Challenges for sentiment analysis

1. Named Entity Recognition: what is actually mentioned by people. For example, Apple is an indispensable good thing in life. Is Apple a technology product or fruit?

2. Resolution of ing: solves the problem of reference to pronouns and noun phrases. For example, we had dinner after watching the movie. Does it mean that movie or dinner is uncomfortable?

3. syntax analysis: What is the subject and object in a sentence, and which of the following is the subject or object in the actual situation of its verb or adjective )?

4. ironic expression: if you are not familiar with the tabulation, you do not know whether it is true or false.

5. Weibo social media can be expressed in a casual manner, such as abbreviations, lack of uppercase letters, spelling, pronunciation, and syntax errors.

Http://stackoverflow.com/questions/4806176/what-are-the-most-challenging-issues-in-sentiment-analysisopinion-mining? Rq = 1