提取文档的主题分数 LDA Gensim Python

Question

在使用 LDA 模型后，我正在尝试为我的数据集中的文档提取主题分数。具体来说，我遵循了这里的大部分代码：https://www.machinelearningplus.com/nlp/topic-modeling-gensim-python/

我已经完成了主题模型并得到了我想要的结果，但是提供的代码只给出了每个文档的最主要主题。有没有一种简单的方法来修改以下代码，以便给我说 5 个最主要的主题的分数？

##dominant topic for each document
def format_topics_sentences(ldamodel=optimal_model, corpus=corpus, texts=data):
    # Init output
    sent_topics_df = pd.DataFrame()

# Get main topic in each document
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 document
    for j, (topic_num, prop_topic) in enumerate(row):
        if j == 0:  # => dominant topic
            wp = ldamodel.show_topic(topic_num)
            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']

# Add original text to the end of the output
contents = pd.Series(texts)
sent_topics_df = pd.concat([sent_topics_df, contents], axis=1)
return(sent_topics_df)


df_topic_sents_keywords = format_topics_sentences(ldamodel=optimal_model, corpus=corpus, texts=data)

# Format
df_dominant_topic = df_topic_sents_keywords.reset_index()
df_dominant_topic.columns = ['Document_No', 'Dominant_Topic', 'Topic_Perc_Contrib', 'Keywords', 'Text']

# Show
df_dominant_topic.head(10)

Answer 1

这是一个粗俗的例子，因为你没有提供数据来重现，但我们可以使用一些 gensim 测试语料库、文本和字典：

from gensim.test.utils import common_texts, common_corpus, common_dictionary
from gensim.models import LdaModel

# train a quick lda model using the common _corpus, _dictionary and _texts from gensim
optimal_model = LdaModel(common_corpus, id2word=common_dictionary, num_topics=10)

然后我们可以将函数稍微改写成：

import pandas as pd

##dominant topic for each document
def format_topics_sentences(ldamodel=optimal_model, 
                            corpus=common_corpus, 
                            texts=common_texts, 
                            n=1):
    """
    A function for extracting a number of dominant topics for a given document
    using an existing LDA model
    """
    # Init output
    sent_topics_df = pd.DataFrame()


    # Get main topic in each document
    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 document
        for j, (topic_num, prop_topic) in enumerate(row):
            # we use range here to iterate over the n parameter
            if j in range(n):  # => dominant topic
                wp = ldamodel.show_topic(topic_num)
                topic_keywords = ", ".join([word for word, prop in wp])
                sent_topics_df = sent_topics_df.append(
                    # and also use the i value here to get the document label
                    pd.Series([int(i), int(topic_num), round(prop_topic, 4), topic_keywords]),
                    ignore_index=True,
                )
            else:
                break
    sent_topics_df.columns = ["Document", "Dominant_Topic", "Perc_Contribution", "Topic_Keywords"]

    # Add original text to the end of the output
    text_col = [texts[int(i)] for i in sent_topics_df.Document.tolist()]
    contents = pd.Series(text_col, name='original_texts')
    sent_topics_df = pd.concat([sent_topics_df, contents], axis=1)
    return sent_topics_df

那么我们可以这样使用函数：

format_topics_sentences(ldamodel=optimal_model, corpus=common_corpus, texts=common_texts, n=2)

n 参数指定要提取的主导主题的数量。