My name is Kei @ airget0919. I am a machine learning engineer in Tokyo. I am writing this article thinking that I may be able to output some practical skills. This time, as a preliminary step to implement natural language processing, I will acquire articles from the Internet and perform a simple analysis.
You can get various information on the net by using scraping.
This time I'm using Python code to get the article.
Use BeautifulSoup
etc. to specify HTML or CSS information and extract the information.
Let's quantify whether the content of the article is negative or positive based on the Word Emotion Polarity Correspondence Table. I think. The word emotion polarity correspondence table defines the negative / positive degree for a word as -1 to 1.
Joy: Joy: Noun: 0.998861 Severe: Severe: Adjective: -0.999755
Etc.
For scraping, I used udemy course. For negative / positive analysis, I referred to this article. In addition, the subject of this analysis was Bungei Online.
Let's start the work according to the above. Let's start by importing the required libraries.
import requests
from bs4 import BeautifulSoup
import re
import itertools
import pandas as pd, numpy as np
import os
import glob
import pathlib
import re
import janome
import jaconv
from janome.tokenizer import Tokenizer
from janome.analyzer import Analyzer
from janome.charfilter import *
First, prepare for scraping.
Get the URL and apply requests
and BeautifulSoup
.
url = "https://bunshun.jp/" #Store Bunshun online link in url
res = requests.get(url) # requests.get()Store url in res using
soup = BeautifulSoup(res.text, "html.parser") #Now you are ready for scraping with Beautiful Soup.
Basically, articles are lined up as li
elements, and the parent element is ʻul` in many patterns.
I'm using the for statement to get the title and URL from the article list.
elems = soup.select("ul") #Since the list of articles was lined up as a li element, its parent element, ul, is specified.
title_list = [] #List to store article titles
url_list = [] #List to store article URLs
for sibling in elems[3]: # elems[3]I had a list I wanted. This for minute gets the article title and URL from the article list and stores them in the list respectively.
if sibling != "\n": #Excluded because line breaks were included
print(sibling.h3.string) #The title was in the h3 tag.
title_list.append(sibling.h3.string.replace('\u3000', ' ')) # \Since there was a part containing u3000, it was converted to blank
print(url + sibling.h3.a["href"]) #The link was stored in the href attribute of the a tag.
url_list.append(url + sibling.h3.a["href"]) #The part below the url obtained above was stored, so I added it.
Former Johnny's MADE leader Hikaru Inaba (29) and former Berryz Kobo idol "Shibuya Hotel Date" << Scoop Shooting >>
https://bunshun.jp//articles/-/40869
"A vast site of about 1,200 m2 near the station" "Land utilization that balances home rebuilding and community contribution" What is the answer given by professionals?
https://bunshun.jp//articles/-/40010
"Let's work 24 hours a day, 365 days a year" ... Watami's "idea education" was still going on
https://bunshun.jp//articles/-/40843
It is a harmful effect of "vertical administration"! Toru Hashimoto talks about "What's wrong with the Go To campaign?"
https://bunshun.jp//articles/-/40877
Police officer spit and arrested for obstructing public duties Yomiuri Shimbun disciplines Seoul bureau reporter
https://bunshun.jp//articles/-/40868
"It's a ridiculous misunderstanding" to the Adachi Ward Council of homosexual discrimination ... What my 81-year-old grandmother's letter told me
https://bunshun.jp//articles/-/40826
Family ramen "Goodwill War" "Yoshimuraya vs. Rokkakuya" Betrayal and succumbing black history
In order to create a link list, create a while statement and get the URL if there is a link to the next page and transition to the page, and if the transitioned page also has a link to the next page, get it Turn the loop of transition. If there is no link on the next page, it will move to the next article. By doing this, you can get the links of all pages of all articles in list format.
news_list = [] #Links to all news articles are stored here.
for pickup_link in url_list: #This for statement retrieves the URL from the URL list.
news = [] #Since news articles are separated by page, we will include a link for each page in this list.
news.append(pickup_link) #Store the first link
pickup_res = requests.get(pickup_link) # requests.get()Get the page from the link using
pickup_soup = BeautifulSoup(pickup_res.text, "html.parser") #Apply Beautiful Soup
while True: #In this while statement, if there is a link to the next page, that link is acquired and the loop is routed to that page.
