Hi,

You need to do that yourself based on the understanding to the text. Another option is to have a prior predictive model to do this job.

HTH.

Samer Sarsam, PhD.

It is rather risky to assume that there are sentences not conveying opinion inside a review. Even if there are no emotive words (you can leave sentences with such words and get rid of the rest, for sure), an emoticon or "I am not going to buy it again" type sentences might be useful depending on your task (searching for particular aspects as "quality of sound" or "size" requires a different approach than searching for semantic reasons why a whole review has one or five stars).

Hi,

For each sentence, you must identify if exist or not the triple "subject-object-opinion words" and after that you can predict the opinion by using the sentiwordnet by example.

One of the sentiment analysis tasks is Subjectivity Detection, which is detecting if a text is objective or subjective.

Objective texts carry some factual information, for example, "The sky is blue," while

subjective texts express somebody's personal views or opinions, for example, "I like

the color blue ".

There are many studies in the literature on this task you can review them, such as:

1- Riloff, E., Wiebe, J., & Phillips, W. (2005, July). Exploiting subjectivity classification to improve information extraction. In Proceedings of the national conference on artificial intelligence (Vol. 20, No. 3, p. 1106). Menlo Park, CA; Cambridge, MA; London; AAAI Press; MIT Press; 1999.

2- Das, A., & Bandyopadhyay, S. (2009, September). Theme detection an exploration of opinion subjectivity. In Affective Computing and Intelligent Interaction and Workshops, 2009. ACII 2009. 3rd International Conference on (pp. 1-6). IEEE.

3- Barbosa, L., & Feng, J. (2010, August). Robust sentiment detection on twitter from biased and noisy data. In Proceedings of the 23rd international conference on computational linguistics: posters (pp. 36-44). Association for Computational Linguistics.

4- Lin, C., He, Y., & Everson, R. (2011). Sentence subjectivity detection with weakly-supervised learning. In Proceedings of 5th International Joint Conference on Natural Language Processing (pp. 1153-1161).

5- Chenlo, J. M., & Losada, D. E. (2014). An empirical study of sentence features for subjectivity and polarity classification. Information Sciences, 280, 275-288.