Sentiment analysis

Jun 9, 2016 at 7:17 PM

Sorry if this question has an easy to find answer but I searched the help documentation and could not find anything. I am confused by the sentiment analysis function, where do I find the list of words that are being used to calculate this measure? And how do I edit the list?

Many thanks,

Jun 10, 2016 at 6:10 PM

Thank you for the interest in NodeXL!You can find the features for text analysis in this NodeXL menu:

NodeXL>Analysis>Graph Metrics>Words and Word Pairs>Options>Sentiment:

There you will see the Word and Word Pair Metrics dialog:


This dialog allows you to edit the lists of words that are to be skipped or categorized as List #1, #2, or #3.

NodeXL ships with a default collection of American English language "Positive", "Negative", and "Angry/Violent" terms.

You may modify or replace these lists.

You may RENAME these lists as well.

You may use any language to populate these lists.

These lists do not need to be "Positive" and "Negative" words. Consider student versus teacher words, or investor versus consumer words, or even lists of words that are distinctive of different languages.


Marked as answer by MarcSmith on 7/5/2016 at 7:05 PM
Jun 11, 2016 at 12:49 PM
Great this helps a lot thanks :)
May 16, 2017 at 5:56 PM
Hi, I am running a sentiment analysis on a dataset of about 24k tweets that I scraped with TAGS. Results of positive/negative word counts and percentages were only reported for about 50 tweets. The cells for the rest were blank. Do you know why? It seems Tweets starting with @ or RT were not calculated. But there are some "regular" tweets not calculated either.

Thank you very much!

May 16, 2017 at 5:57 PM
Never mind. Sorry. My pro version worked fine. It seems the basic version doesn't do the job. Thanks!