How can I pick out some positive terms from my term list?

[utincas]

Hello all, 

I'm new to Knime, and I've some difficulties in handling it.

I have grabbed some comments for a product from one on-line store. And I want to see if the customers like this product or not, so I am going to look for some positive terms appearing in the comments.

and now I have a terms list like this, but I don't know how/which nodes should I use to separate these terms into a positive group and a negative group. I want some advice for this, thanks!

[Iris]

Here is a recent blog post about sentiment analysis: https://www.knime.org/blog/sentiment-analysis which should help you get started.

I will also move your blog post to the text processing forum, where the experts are :)

[kilian.thiel]

Hi utincas,

what you need is a list with positive and negative words. When you have these lists / dictionaries you can use the Dictionary Tagger node to find and tag these words inside your documents. Use the Dictionary Tagger node directly after e.g. the Strings to Document node (or any parser node). In the Dictionary Tagger node you can specify with which tag the words of the dictionaries should be tagged e.g. with tag POSITIVE or NEGATIVE of the Sentiment tag set.

Once the terms are tagged you can create a bag of words and count them or filter all terms except the tagged ones.

For dictionaries for English language see MPQA corpus and the subjectivity lexicon:

http://mpqa.cs.pitt.edu/

http://mpqa.cs.pitt.edu/lexicons/subj_lexicon/

Cheers, Kilian