try: #If there is a link to the next page at the transition destination, this loop will be repeated forever.
next_link = pickup_soup.find("a", class_="next menu-link ga_tracking")["href"] # next menu-link ga_The href attribute of the a tag with the class tracking was the link to the next page.
next_link = url + next_link
next_res = requests.get(next_link) # requests.get()And BeautifulSoup are used to get the page information of the transition destination.
pickup_soup = BeautifulSoup(next_res.text, "html.parser")
news.append(next_link) #Add each page information to news.
except Exception: #If there is no link to the next page, this process will be performed.
news_list.append(news) #The URL of all the articles in the title is stored in news, so news_Store in list.
break
display(news_list) #Display the created URL list.
[['https://bunshun.jp//articles/-/40869',
'https://bunshun.jp//articles/-/40869?page=2',
'https://bunshun.jp//articles/-/40869?page=3',
'https://bunshun.jp//articles/-/40869?page=4'],
['https://bunshun.jp//articles/-/40010',
'https://bunshun.jp//articles/-/40010?page=2'],
['https://bunshun.jp//articles/-/40843',
'https://bunshun.jp//articles/-/40843?page=2',
'https://bunshun.jp//articles/-/40843?page=3',
'https://bunshun.jp//articles/-/40843?page=4'],
['https://bunshun.jp//articles/-/40877',
'https://bunshun.jp//articles/-/40877?page=2'],
['https://bunshun.jp//articles/-/40868',
'https://bunshun.jp//articles/-/40868?page=2'],
['https://bunshun.jp//articles/-/40826',
'https://bunshun.jp//articles/-/40826?page=2',
'https://bunshun.jp//articles/-/40826?page=3',
'https://bunshun.jp//articles/-/40826?page=4'],
['https://bunshun.jp//articles/-/40752',
'https://bunshun.jp//articles/-/40752?page=2',
'https://bunshun.jp//articles/-/40752?page=3',
'https://bunshun.jp//articles/-/40752?page=4'],
['https://bunshun.jp//articles/-/40862',
'https://bunshun.jp//articles/-/40862?page=2',
'https://bunshun.jp//articles/-/40862?page=3'],
['https://bunshun.jp//articles/-/40841',
'https://bunshun.jp//articles/-/40841?page=2',
'https://bunshun.jp//articles/-/40841?page=3',
'https://bunshun.jp//articles/-/40841?page=4',
'https://bunshun.jp//articles/-/40841?page=5'],
['https://bunshun.jp//articles/-/40694',
'https://bunshun.jp//articles/-/40694?page=2',
'https://bunshun.jp//articles/-/40694?page=3',
'https://bunshun.jp//articles/-/40694?page=4']]
Now that you have created the URL list with the code above, follow the link to get the article body.
I can get only the text by applying .text
, but I am applying .text
by turning the for statement in detail.
Therefore, I created a list that stores the text while creating and storing some empty characters (or empty list).
news_page_list = [] #The text of all articles is stored here.
for news_links in news_list: #This for statement retrieves a linked list of a title from the list of URLs.
news_page = '' #We will add the text obtained from each page here.
for news_link in news_links: #Extract the links one by one from the link list in the title.
news_res = requests.get(news_link) # requests.get()And use Beautiful Soup to get article information.
news_soup = BeautifulSoup(news_res.text, "html.parser")
news_soup = news_soup.find(class_=re.compile("article-body")).find_all("p") # article-The body was stored in the p tag directly under the tag with id body.
news_phrase = '' #Stores the phrase in the body of the page
for news in news_soup: #I was able to get only the body phrase by applying text by turning it with a for statement.
news_phrase += news.text.replace('\u3000', ' ') #Add the acquired phrase. Because it is a character string+I was able to add it with.
news_page += news_phrase #If you can get one page of phrases, new_Add to page
news_page_list.append(news_page) #All text for one title is new_news when stored on page_page_Added to list. Since this is a list type, use append.
for i in range(1, 4): #Let's display a part of the acquired text. It seems that I was able to get it successfully.
