log inhelp

 | 

  • If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • You already know Dokkio is an AI-powered assistant to organize & manage your digital files & messages. Very soon, Dokkio will support Outlook as well as One Drive. Check it out today!

View
 

Inez - Topic Modeling Test Page

Page history last edited by Inez 10 years, 6 months ago

With the Java TopicModelingTool, I tried to process my own fifty-page paper on Elizabeth Bishop, as I think the topic modeling might make more sense to be used on an article than a literary text, and it's a text file on hand. But I'm also curious to see how the topics written by me would be read by the machine. It turned out that I do like the result, which looks beautifully Bishop-esque.

 

List of Topics

1. poem crusoe goat island dream dead thought lines expressed living
2. eye eyes gaze tear gazes water idea glass derrida reflection
3. darwin animal human world reading part read encounter awareness letter
4. mouse vision ibid things natural species similar jarrell strange scene
5. animals moment life observation poetic description nature earlier story cage
6. print elizabeth university prose press made robert ed york letters
7. man moth solitude face small caught lonely art half young
8. fish death hand isinglass air experience make contrast image descriptions
9. bishop word inside unexpected words minute sudden wrote famous encounters
10. moore poems time poetry poet feeling back works long years

 

 

When it comes to a larger archive, this process could be, of course, much more telling. So trying with PMLA is very interesting. I searched with key word to find my interested topics. Those searches that yielded no results can tell as much as those are in the pool. For example, "english," "american," "french" and "latin" are among many topics, "german," "italian," "indian" got into two topics, "chinese," "spanish" got one, "portuguese" and "arabic" got none. "queer" got one, "garbage" got none. "shakespeare" got one, "milton" got none. "media" got one, "digital" got none. etc.

 

I found this lovely topic that makes me wonder at most:

 

14 irish sonnet sonnets yeats ireland defoe pound robinson crusoe dublin bloom plague epiphany celtic thorpe

 

And to look at the yearly proportion of this gender and sexuality topic clearly shows the development and trend of this study:

http://agoldst.github.io/dfr-browser/demo/#/topic/18

Comments (0)

You don't have permission to comment on this page.

Printable version