问题描述
使用横向狄利克雷分配确定主题
tokenized = tweets['initial']
dictionary = corpora.Dictionary(tokenized)
dictionary.filter_extremes(no_below=1,no_above=0.8)
corpus = [dictionary.doc2bow(tokens) for tokens in tokenized]
ldamodel = gensim.models.ldamodel.Ldamodel(corpus,num_topics = 5,id2word=dictionary,passes=15)
ldamodel.save('mOdel.gensim')
topics = ldamodel.print_topics(num_words=10)
for topic in topics:
print(topic)
You can see output of this block from here
显示哪个主题代表推文的数据框dataframe that should do that but not
def dominant_topic(ldamodel,corpus,content):
#Function to find the dominant topic in each query
sent_topics_df = pd.DataFrame()
# Get main topic in each query
for i,row in enumerate(ldamodel[corpus]):
row = sorted(row,key=lambda x: (x[1]),reverse=True)
# Get the Dominant topic,Perc Contribution and Keywords for each query
for j,(topic_num,prop_topic) in enumerate(row):
if j == 0: # => dominant topic
wp = ldamodel.show_topic(topic_num,topn=20)
topic_keywords = ",".join([word for word,prop in wp])
sent_topics_df = sent_topics_df.append(pd.Series([int(topic_num),round(prop_topic,4),topic_keywords]),ignore_index=True)
else:
break
sent_topics_df.columns = ['Dominant_Topic','Perc_Contribution','Topic_Keywords']
contents = pd.Series(content)#noisy data
sent_topics_df = pd.concat([sent_topics_df,contents],axis=1)
return(sent_topics_df)
df_dominant_topic = dominant_topic(ldamodel=ldamodel,corpus=corpus,content=tweets['initial'])
df_dominant_topic.head(5)
我只能看到第一个。这就是问题所在。
解决方法
暂无找到可以解决该程序问题的有效方法,小编努力寻找整理中!
如果你已经找到好的解决方法,欢迎将解决方案带上本链接一起发送给小编。
小编邮箱:dio#foxmail.com (将#修改为@)