print("<%s>" % i, news_page_list[i][:500], end="\n\n")
<1> Mitsui Home continues to be selected as a "land utilization partner" by many land owners. Through interviews with the company's sales staff, who are professionals in land utilization, this project will reveal the reasons why Mitsui Home is selected as a partner. This time, we will introduce an example of rebuilding a vast home site of about 1,200 m2 in Fuchu City, Tokyo, into a total of two buildings: a clinic + rental housing + home and a nursery school. We spoke with Toshito Nishijima, the head of the Tokyo West Area Sales Group, Tokyo Consulting Sales Department, who was in charge of this matter. About a 5-minute walk from the nearest station on the Keio Line, which extends west from the city center, was the home of the landowner, who had continued since before the war. The site area is approximately 1,200 m2. The 50-year-old home was aging and had to be considered for rebuilding. The owner in his 70s is a family of landowners who have continued to this area for generations, and own multiple rental condominiums around their homes. The beginning of this project was to make effective use of the vast site of about 1,200 m2, taking advantage of the rebuilding of a dilapidated home that was about 50 years old. It was early 2018 that the bank brought us a consultation. First, I met the owner with a plan of home + rental housing. Then oh
<2> Watami Co., Ltd. continues to be charged with labor issues. On October 2, the director of the "Watami's Home Meal" sales office announced a series of recommendations from the Labor Standards Inspection Office to correct unpaid overtime, long working hours exceeding 175 hours a month, and falsification of time cards by bosses. It is. Overtime work of 175 hours a month due to "white company" promotion Watami's correction recommendation from the Labor Bureau due to unpaid overtime fee Why did Watami not become a white company? As for the attendance "tampering" system, Mr. A lost his sense of day and night after working long hours, and even lived with fear that he would not wake up if he slept as it was. "If I worked as it was, I would have died," A asserts. Currently, he has a mental illness and is on leave while applying for an industrial accident. However, why did Mr. A continue to work hard while feeling the danger of his life? Behind this was Watami's system of "idea education" that worked on the consciousness of workers and made them accept harsh labor. "I'm doing such a good job, so I'll do my best even if it's painful." "It's not painful even if it's painful. Rather, it helps me." Mr. A told himself during overwork. In fact, A
<3> "Regulatory reform" "Administrative reform" "Vertical breakthrough". After the inauguration of Yoshihide Suga's administration, the word "reform" came to be heard well. At the inaugural press conference on September 16, Prime Minister Suga declared that "regulatory reform will be in the middle of the administration." How will Japan change as a result of this "reform"? Toru Hashimoto, who has a close relationship with Prime Minister Suga, talked about the goals of the Suga administration's "reform" in an interview with the November issue of "Bungeishunju." From his own experience, Mr. Hashimoto says that it is important to have a sense of "funny" in order to proceed with "reform." "For" reforming power, "it is extremely important to always keep an antenna on things around you, and if you feel that this is strange, immediately say it. Then, fix it each time. Even when I was the governor / mayor, it was a series of such tasks. For example, when I got in a public car, five newspapers were quickly inserted into the magazine rack. By the time I arrived at the government building. It's nice because you can check the news, but when you enter the governor's office, 5 papers are on the desk, and when you go to the governor's reception room, 5 papers again ... Isn't it? "" What's going on, this new
The information obtained by scraping up to now is stored in one DataFrame. This not only makes the data easier to see, but also easier to handle. If possible, all you have to do is process the data and perform negative / positive analysis!
new_no_list = [x for x in range(len(title_list))] #News No. I will use it later.Create
newslist = np.array([new_no_list, title_list, url_list, news_page_list]).T #Np in preparation for storage in DataFrame.Store it in the array list and transpose it.
