Conversation

JSON Baumgartner

@jasonbaumgartne

Sentiment Lexicons -- Stanford's "Socialsent" scores all adjectives based on their sentiment values: nlp.stanford.edu/projects/socia

3

9

24

Jacobus Vanderwilt

Replying to

@jasonbaumgartne

I propose to use grammatical categories for sentiment analysis and use adjective meanings as secondary information. SUBJECT VERB OBJECT triples improve Text Mining.

Linguistics improves statistical classification with KLD, NB, TF*IDF, K-NN: the positive effects of...

Four statistical classification algorithms are executed with and without Stanford Parser-derived SVO triples. The output is improved in a major way. Thus this is an important method of feature...

1

1

JSON Baumgartner

@jasonbaumgartne

Replying to

This is awesome. Thanks for sharing this!

12:19 PM · May 30, 2018Twitter Web Client

Jacobus Vanderwilt

Replying to

@jasonbaumgartne

Thank you, sir

1