: var(--sklearn-color-fitted-level-3);\n",
|
1288 |
+
"}\n",
|
1289 |
+
"\n",
|
1290 |
+
".sk-estimator-doc-link:hover span {\n",
|
1291 |
+
" display: block;\n",
|
1292 |
+
"}\n",
|
1293 |
+
"\n",
|
1294 |
+
"/* \"?\"-specific style due to the `<a>` HTML tag */\n",
|
1295 |
+
"\n",
|
1296 |
+
"#sk-container-id-1 a.estimator_doc_link {\n",
|
1297 |
+
" float: right;\n",
|
1298 |
+
" font-size: 1rem;\n",
|
1299 |
+
" line-height: 1em;\n",
|
1300 |
+
" font-family: monospace;\n",
|
1301 |
+
" background-color: var(--sklearn-color-background);\n",
|
1302 |
+
" border-radius: 1rem;\n",
|
1303 |
+
" height: 1rem;\n",
|
1304 |
+
" width: 1rem;\n",
|
1305 |
+
" text-decoration: none;\n",
|
1306 |
+
" /* unfitted */\n",
|
1307 |
+
" color: var(--sklearn-color-unfitted-level-1);\n",
|
1308 |
+
" border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
|
1309 |
+
"}\n",
|
1310 |
+
"\n",
|
1311 |
+
"#sk-container-id-1 a.estimator_doc_link.fitted {\n",
|
1312 |
+
" /* fitted */\n",
|
1313 |
+
" border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
|
1314 |
+
" color: var(--sklearn-color-fitted-level-1);\n",
|
1315 |
+
"}\n",
|
1316 |
+
"\n",
|
1317 |
+
"/* On hover */\n",
|
1318 |
+
"#sk-container-id-1 a.estimator_doc_link:hover {\n",
|
1319 |
+
" /* unfitted */\n",
|
1320 |
+
" background-color: var(--sklearn-color-unfitted-level-3);\n",
|
1321 |
+
" color: var(--sklearn-color-background);\n",
|
1322 |
+
" text-decoration: none;\n",
|
1323 |
+
"}\n",
|
1324 |
+
"\n",
|
1325 |
+
"#sk-container-id-1 a.estimator_doc_link.fitted:hover {\n",
|
1326 |
+
" /* fitted */\n",
|
1327 |
+
" background-color: var(--sklearn-color-fitted-level-3);\n",
|
1328 |
+
"}\n",
|
1329 |
+
"</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[('preprocessor', TextPreprocessor()),\n",
|
1330 |
+
" ('vectorizer',\n",
|
1331 |
+
" TfidfVectorizer(max_df=0.9, min_df=500,\n",
|
1332 |
+
" stop_words=['и', 'в', 'во', 'не', 'что', 'он',\n",
|
1333 |
+
" 'на', 'я', 'с', 'со', 'как', 'а',\n",
|
1334 |
+
" 'то', 'все', 'она', 'так', 'его',\n",
|
1335 |
+
" 'но', 'да', 'ты', 'к', 'у', 'же',\n",
|
1336 |
+
" 'вы', 'за', 'бы', 'по', 'только',\n",
|
1337 |
+
" 'ее', 'мне', ...])),\n",
|
1338 |
+
" ('lsa',\n",
|
1339 |
+
" Pipeline(steps=[('truncatedsvd',\n",
|
1340 |
+
" TruncatedSVD(n_components=500)),\n",
|
1341 |
+
" ('normalizer', Normalizer(copy=False))])),\n",
|
1342 |
+
" ('classifier', LogisticRegression())])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\"> Pipeline<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.4/modules/generated/sklearn.pipeline.Pipeline.html\">?<span>Documentation for Pipeline</span></a><span class=\"sk-estimator-doc-link fitted\">i<span>Fitted</span></span></label><div class=\"sk-toggleable__content fitted\"><pre>Pipeline(steps=[('preprocessor', TextPreprocessor()),\n",
|
1343 |
+
" ('vectorizer',\n",
|
1344 |
+
" TfidfVectorizer(max_df=0.9, min_df=500,\n",
|
1345 |
+
" stop_words=['и', 'в', 'во', 'не', 'что', 'он',\n",
|
1346 |
+
" 'на', 'я', 'с', 'со', 'как', 'а',\n",
|
1347 |
+
" 'то', 'все', 'она', 'так', 'его',\n",
|
1348 |
+
" 'но', 'да', 'ты', 'к', 'у', 'же',\n",
|
1349 |
+
" 'вы', 'за', 'бы', 'по', 'только',\n",
|
1350 |
+
" 'ее', 'мне', ...])),\n",
|
1351 |
+
" ('lsa',\n",
|
1352 |
+
" Pipeline(steps=[('truncatedsvd',\n",
|
1353 |
+
" TruncatedSVD(n_components=500)),\n",
|
1354 |
+
" ('normalizer', Normalizer(copy=False))])),\n",
|
1355 |
+
