西野カナに「恋」とは何か聞いてみた

import requests
from bs4 import BeautifulSoup

base_url = "https://www.uta-net.com"
target_url = 'https://www.uta-net.com/search/?Aselect=1&Keyword=...&Bselect=3&x=0&y=0'
music_num = 167

r = requests.get(target_url)

soup = BeautifulSoup(r.text, "html.parser")
url_list = []
#曲一覧から各曲のURLを取り出してリストに入れる
for i in range(music_num):
   href = soup.find_all("td", attrs={"class": "side td1"})[i].contents[0].get("href")
   url_list.append(href)

kashi = ""
#曲ごとにRequestを送り歌詞を抽出する
for i in range(music_num):
   target_url = base_url + url_list[i]
   r = requests.get(target_url)
   soup = BeautifulSoup(r.text, "html.parser")

   for string in soup.find_all("div", attrs={"id": "kashi_area"})[0].strings:
       kashi += string

with open('kashi.txt', mode = 'w', encoding = 'utf-8') as fw:
   fw.write(kashi)

import re
# 英数字の削除
kashi = re.sub("[a-xA-Z0-9_]","",kashi)
# 記号の削除
kashi = re.sub("[!-/:-@[-`{-~]","",kashi)
# 空白・改行の削除
kashi = re.sub(u'\n\n', '\n', kashi)
kashi = re.sub(u'\r', '', kashi)

def tokenize(text):
    t = Tokenizer()
    tokens = t.tokenize(text)
    word = []
    stop_word = create_stop_word()
    for token in tokens:
        part_of_speech = token.part_of_speech.split(",")[0]
        if part_of_speech == "名詞":
            if not token.surface in stop_word:
                word.append(token.surface)
        if part_of_speech == "動詞":
            if not token.base_form in stop_word:
                word.append(token.base_form)
        if part_of_speech == "形容詞":
            if not token.base_form in stop_word:
                word.append(token.base_form)
        if part_of_speech == "形容動詞":
            if not token.base_form in stop_word:
                word.append(token.base_form)

    return word

窓   名詞,一般,*,*,*,*,窓,マド,マド
の   助詞,連体化,*,*,*,*,の,ノ,ノ
外   名詞,一般,*,*,*,*,外,ソト,ソト
を   助詞,格助詞,一般,*,*,*,を,ヲ,ヲ
眺め 動詞,自立,*,*,一段,連用形,眺める,ナガメ,ナガメ
ながら 助詞,接続助詞,*,*,*,*,ながら,ナガラ,ナガラ
考え 動詞,自立,*,*,一段,連用形,考える,カンガエ,カンガエ
てる 動詞,非自立,*,*,一段,基本形,てる,テル,テル

def create_stop_word():
    target_url = 'http://svn.sourceforge.jp/svnroot/slothlib/CSharp/Version1/SlothLib/NLP/Filter/StopWord/word/Japanese.txt'
    r =requests.get(target_url)
    soup=BeautifulSoup(r.text, "html.parser")
    stop_word=str(soup).split()
    #自分で追加
    my_stop_word=['いる','する','させる','の','色','真夏','身体','最初','知る','られる']
    stop_word.extend(my_stop_word)
    return stop_word

model = word2vec.Word2Vec(sentence, size=200, min_count=4, window=4, iter=50)

.most_similar(positive=["単語"])

from gensim import corpora
from janome.tokenizer import Tokenizer
from gensim.models import word2vec
import matplotlib.pyplot as plt
from wordcloud import WordCloud
import re
import requests
from bs4 import BeautifulSoup

with open("kashi.txt", "r", encoding="utf-8") as f:
    kashi = f.read()

# 英数字の削除
kashi = re.sub("[a-xA-Z0-9_]","",kashi)
# 記号の削除
kashi = re.sub("[!-/:-@[-`{-~]","",kashi)
# 空白・改行の削除
kashi = re.sub(u'\n\n', '\n', kashi)
kashi = re.sub(u'\r', '', kashi)
# 品詞を取り出し「名詞、動詞、形容詞、形容動詞」のリスト作成
def tokenize(text):
    t = Tokenizer()
    tokens = t.tokenize(text)
    word = []
    stop_word = create_stop_word()
    for token in tokens:
        part_of_speech = token.part_of_speech.split(",")[0]
        if part_of_speech == "名詞":
            if not token.surface in stop_word:
                word.append(token.surface)
        if part_of_speech == "動詞":
            if not token.base_form in stop_word:
                word.append(token.base_form)
        if part_of_speech == "形容詞":
            if not token.base_form in stop_word:
                word.append(token.base_form)
        if part_of_speech == "形容動詞":
            if not token.base_form in stop_word:
                word.append(token.base_form)

    return word

def create_stop_word():
    target_url = 'http://svn.sourceforge.jp/svnroot/slothlib/CSharp/Version1/SlothLib/NLP/Filter/StopWord/word/Japanese.txt'
    r =requests.get(target_url)
    soup=BeautifulSoup(r.text, "html.parser")
    stop_word=str(soup).split()
    #自分で追加
    my_stop_word=['いる','する','させる','の','色','真夏','身体','最初','知る','られる']
    stop_word.extend(my_stop_word)
    return stop_word

sentence = [tokenize(kashi)]
model = word2vec.Word2Vec(sentence, size=200, min_count=4, window=4, iter=50)
print(model.wv.most_similar(positive=[u"恋"], topn=10))

model.wv.most_similar(positive=[u"単語1", u"単語2"], negative=[u"単語3"], topn=10)

model.wv.most_similar(positive=[u"恋"], negative=["心"], topn=3)

model.wv.most_similar(positive=[u"恋", u"嘘"], topn=3)