newslist = pd.DataFrame(newslist, columns=['News No.', 'title', 'url', 'news_page_list']) #Store in DataFrame by specifying the column name
newslist = newslist.astype({'News No.':'int64'}) # あとでテーブルを結合するためにNews No.To int64 type
display(newslist)
News No. th> | title | url | news_page_list | |
---|---|---|---|---|
0 | 0 | Former Johnny's MADE leader Hikaru Inaba (29) and former Berryz Kobo idol "Shibuya Hotel Date ... td> | https://bunshun.jp//articles/-/40869 | Ryota Yamamoto (30) of the popular unit "Space Six" in Johnny's Jr. goes to an illegal dark slot shop ... td> |
1 | 1 | "A vast site of about 1,200 m2 near the station" "Land utilization that balances home rebuilding and community contribution" Professionals put out ... td> | https://bunshun.jp//articles/-/40010 | Mitsui Home continues to be selected as a "land utilization partner" by many land owners. Land utilization ... td> |
2 | 2 | "Let's work 24 hours a day, 365 days a year" ... Watami's "idea education" was still going on. Td> | https://bunshun.jp//articles/-/40843 | Watami Co., Ltd. continues to be charged with labor issues. October 2nd, Director of "Watami's Home Meal" Sales Office ... td> |
3 | 3 | It's a bad effect of "vertical administration"! Toru Hashimoto talks about "What's wrong with the Go To campaign?" Td> | https://bunshun.jp//articles/-/40877 | "Regulatory reform" "Administrative reform" "Vertical breakthrough". After the inauguration of Yoshihide Suga's administration, I often heard the word "reform" ... td> |
4 | 4 | Spitting police officer arrested for obstructing public duties Yomiuri Shimbun disciplines Seoul bureau reporter td> | https://bunshun.jp//articles/-/40868 | A reporter (34) from the Seoul branch of the Yomiuri Shimbun was arrested by South Korean authorities on suspicion of obstructing the execution of public affairs in mid-July ... td> |
5 | 5 | "It's a ridiculous misunderstanding" to the Adachi Ward Council of homosexual discrimination ... What my 81-year-old grandmother's letter told me td> | https://bunshun.jp//articles/-/40826 | "Grandma is angry about the Adachi Ward Assembly and seems to write a letter." This LINE came from my mother ... td> |
6 | 6 | Iekei Ramen "Goodwill War" "Yoshimuraya vs. Rokkakuya" Betrayal and Succumbing Black History https://bunshun.jp//articles/-/40752 |
"'Rokkakuya' went bankrupt, but the number of stores of'Iekei Ramen'as a genre is increasing year by year, and ... td>
| |
7 | 7 | << Hirate school and graduation >> Keyakizaka46, Shiori Sato wrote "Sad things in my activities" Sakurazaka46's "Sudden detention ... td> | https://bunshun.jp//articles/-/40862 | << Good evening everyone. Today, I have something to tell everyone who is always supporting me. I, Sato ... td> |
8 | 8 | Was the Abe administration's greatest achievement the "Ainu Museum"? The truth of "Upopoi" with 20 billion yen td> | https://bunshun.jp//articles/-/40841 | ──Is that really okay? On the way back, while renting a car on the rainy Hokkaido Expressway, I felt that way ... td> |
9 | 9 | Not "length" ... "Unexpected words" that hairdressers teach when cutting hair td> | https://bunshun.jp//articles/-/40694 | The comforter is getting more and more comfortable these days. Wearing a thin jacket for the sudden temperature difference between morning and night ... td> |
Use the "Word Emotion Polarity Correspondence Table" as the negative / positive judgment criteria. Download it to your working directory in advance. Prepare this "Word Emotion Polarity Correspondence Table" into a form for use in analysis.
p_dic = pathlib.Path('/work/dic') #Pass the path to the dic folder in the work directory. The file of "Word Emotion Polarity Correspondence Table" is placed here.
for i in p_dic.glob('*.txt'): #Find the file in question.
with open (i, 'r', encoding='utf-8') as f:
x = [i.replace('\n', '').split(':') for i in f.readlines()] #Read line by line.
posi_nega_df = pd.DataFrame(x, columns = ['Uninflected word', 'reading', 'Part of speech', 'Score']) # reading込んだデータをDataFrameに格納します。
posi_nega_df['reading'] = posi_nega_df['reading'].apply(lambda x : jaconv.hira2kata(x)) #Convert Hiragana to Katakana(同じreadingのものが含まれており、重複を無くす為のようです。)
posi_nega_df = posi_nega_df[~posi_nega_df[['Uninflected word', 'reading', 'Part of speech']].duplicated()] #Remove duplicates.
posi_nega_df.head()
Basic form th> | Reading th> | Part of speech th> | Score th> | |
---|---|---|---|---|
0 | excellent td> | Sugurel td> | verb td> | 1 |
1 | good td> | Yoi td> | Adjectives td> | 0.999995 |
2 | Rejoice td> | Yorokobu td> | verb td> | 0.999979 |
3 | Compliment td> | Homel td> | verb td> | 0.999979 |
4 | Congratulations td> | Medetai td> | Adjectives td> | 0.999645 |
Morphological analysis of the article text is made into a form that can be used for analysis.
