如何对excel文件中包含的文本进行聚类？答案

【问题标题】：How to perform clustering on text contained in an excel file?如何对excel文件中包含的文本进行聚类？
【发布时间】：2019-12-09 14:16:22
【问题描述】：

我正在尝试使用 excel 文件中包含的文本创建集群，但我收到错误“AttributeError: 'int' object has no attribute 'lower'”。

Sample.xlsx 是一个包含如下数据的文件：

我创建了一个名为 corpus 的列表，它根据每一行都有唯一的文本，我在对语料库进行矢量化时遇到了这个问题。

'''蟒蛇

import pandas as pd
import numpy as np


data=pd.read_excel('sample.xlsx') 
idea=data.iloc[:,0:1] #Selecting the first column that has text. 

#Converting the column of data from excel sheet into a list of documents, where each document corresponds to a group of sentences.
corpus=[]
for index,row in idea.iterrows():
    corpus.append(row['_index_text_data']) 

#Count Vectoriser then tidf transformer

from sklearn.feature_extraction.text import CountVectorizer

vectorizer = CountVectorizer()
X = vectorizer.fit_transform(corpus) #ERROR AFTER EXECUTING THESE #LINES

#vectorizer.get_feature_names()

#print(X.toarray())     

from sklearn.feature_extraction.text import TfidfTransformer

transformer = TfidfTransformer(smooth_idf=False)
tfidf = transformer.fit_transform(X)
print(tfidf.shape )                        

from sklearn.cluster import KMeans

num_clusters = 5 #Change it according to your data.
km = KMeans(n_clusters=num_clusters)
km.fit(tfidf)
clusters = km.labels_.tolist()

idea={'Idea':corpus, 'Cluster':clusters} #Creating dict having doc with the corresponding cluster number.
frame=pd.DataFrame(idea,index=[clusters], columns=['Idea','Cluster']) # Converting it into a dataframe.

print("\n")
print(frame) #Print the doc with the labeled cluster number.
print("\n")
print(frame['Cluster'].value_counts()) #Print the counts of doc belonging `#to each cluster.`

预期结果：

错误：“AttributeError: 'int' 对象没有属性 'lower'”

【问题讨论】：

标签： python python-3.x nlp cluster-analysis

【解决方案1】：

如果有人正在寻找这个问题的答案，那么只需在 for 循环之后在上面的代码中使用 '''corpus = [str (item) for item in corpus]''' 将整个语料库转换为文本。

新代码：

corpus=[]
for index,row in idea.iterrows():
    corpus.append(row['_index_text_data'])
    corpus = [str (item) for item in corpus]

【讨论】：