" ('classifier', LogisticRegression())])</pre></div> </div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\">TextPreprocessor</label><div class=\"sk-toggleable__content fitted\"><pre>TextPreprocessor()</pre></div> </div></div><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\"> TfidfVectorizer<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.4/modules/generated/sklearn.feature_extraction.text.TfidfVectorizer.html\">?<span>Documentation for TfidfVectorizer</span></a></label><div class=\"sk-toggleable__content fitted\"><pre>TfidfVectorizer(max_df=0.9, min_df=500,\n",
|
1356 |
+
" stop_words=['и', 'в', 'во', 'не', 'что', 'он', 'на', 'я', 'с',\n",
|
1357 |
+
" 'со', 'как', 'а', 'то', 'все', 'она', 'так', 'его',\n",
|
1358 |
+
" 'но', 'да', 'ты', 'к', 'у', 'же', 'вы', 'за', 'бы',\n",
|
1359 |
+
" 'по', 'только', 'ее', 'мне', ...])</pre></div> </div></div><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-4\" type=\"checkbox\" ><label for=\"sk-estimator-id-4\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\"> lsa: Pipeline<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.4/modules/generated/sklearn.pipeline.Pipeline.html\">?<span>Documentation for lsa: Pipeline</span></a></label><div class=\"sk-toggleable__content fitted\"><pre>Pipeline(steps=[('truncatedsvd', TruncatedSVD(n_components=500)),\n",
|
1360 |
+
" ('normalizer', Normalizer(copy=False))])</pre></div> </div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-5\" type=\"checkbox\" ><label for=\"sk-estimator-id-5\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\"> TruncatedSVD<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.4/modules/generated/sklearn.decomposition.TruncatedSVD.html\">?<span>Documentation for TruncatedSVD</span></a></label><div class=\"sk-toggleable__content fitted\"><pre>TruncatedSVD(n_components=500)</pre></div> </div></div><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-6\" type=\"checkbox\" ><label for=\"sk-estimator-id-6\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\"> Normalizer<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.4/modules/generated/sklearn.preprocessing.Normalizer.html\">?<span>Documentation for Normalizer</span></a></label><div class=\"sk-toggleable__content fitted\"><pre>Normalizer(copy=False)</pre></div> </div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-7\" type=\"checkbox\" ><label for=\"sk-estimator-id-7\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow fitted\"> LogisticRegression<a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.4/modules/generated/sklearn.linear_model.LogisticRegression.html\">?<span>Documentation for LogisticRegression</span></a></label><div class=\"sk-toggleable__content fitted\"><pre>LogisticRegression()</pre></div> </div></div></div></div></div></div>"
|
1361 |
+
],
|
1362 |
+
"text/plain": [
|
1363 |
+
"Pipeline(steps=[('preprocessor', TextPreprocessor()),\n",
|
1364 |
+
" ('vectorizer',\n",
|
1365 |
+
" TfidfVectorizer(max_df=0.9, min_df=500,\n",
|
1366 |
+
" stop_words=['и', 'в', 'во', 'не', 'что', 'он',\n",
|
1367 |
+
" 'на', 'я', 'с', 'со', 'как', 'а',\n",
|
1368 |
+
" 'то', 'все', 'она', 'так', 'его',\n",
|
1369 |
+
" 'но', 'да', 'ты', 'к', 'у', 'же',\n",
|
1370 |
+
" 'вы', 'за', 'бы', 'по', 'только',\n",
|
1371 |
+
" 'ее', 'мне', ...])),\n",
|
1372 |
+
" ('lsa',\n",
|
1373 |
+
" Pipeline(steps=[('truncatedsvd',\n",
|
1374 |
+
" TruncatedSVD(n_components=500)),\n",
|
1375 |
+
" ('normalizer', Normalizer(copy=False))])),\n",
