tweet classification how to identify type of conversation [closed]

Question

Closed. This question is off-topic. It is not currently accepting answers.

---

Want to improve this question? Update the question so it's on-topic for Stack Overflow.

Closed 5 years ago.

I've been reading papers on tweet classification and sentiment analysis. So far all of it was about positive and negative classification. What about if you want to identify the kind of communication or tweet it is (inquiry, suggestion, news, etc.)? Also, would naive Bayes be possible to use for it?

Answer 1

In Sentiment Analysis and Opinion Mining (http://www.amazon.com/gp/product/B009KET3PU) there are some chapters on spam detection in online reviews. I imagine some of those techniques would apply very well.

For learning features, beyond just the words and punctuation and sentence patterns, the number of retweets might be useful (questions don't get retweeted as much as news, for instance). (Also what people write with the retweet: "This is great: ..." is more likely to be news, whereas "Anyone know?" is likely to be a question, whereas "What do you think of this?" or "I'd never thought of this!" might introduce a tip or suggestion.)