wordcloudで凛として時雨の歌詞を可視化してみた

凛として時雨の歌詞でよく使われる単語を可視化してみた。
歌詞データをスクレイピング -> janomeで分かち書き -> wordcloudで可視化という流れになっている。

コード

# -*- coding: utf-8 -*-

from bs4 import BeautifulSoup
import urllib.request as req
from janome.tokenizer import Tokenizer
#jupyterでmatplotlibを使うときに記述
%matplotlib inline
import matplotlib.pyplot as plt
from wordcloud import WordCloud

t = Tokenizer()


def getInfo(url, link_list = []):
    
    res = req.urlopen(url)
    soup = BeautifulSoup(res, "html.parser")
    class_name = "side td1".strip()
    name_list = soup.find_all("td", class_ = class_name)
    
    for name in name_list:
        s = BeautifulSoup(str(name), "html.parser")
        link_list.append("https://www.uta-net.com" + s.find("a")["href"])
        
    return link_list


def wakati(word_list = []):
    
    link_list = getInfo("https://www.uta-net.com/artist/7840/6/")
    for link in link_list:
        res = req.urlopen(link)
        soup = BeautifulSoup(res, "html.parser")
        kashi = soup.find_all("div", id = "kashi_area")
        for doc in kashi:
            tokens = t.tokenize(doc.text)
            for token in tokens:
                if token.part_of_speech.split(',')[0] == u"名詞" or (token.part_of_speech.split(',')[0] == u"動詞" and len(token.surface) > 2):
                    word_list.append(token.surface)

    return word_list


def wordcloud():

    
    word_list = wakati()
    words = " ".join(word_list)
    fpath = '/usr/share/fonts/truetype/fonts-japanese-gothic.ttf'
    wordcloud = WordCloud(background_color="white", font_path=fpath, width=900, height=500).generate(words)
                         
    plt.figure(figsize=(15,12))
    plt.imshow(wordcloud)
    plt.axis("off")
    plt.show()
    wordcloud.to_file('wordcloud.png')
    

create_wordcloud()