|
1376 |
+
" ('classifier', LogisticRegression())])"
|
1377 |
+
]
|
1378 |
+
},
|
1379 |
+
"execution_count": 54,
|
1380 |
+
"metadata": {},
|
1381 |
+
"output_type": "execute_result"
|
1382 |
+
}
|
1383 |
+
],
|
1384 |
+
"source": [
|
1385 |
+
"import re\n",
|
1386 |
+
"import pandas as pd\n",
|
1387 |
+
"import numpy as np\n",
|
1388 |
+
"from sklearn.base import BaseEstimator, TransformerMixin\n",
|
1389 |
+
"from sklearn.feature_extraction.text import TfidfVectorizer\n",
|
1390 |
+
"from sklearn.decomposition import TruncatedSVD\n",
|
1391 |
+
"from sklearn.pipeline import Pipeline, FeatureUnion\n",
|
1392 |
+
"from sklearn.linear_model import LogisticRegression\n",
|
1393 |
+
"from sklearn.preprocessing import Normalizer\n",
|
1394 |
+
"import joblib\n",
|
1395 |
+
"import nltk\n",
|
1396 |
+
"from nltk.corpus import stopwords\n",
|
1397 |
+
"from pymorphy2 import MorphAnalyzer\n",
|
1398 |
+
"\n",
|
1399 |
+
"nltk.download('stopwords')\n",
|
1400 |
+
"nltk.download('punkt')\n",
|
1401 |
+
"\n",
|
1402 |
+
"class TextPreprocessor(BaseEstimator, TransformerMixin):\n",
|
1403 |
+
" def __init__(self):\n",
|
1404 |
+
" self.stop_words = set(stopwords.words('russian'))\n",
|
1405 |
+
" self.morph = MorphAnalyzer()\n",
|
1406 |
+
"\n",
|
1407 |
+
" def preprocess_text(self, text):\n",
|
1408 |
+
" # Удаление всего, что не является буквами или знаками препинания\n",
|
1409 |
+
" clean_pattern = re.compile(r'[^a-zA-Zа-яА-ЯёЁ0-9.,!?;:\\s]')\n",
|
1410 |
+
" text = clean_pattern.sub('', text)\n",
|
1411 |
+
" url_pattern = re.compile(r'http\\S+|www\\S+|https\\S+')\n",
|
1412 |
+
" text = url_pattern.sub(r'', text)\n",
|
1413 |
+
" text = text.translate(str.maketrans('', '', string.punctuation))\n",
|
1414 |
+
" text = text.lower()\n",
|
1415 |
+
" tokens = text.split()\n",
|
1416 |
+
" lemmatized_text = ' '.join([self.morph.parse(word)[0].normal_form for word in tokens if word not in self.stop_words])\n",
|
1417 |
+
" return lemmatized_text\n",
|
1418 |
+
"\n",
|
1419 |
+
" def fit(self, X, y=None):\n",
|
1420 |
+
" return self\n",
|
1421 |
+
"\n",
|
1422 |
+
" def transform(self, X, y=None):\n",
|
1423 |
+
" return X.apply(self.preprocess_text)\n",
|
1424 |
+
"\n",
|
1425 |
+
"\n",
|
1426 |
+
"# Load and preprocess the dataset\n",
|
1427 |
+
"df = pd.read_json('data/healthcare_facilities_reviews.jsonl', lines=True)\n",
|
1428 |
+
"df = df[['sentiment', 'content']]\n",
|
1429 |
+
"df['cleaned_text'] = df['content'].apply(TextPreprocessor().preprocess_text)\n",
|
1430 |
+
"df['sentiment'] = df['sentiment'].apply(lambda x: 1 if x == 'negative' else 0)\n",
|
1431 |
+
"\n",
|
1432 |
+
"# Split the dataset (this is only for training purposes)\n",
|
1433 |
+
"from sklearn.model_selection import train_test_split\n",
|
1434 |
+
"train_df, test_df = train_test_split(df, test_size=0.2, random_state=42)\n",
|
1435 |
+
"\n",
|
1436 |
+
"# Create the pipeline\n",
|
1437 |
+
"vectorizer = TfidfVectorizer(\n",
|
1438 |
+
" max_df=0.9,\n",
|
1439 |
+
" min_df=500,\n",
|
1440 |
+
" stop_words=stopwords.words('russian')\n",
|
1441 |
+
")\n",
|
1442 |
+
"\n",
|
1443 |
+
"lsa = TruncatedSVD(n_components=500)\n",
|
1444 |
+
"\n",
|
1445 |
+
"pipeline = Pipeline([\n",
|
1446 |
+
" ('preprocessor', TextPreprocessor()),\n",
|
1447 |
+
" ('vectorizer', vectorizer),\n",
|
1448 |
+
" ('lsa', make_pipeline(lsa, Normalizer(copy=False))),\n",
|
1449 |
+
" ('classifier', LogisticRegression())\n",
|
1450 |
+
"])\n",
|
1451 |