Use Tokenizer ()
and ʻUnicodeNormalizeCharFilter ()` for morphological analysis.
Extract words, uninflected words, part of speech, and readings and store them in a DataFrame.
Then, merge the article DataFrame with the "Word Emotion Polarity Correspondence Table" to score the words contained in the article.
The table is shown below.
The word "popular" has a high score and was judged to be a positive word.
Why did the other words get that score? There is also something like that, but let's proceed without worrying about it.
i = 0 #This i is news No.It is used when acquiring.
t = Tokenizer()
char_filters = [UnicodeNormalizeCharFilter()]
analyzer = Analyzer(char_filters=char_filters, tokenizer=t)
word_lists = []
for i, row in newslist.iterrows(): #Increase i one by one News No.will do.
for t in analyzer.analyze(row[3]): #The text is stored in the third column of the extracted label.
surf = t.surface #word
base = t.base_form #Uninflected word
pos = t.part_of_speech #Part of speech
reading = t.reading #reading
word_lists.append([i, surf, base, pos, reading]) # word_Add to lists
word_df = pd.DataFrame(word_lists, columns=['News No.', 'word', 'Uninflected word', 'Part of speech', 'reading'])
word_df['Part of speech'] = word_df['Part of speech'].apply(lambda x : x.split(',')[0]) # Part of speechは複数格納されるが最初の1つのみ利用
display(word_df.head(10)) #Display the created text table
print("↓ ↓ ↓ ↓ ↓ ↓ ↓ Merge with word emotion polarity correspondence table ↓ ↓ ↓ ↓ ↓ ↓ ↓")
score_result = pd.merge(word_df, posi_nega_df, on=['Uninflected word', 'Part of speech', 'reading'], how='left') #Merge text table and word emotion polarity correspondence table
display(score_result.head(10)) #Display the created score table. I understand that the score of the word "popularity" is high, but the others are subtle ...
News No. th> | word th> | Basic form th> | Part of speech th> | Reading th> | |
---|---|---|---|---|---|
0 | 0 | Ja td> | Ja td> | noun td> | Ja td> |
1 | 0 | Needs td> | Needs td> | noun td> | Needs td> |
2 | 0 | Jr | Jr | noun td> | * |
3 | 0 | . | . | noun td> | * |
4 | 0 | Insidetd> Inside | td> | noun td> | Nai td> |
5 | 0 | td> | td> | Particles td> | ノ td> |
6 | 0 | Popular td> | Popular td> | noun td> | Ninki td> |
7 | 0 | unit td> | unit td> | noun td> | unit td> |
8 | 0 | 「 | 「 | symbol td> | 「 |
9 | 0 | Universe td> | Universe td> | noun td> | Uchu td> |
↓ ↓ ↓ ↓ ↓ ↓ ↓ Merge with word emotion polarity correspondence table ↓ ↓ ↓ ↓ ↓ ↓ ↓
News No. th> | word th> | Basic form th> | Part of speech th> | Reading th> | Score th> | |
---|---|---|---|---|---|---|
0 | 0 | Ja td> | Ja td> | noun td> | Ja td> | NaN |
1 | 0 | Needs td> | Needs td> | noun td> | Needs td> | -0.163536 |
2 | 0 | Jr | Jr | noun td> | * | NaN |
3 | 0 | . | . | noun td> | * | NaN |
4 | 0 | Insidetd> Inside | td> | noun td> | Nai td> | -0.74522 |
5 | 0 | td> | td> | Particles td> | ノ td> | NaN |
6 | 0 | Popular td> | Popular td> | noun td> | Ninki td> | 0.96765 |
7 | 0 | unit td> | unit td> | noun td> | unit td> | -0.155284 |
8 | 0 | 「 | 「 | symbol td> | 「 | NaN |
9 | 0 | Universe td> | Universe td> | noun td> | Uchu td> | -0.515475 |
Evaluate the negative / positive degree of the entire article using the table created earlier.
