Gensim LDA 提供主题 ID 的输出，但概率加起来不等于 1

Question

我已经训练了 LDA 模型来聚类 100 个主题，据我所知，每个主题都应该以一定的概率输出，加起来等于 1。

但是当我运行这段代码时，我只得到了 2 个主题。

请帮忙。

text = "A blood cell, also called a hematocyte, is a cell produced by hematopoiesis and normally found in blood."

# transform text into the bag-of-words space
bow_vector = dictionary.doc2bow(tokenize(text))
lda_vector = lda_model[bow_vector]
print("LDA Output: ", lda_vector)
print("\nTop Keywords from highest prob Topic: ",lda_model.print_topic(max(lda_vector, key=lambda item: item[1])[0]))
print("\n\nAddition of all the probabilities from LDA output:",functools.reduce(lambda x,y:x+y,[i[1] for i in lda_vector]))

LDA 输出：[(64, 0.6952628), (69, 0.18223721)]

来自最高概率主题的热门关键字：0.042*“健康”+ 0.032*“医疗”+ 0.017*“患者”+ 0.016*“癌症”+ 0.015*“医院”+ 0.015*“说”+ 0.015*“治疗” " + 0.012*"医生" + 0.012*"护理" + 0.012*"药物"

LDA 输出的所有概率相加：0.8775

Answer 1

如果将LdaModel 的参数minimum_probability 设置为0，则总和将为1（或接近1，由于近似错误）。它控制过滤为文档返回的主题。