In this Lab we were asked to use the Topic Modeling Tool that is a tool of MALLET. These are the souvenirs of playing around with this program.
In this first picture I used my small corpus from that I compiled from Lab 3. I changed the number of topics to 15 and these were the topics that appeared from my corpus. I feel like these topics makes sense because most of my novels surrounded women, kings, the ideas of good and ladies.
Next I reduced my number of topics to 5 in the hopes that it would narrow down the topics in my corpus. The topics that it gave did not make sense compared to the number of topics of 15. This showed me that the small the topics the more random your topics may be.
Then I went into the advanced setting and played around with the number of iterations and I made it 150 and then also made the number of topics 15 again. The lists or words changed and then there were some words that did not even make sense when I changed the number iterations. This is the point of this souvenir.
This is the output of topics for when I did 10 topics and then didn’t change anything in the advanced section. This can be found in my desktop and it just is a good tool to find the topics in the document.
Next, this is an example of when I click on one of the topics and then it shows me in what texts the topics appears in the order that shows the most significance. This just helps when trying to find exactly where the topics are and how they are in context.
Last, is the example of the txt file that it opens when clicking on the 1st topic of significance. Then you are given the top documents in the txt file.