result = []
for i in range(len(score_result['News No.'].unique())): # News No.Use to turn the for statement.
temp_df = score_result[score_result['News No.']== i]
text = ''.join(list(temp_df['word'])) # 1タイトル内の全てのwordをつなげる。
score = temp_df['Score'].astype(float).sum() # 1タイトル内のScoreを全て足し合わせる。➡︎累計Score
score_r = score/temp_df['Score'].astype(float).count() # 本文の長さに影響されないように単語数で割り算する。➡︎標準化Score
result.append([i, text, score, score_r])
ranking = pd.DataFrame(result, columns=['News No.', 'text', 'Cumulative score', 'Standardized score']).sort_values(by='Standardized score', ascending=False).reset_index(drop=True) # Standardized scoreで並び替えてDataFrameに格納
ranking = pd.merge(ranking, newslist[['News No.', 'title', 'url']], on='News No.', how='left') # News No.Merge by criteria. Add a title and URL.
ranking = ranking.reindex(columns=['News No.', 'title', 'url', 'text', 'Cumulative score', 'Standardized score']) #Sort columns
display(ranking)
News No. th> | title | url | text th> | Cumulative score th> | Standardized score th> | |
---|---|---|---|---|---|---|
0 | 6 | Iekei Ramen "Goodwill War" "Yoshimuraya vs. Rokkakuya" Betrayal and Succumbing Black History https://bunshun.jp//articles/-/40752 |
"'Rokkakuya' went bankrupt, but the number of stores of'Iekei Ramen'as a genre is increasing year by year, and ... td>
| -238.437124 |
-0.408983 |
|
1 | 1 | "A vast site of about 1,200 m2 near the station" "Land utilization that balances home rebuilding and community contribution" Professionals put out ... td> | https://bunshun.jp//articles/-/40010 | Mitsui Home continues to be selected as a "land utilization partner" by many land owners. Land utilization ... td> | -315.299051 | -0.438524 |
2 | 7 | << Hirate school and graduation >> Keyakizaka46, Shiori Sato wrote "Sad things in my activities" Sakurazaka46's "Sudden detention ... td> | https://bunshun.jp//articles/-/40862 | << Good evening everyone. Today, I have something to tell everyone who is always supporting me. I, Sato ... td> | -136.887378 | -0.447344 |
3 | 5 | To the Adachi Ward Council of homosexual discrimination, "It's a ridiculous misunderstanding." | https://bunshun.jp//articles/-/40826 | "Grandma is angry about the Adachi Ward Assembly and seems to write a letter." My mother sent me this LINE ... td> | -213.244051 | -0.460570 |
4 | 9 | Not "length" ... "Unexpected words" that hairdressers teach when cutting hair td> | https://bunshun.jp//articles/-/40694 | The comforter is getting more and more comfortable these days. I wore a thin coat due to the sudden temperature difference between morning and night ... td> | -192.702889 | -0.475810 |
5 | 8 | Was the Abe administration's greatest achievement the "Ainu Museum"? The truth of "Upopoi" with 20 billion yen td> | https://bunshun.jp//articles/-/40841 | ──Is that really okay? On the way back, while renting a car on the rainy Hokkaido Expressway, I feel like that ... td> | -483.393151 | -0.476719 |
6 | 0 | Former Johnny's MADE leader Hikaru Inaba (29) and former Berryz Kobo idol "Shibuya Hotel Date ... td> | https://bunshun.jp//articles/-/40869 | Ryota Yamamoto (30) of the popular unit "Space Six" in Johnny's Jr. goes to an illegal dark slot shop ... td> | -196.888853 | -0.479048 |
7 | 3 | It's a bad effect of "vertical administration"! Toru Hashimoto talks about "What's wrong with the Go To campaign?" Td> | https://bunshun.jp//articles/-/40877 | "Regulatory reform" "Administrative reform" "Vertical breakthrough". After the inauguration of Yoshihide Suga's administration, I often heard the word "reform" ... td> | -94.718989 | -0.480807 |
8 | 4 | Spitting police officer arrested for obstructing public duties Yomiuri Shimbun disciplines Seoul bureau reporter td> | https://bunshun.jp//articles/-/40868 | A reporter (34) from the Yomiuri Shimbun Seoul Bureau was arrested by South Korean authorities on suspicion of obstructing the execution of public affairs in mid-July ... td> | -144.916148 | -0.489582 |
9 | 2 | "Let's work 24 hours a day, 365 days a year" ... Watami's "idea education" was still going on. Td> | https://bunshun.jp//articles/-/40843 | Watami Co., Ltd. continues to be charged with labor issues. On October 2nd, the director of the "Watami's Home Meal" sales office ... td> | -321.838102 | -0.528470 |
Let's display the most positive articles and the most negative articles.
