|
Abstract
|
Online product review websites have become excellent platforms for customers
to share their opinions about a variety of products and services in the form of online reviews.
Despite being an invaluable source of information for both consumers and rms, the quality
of online reviews varies greatly. To tackle the problem of low quality reviews, in this
paper, we address reviewer credibility problem and propose a novel framework for ranking
reviewers in terms of credibility based on Interval Type-2 Fuzzy Analytical Hierarchy
Process (IT2 FAHP) and technique for order performance by similarity to ideal solution
(TOPSIS). The novel IT2 FAHP was used to determine weights of features representing
reviewers, where the interval type-2 trapezoidal fuzzy numbers were used predominantly
and TOPSIS method was used to obtain the nal ranking of reviewers. To illustrate an
application of the proposed framework, we conducted an experimental study using real data
crawled from Epinions. This proposed framework provides a more e ective and systematic
approach, especially for rms, to nd highly credible reviewers and select their reviews and
opinions for opinion mining.
|