Topic-Detection/preprocess.py at master · AUT-Twitter-Analytics/Topic-Detection · GitHub

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
import re, hazm, nltk
import pandas as pd
from nltk.corpus import stopwords
from nltk.stem import PorterStemmer
from nltk.stem import WordNetLemmatizer

nltk.download("wordnet")
nltk.download("stopwords")


class Preprocess:
    def fit(self, data, lang):
        is_fa = data["lang"] == "fa"
        fa_data = data[is_fa]

        is_en = data["lang"] == "en"
        en_data = data[is_en]

        if lang == "fa":
            self.data = self.clean_fa(fa_data)
        elif lang == "en":
            self.data = self.clean_en(en_data)

    def clean_fa(self, data):
        data.text = self.fa_normalize(data.text)
        data.text = self.tokenizer(data.text)

        stemmer = hazm.Stemmer()
        lemmatizer = hazm.Lemmatizer()
        stopwords = hazm.stopwords_list()
        alphabet = set(list("ابپتثجچحخدذرزژسشصضطظعغفقکگلمنوهی"))

        data.text = data.apply(
            lambda row: self.stemLemmaStopWord(
                stemmer, lemmatizer, stopwords, alphabet, row.text
            ),
            axis=1,
        )
        return data

    def clean_en(self, data):
        data.text = self.normalize(data.text)
        data.text = self.tokenizer(data.text)

        stemmer = PorterStemmer()
        lemmatizer = WordNetLemmatizer()
        stopwords = set(stopwords.words("english"))
        alphabet = set(list("abcdefghijklmnopqrstuvwxyz"))

        data.text = data.apply(
            lambda row: self.stemLemmaStopWord(
                stemmer, lemmatizer, stopwords, alphabet, row.text
            ),
            axis=1,
        )
        return data

    def tokenizer(self, text):
        text = text.str.split(" ")
        return text

    def stemLemmaStopWord(self, stemmer, lemmatizer, stopwords, alphabet, tokens):
        final_tokens = []
        for token in tokens:
            stemmed_token = stemmer.stem(lemmatizer.lemmatize(token))
            if "#" in stemmed_token:
                stemmed_token = stemmed_token.split("#")[0]
            if (
                token not in stopwords
                and stemmed_token not in stopwords
                and not token == ""
                and stemmed_token not in alphabet
            ):
                final_tokens.append(stemmed_token)
        return final_tokens

    def fa_normalize(self, text):
        text = text.replace(to_replace=r"[ئيی]", value="ی", regex=True)
        text = text.replace(to_replace=r"[ك]", value="ک", regex=True)
        text = text.replace(to_replace=r"[ؤ]", value="و", regex=True)
        text = text.replace(to_replace=r"[ة]", value="ه", regex=True)
        text = text.replace(to_replace=r"[إأآا]", value="ا", regex=True)
        text = text.replace(
            to_replace=r"[^ابپتثجچحخدذرزژسشصضطظعغفقکگلمنوهی]", value=" ", regex=True
        )
        text = text.replace(to_replace=r"(.)\1+", value=r"\1", regex=True)
        text = text.replace(to_replace=r"[^\S\n\t]+", value=" ", regex=True)
        return text

    def en_normalize(self, text):
        text = text.replace(to_replace=r"@([A-Za-z0-9_]+)", value="", regex=True)
        text = text.replace(to_replace=r"http([^\s\\]+)", value="", regex=True)
        text = text.str.lower()
        text = text.replace(to_replace=r"[^a-z]", value=" ", regex=True)
        text = text.replace(to_replace=r"(.)\1+", value=r"\1", regex=True)
        text = text.replace(to_replace=r"[^\S\n\t]+", value=" ", regex=True)
        return text