Search code examples
javastanford-nlp

Sentiment analysis of entity using stanford nlp

I extracted all the entities present in a particular sentence. For example if my sentence is

  infrastructure is good, Work-culture is pathetic,hikes are not good either

I have developed a code that gives me entity. now i need sentiment based upon entities. my output should be something like

infrastructure--> positive
work-culture--> negative
hikes--> negative

how am i supposed to do that?

Solution

  • If you are done with the coding next thing which is the most challenging part is to train the system with proper content. I have worked in Google prediction API for same sentiment analysis. You need content for the matter, means if it is a movie review then the training content should contains lots of movie review. I can tell you I have trained a system for movie review analysis with 30 movie review contents(15 positive and 15 negative). Still the system does not give 80% perfect result.