print("<<Positive 1st place>>", end="\n\n")
for i in range(1, 4):
print(ranking.iloc[0, i])
<< Positive 1st place >>
Family ramen "Goodwill War" "Yoshimuraya vs. Rokkakuya" Betrayal and succumbing black history
print("<<Negative 1st place>>", end="\n\n")
for i in range(1, 4):
print(ranking.iloc[-1, i])
<< Negative 1st place >>
"Let's work 24 hours a day, 365 days a year" ... Watami's "idea education" was still going on
https://bunshun.jp//articles/-/40843
Watami Co., Ltd. continues to be charged with labor issues. On October 2, the director of the "Watami's Home Meal" sales office announced a series of recommendations from the Labor Standards Inspection Office to correct unpaid overtime, long working hours exceeding 175 hours a month, and falsification of time cards by bosses. It is. Overtime work of 175 hours a month due to Watami of "white company" promotion Recommendation from Labor Bureau for correction due to unpaid overtime fee Why did Watami not become a white company? Attendance "tampering" system without permission Mr. A also worked long hours, day and night I lost that feeling, and even lived with fear, "If I sleep like this, I may not wake up anymore." "If I worked as it was, I would have died," A asserts. Currently, he has a mental illness and is on leave while applying for an industrial accident. However, why did Mr. A continue to work hard while feeling the danger of his life? Behind this was Watami's system of "idea education" that worked on the consciousness of workers and made them accept harsh labor. "I'm doing such a good job, so I'll do my best even if it's painful." "It's not painful even if it's painful. Rather, it helps me." Mr. A told himself during overwork. In fact, Mr. A was "pride" in Watami's home-cooking work. Certainly, Watami's "home-cooked meal" has a part that can be called "social contribution." Watami's home-cooked meals have the concept of regularly delivering cheap meals to elderly people who have difficulty procuring meals. The delivery destinations of Mr. A's sales office were all such people. Elderly people living alone who go to day service every other day and have trouble eating when they are at home every other day. Elderly people, people with disabilities, and people who care for their parents who ask for two meals, one for lunch and one for dinner, and eat only Watami's home meal. Due to the difficult living conditions, there were various users who could not spend time, time and money on meals and had to rely on Watami. For Mr. A, who had worked in the field of long-term care and education until then, the home-cooking job that benefits society and the community was very rewarding. However, if they try to make a profit through such "support", they will be supported by the sacrifice of workers due to long working hours and low wages. It can be called a "poverty / black company business". At first, Mr. A calmly thought that job satisfaction and harsh labor should be separated, and was dissatisfied with Watami about the difficulty of work. However, poor working conditions were no longer a problem in Mr. A's mind. Mr. A recalls that it was Watami's "idea education" that brought about that change. "Work 24 hours a day, 365 days a year until you die." Many of you may know this phrase. It is a word that was written in a 400-page book called "Philosophy Collection", which was edited by excerpting Miki Watanabe's texts over the past 30 years. Immediately after joining the company, Mr. A was given a "philosophy collection" by the company and was told to carry it with him. Now, under criticism, extreme expressions such as "work until death" have been deleted. Still, let's quote some of the impressive expressions that remained undeleted in the "Philosophy Collection" (2016 edition) handed to Mr. A. <I don't think work is just a way to make money. I believe that work is the person's "way of life" itself and "the only means of self-actualization." That's why even at seminars for new graduates, while being told that it's out of date, he says, "Let's work 24 hours a day, 365 days a year."