+
"\n",
|
1452 |
+
"# Train the model\n",
|
1453 |
+
"X_train = train_df['cleaned_text']\n",
|
1454 |
+
"y_train = train_df['sentiment']\n",
|
1455 |
+
"pipeline.fit(X_train, y_train)\n",
|
1456 |
+
"\n",
|
1457 |
+
"# Save the model\n",
|
1458 |
+
"# joblib.dump(pipeline, 'logistic_regression_pipeline.pkl')\n"
|
1459 |
+
]
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"cell_type": "code",
|
1463 |
+
"execution_count": 55,
|
1464 |
+
"metadata": {},
|
1465 |
+
"outputs": [
|
1466 |
+
{
|
1467 |
+
"data": {
|
1468 |
+
"text/plain": [
|
1469 |
+
"['logistic_regression_pipeline.pkl']"
|
1470 |
+
]
|
1471 |
+
},
|
1472 |
+
"execution_count": 55,
|
1473 |
+
"metadata": {},
|
1474 |
+
"output_type": "execute_result"
|
1475 |
+
}
|
1476 |
+
],
|
1477 |
+
"source": [
|
1478 |
+
"# Save the model for future use\n",
|
1479 |
+
"joblib.dump(pipeline, 'logistic_regression_pipeline.pkl')"
|
1480 |
+
]
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"cell_type": "code",
|
1484 |
+
"execution_count": 56,
|
1485 |
+
"metadata": {},
|
1486 |
+
"outputs": [],
|
1487 |
+
"source": [
|
1488 |
+
"# Load the model (if not already loaded)\n",
|
1489 |
+
"pipeline_test= joblib.load('logistic_regression_pipeline.pkl')"
|
1490 |
+
]
|
1491 |
+
},
|
1492 |
+
{
|
1493 |
+
"cell_type": "code",
|
1494 |
+
"execution_count": 61,
|
1495 |
+
"metadata": {},
|
1496 |
+
"outputs": [
|
1497 |
+
{
|
1498 |
+
"name": "stdout",
|
1499 |
+
"output_type": "stream",
|
1500 |
+
"text": [
|
1501 |
+
"Predicted class: 1\n",
|
1502 |
+
"Predicted proba: 0.898\n"
|
1503 |
+
]
|
1504 |
+
}
|
1505 |
+
],
|
1506 |
+
"source": [
|
1507 |
+
"# Sample text for prediction\n",
|
1508 |
+
"sample_text = \"Ужасная клиника, обслуживание из рук вон плохое, хотеловь бы выразить свое разочарование данным заведением. Советую обходить его мимо.\"\n",
|
1509 |
+
"\n",
|
1510 |
+
"# Use the pipeline to predict the class\n",
|
1511 |
+
"predicted_class = pipeline_test.predict(pd.Series([sample_text]))\n",
|
1512 |
+
"predicted_prob = pipeline_test.predict_proba(pd.Series([sample_text]))\n",
|
1513 |
+
"print(f\"Predicted class: {predicted_class[0]}\")\n",
|
1514 |
+
"print(f\"Predicted proba: {round(predicted_prob[0][1], 3)}\")"
|
1515 |
+
]
|
1516 |
+
}
|
1517 |
+
],
|
1518 |
+
"metadata": {
|
1519 |
+
"kernelspec": {
|
1520 |
+
"display_name": "base",
|
1521 |
+
"language": "python",
|
1522 |
+
"name": "python3"
|
1523 |
+
},
|
1524 |
+
"language_info": {
|
1525 |
+
"codemirror_mode": {
|
1526 |
+
"name": "ipython",
|
1527 |
+
"version": 3
|
1528 |
+
},
|
1529 |
+
"file_extension": ".py",
|
1530 |
+
"mimetype": "text/x-python",
|
1531 |
+
"name": "python",
|
1532 |
+
"nbconvert_exporter": "python",
|
1533 |
+
"pygments_lexer": "ipython3",
|
1534 |
+
"version": "3.10.14"
|
1535 |
+
}
|
1536 |
+
},
|
1537 |
+
"nbformat": 4,
|
1538 |
+
"nbformat_minor": 2
|
1539 |
+
}
|
pages/policlinic.py
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
import joblib
|
3 |
+
import pandas as pd
|
4 |
+
|
5 |
+
# Load the trained pipeline
|
6 |
+
pipeline = joblib.load('logistic_regression_pipeline.pkl')
|
7 |
+
|
8 |
+
# Streamlit application
|
9 |
+
st.title('Классификация отзывов на русском языке')
|
10 |
+
|
11 |
+
input_text = st.text_area('Введите текст отзыва')
|
12 |
+
|
13 |
+
if st.button('Предсказать'):
|
14 |
+
prediction = pipeline.predict(pd.Series([input_text]))
|
15 |
+
st.write(f'Предсказанный класс с помощью логрег: {prediction[0]}')
|