LDA results (number of topics and words per topic)

ayaRizk

Hi, I am trying to extract topics from files and have set the number of topics to be extracted to 50 and the top words per topic to 20 (defaults are 10 and 5 respectively). But when I run the process I get about 90 topics (that also varies with different runs) and more than 50 words per topic.

I am not sure what I'm doing wrong. Attaching the process here - appreciate the help!

Edit: Now I realized the source of the problem. The parent process "Optimize parameters" has option to set max number of topics and words per topic as the chosen parameters. I was setting their values in the LDA operator and forgetting the ones in the parent operator, which seems to override the LDA values.

Regards,
Aya

<参数键= value = " /“文件名”Users/ayari88/Documents/Research/AFA/ROBOT/RapidMiner/sv_stopwords_418.csv"/>

<参数键= "从文档中提取主题(LDA).number_of_topics" value="[1.0;100.0;10;linear]"/>

<参数键= "从文档中提取主题(LDA).iterations" value="[1.0;100.0;10;linear]"/>

<参数键= "从文档中提取主题(LDA).top_words_per_topic" value="[1.0;100.0;10;linear]"/>

<连接from_op = "优化参数(网格)”_port="performance" to_port="result 1"/>

<连接from_op = "优化参数(网格)”_port="parameter set" to_port="result 2"/>

<连接from_op = "优化参数(网格)”_port="output 1" to_port="result 3"/>

<连接from_op = "优化参数(网格)”_port="output 2" to_port="result 4"/>