SentiWordNet: A Publicly Available Lexical Resource for Opinion Mining
Citation
SENTIWORDNET: A Publicly Available Lexical Resource for Opinion Mining. In Proceedings of the 5th Conference on Language Resources and Evaluation (LREC’06),417-422.
Online version
Summary
Key Ideas There are mainly three subtasks related to opinion mining. First, to determine if the text has subjective expressions or is it objective. A given text can belong to either subjective class or objective class. This is coined as SO-polarity. Second, to determine if the subjective text expresses a positive or negative opinion. This is coined as PN-polarity. Third, to determine the strength of positive or negative opnion.
Previous work of assigning the PN polarity has been limited to confidence of assigned label. In this paper the score is the strength of positive or negative opinion as compared to the confidence of the positive netiave label.
the grading approach allows scope for subtle subjectivity in words naunces. Where as the previous approaches assigned a hard binary labels.
In the previous paper the authors solve the problem of determining whether a sentence is subjective or not and if it is subjective does it
have the postive or negative polarity. The combined problem is tougher than just determining the polarity of a
sentence.
determining setence subjectivity by performing binary text categorization under categories Objective and Subjective (Pang and Lee, 2004; Yu and Hatzivassiloglou, 2003);
to aid these tasks, recent work has tackled the issue of identifying the orirentaiton of subjective terms contained in text. i.e determine whether a term that carries opinoanted content has a positive or negative conotation this is the key component in identifying the oreintation of the documents. There has been combined approaches to solve the above problem but they lack realism that
The very first work in determining the term orientation has been to determine the orientatino of subjective adjectives extracted from large unlabelled document set.The underlying intuition is that the act of conjoining adjectives is subject to linguistic constraints on the orientation of the adjectives involved; e.g. and usually conjoins adjectives of equal orientation, while but conjoins adjectives of opposite orientation. The authors generate a graph where terms are nodes connected by “equal-orientation” or “opposite-orientation” edges, depending on the conjunctions extracted from the document set. A clustering algorithm then partitions the graph into a Positive cluster and a Negative cluster, based on a relation of similarity induced by the edges
