0

I'm trying to do Twitter sentiment analysis on Chinese, French, Norwegian, Russian, Persian, Hebrew, German, Hindi, Indonesian, Japanese, Korean, Portuguese, Turkish and Arabic. Just wanna ask if there are any packages for these languages and in general what languages have supporting packages to do Twitter sentiment analysis in? The programming language doesn't matter. Thx!!!

Ham Huang
  • 1
  • 1
  • https://boostlog.io/@anshulc95/twitter-sentiment-analysis-using-nodejs-5ad1331247018500491f3b6a – JeffProd Jul 13 '18 at 12:00
  • I think you could google "twitter sentiment analysis packages", or something like that, then read some blogs, do some experimentation, and then come back to StackOverflow with more specific questions. (It is discouraged to ask for software/package recommendations.) – Darren Cook Jul 13 '18 at 17:24

1 Answers1

0

From my own experience, the best way to use sentiment classifiers on twitter is to build your own bag of words in the language of your choice.

I recommend using Vader, a python tool for lexicon-based analysis specifically designed for social media. However, Vader is available only in English. In that case, what you should do is obtain lists of words in other languages and translate them using the Google API. Then you employ Vader.

You can check this page for details: https://towardsdatascience.com/almost-real-time-twitter-sentiment-analysis-with-tweep-vader-f88ed5b93b1c

This is not the most appropriate procedure. It is best to focus on one or two languages because there are cultural differences that will reduce the accuracy of the algorithm.

I hope you find it helpful. Greeting.

aKratos
  • 191
  • 10