> <I also at a company information session, "Work is life itself. Don't use any means to do this. Let's improve our humanity through work. ”> <Watami Takushoku's greatest product is known as" people. " Magokoro-san (Note: Home-meal delivery person) is a person who carries a lunch box with a "heart" and receives "thank you". I pray that no one will misunderstand and think that it is a job to carry a "lunch box" and receive "money"> In this way, the book encourages working to thank customers for "thank you". Then, criticism of working for wages and justification for sacrificing oneself for the benefit of Watami were written everywhere. When Mr. A, who joined Watami, had any trouble, he was asked by the branch president, "Did you read Chapter x of the Philosophy Collection properly?", Because he did not fully understand Miki Watanabe's "idea." It was noted that. In addition, there is monthly counseling by the area manager, where the impressions of this month's "in-house newsletter" (published monthly and many of the texts of the philosophy collection are excerpts from here) and impressions of any part of the philosophy collection. Was to be written. In addition, a report was imposed once every four months. Two impressions were forced on the scope of the chapter specified from the collection of philosophies and the words written by Mr. Watanabe in the company newsletter for the last four months. Originally, Mr. A felt that submitting these impressions was "unpleasant." However, unwillingly, while it was being written continuously, he said, "I felt that Miki Watanabe's ideas were being planted somewhere." It was the existence of the "video letter" that deepened that. Every month, a "video letter" starring Miki Watanabe was provided to the sales office. The narrator of the popular TV program "Jonetsu Tairiku" is appointed, and Miki Watanabe appears every time, and it is a 30-minute video that explains the splendor of Watami's business. Monthly impressions were also required for this video. Moreover, it is not only the director who writes. Even the deliveryman, who was supposed to be a sole proprietor, was included in the contract to watch a video letter every month and write his impressions. When the delivery staff are shown this video at the sales office, they write their impressions by hand in their own impression column (there is a space for writing about 60 characters) on the sheet where the names of the director and the delivery staff are listed. .. The delivery staff only pays a hundred and several tens of yen per delivery destination, but there is no new reward for watching this video or writing their impressions. Furthermore, if the deliveryman's impressions included criticism of Watami or dissatisfaction with the work, the director was instructed to mark the part and add a comment. As an original device, Mr. A politely commented on the impressions of all the delivery staff with a number of characters that exceeded the impressions of the delivery staff. And it is said that this "comment" played a major role in Mr. A's "idea formation." Mr. A was told by his boss's area manager, "It is also the director's job to" take "the delivery staff to write impressions that praise Watami." They are trying to "control" the impressions of the delivery staff. However, Mr. A did not "tamper" with the impressions of the delivery staff. First, as the director, Mr. A explained in his impressions column prior to the delivery staff the wonders of Watami's business and working at Watami. Then, the delivery staff who write later will naturally be aware of the director's "model answer" and it will be difficult to write negative impressions. Even so, if there was a part of the deliveryman's impression that questioned Watami, he pointed it out in the comment section and said, "Let's share the significance with me again." Over the course of several months, the negative texts disappeared from the comments of the delivery staff, and at least ostensibly, the impressions of praising Watami began to grow. This monthly comment, which repeats the splendor of Watami, has steadily influenced "idea formation." However, what really changed was not the deliveryman, but Mr. A, the director. Mr. A, who should have dared to continue to "praise" Watami because of the necessity of his work, gradually praised Watami's business and labor unconditionally, and he said that he really became aware that he would not feel dissatisfied with labor problems. Through the act of instructing the impressions of the delivery staff many times, Mr. A's own consciousness was "educated". Here, let's quote the impression of the video letter that Mr. A wrote to show to the delivery person. It is an impression that I saw the video of Watami's SDGs efforts and Watami's efforts to support schools in Cambodia.
I feel like I've been able to evaluate it correctly!
That's all for the explanation. Thank you very much for reading.
Recommended Posts