A Cross-domain Deceptive Opinion Detection by Genetic Algorithm

DOI

10.2991/icmmct-16.2016.123How to use a DOI?

Keywords

Deceptive reviews, Cross-domain, Genetic algorithm, Spectral clustering

Abstract

With the maturing of the electronic commerce,many people will choose to shop online.At the same time the goods's reviews will influence the decision of people to buy goods,which has led businesses to deliberately write deceptive opinion to improve the sales of goods or undercut rivals .Aiming at the problem of the deceptive reviews,this paper proposes a deceptive reviews detection based on cross-domain and genetic algorithm.To define the data of feature in the source domain ,using genetic algorithm for feature to get the optimal feature set and constructing the feature incidence matrix M y the similarity of documents.it reducing the dimensionality of M , by using spectral clustering algorithm,to obtain a corresponding space mapping.then, with the help of sentiment classifiers,training and being classified in the target domain.Experimental results points that the feasibility and advantages of the method in recognizing the problem of deceptive reviews.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Qiao-jing Tang
AU  - Wei-hua Li
AU  - Jing Zhao
PY  - 2016/03
DA  - 2016/03
TI  - A Cross-domain Deceptive Opinion Detection by Genetic Algorithm
BT  - Proceedings of the 2016 4th International Conference on Machinery, Materials and Computing Technology
PB  - Atlantis Press
SP  - 621
EP  - 625
SN  - 2352-5401
UR  - https://doi.org/10.2991/icmmct-16.2016.123
DO  - 10.2991/icmmct-16.2016.123
ID  - Tang2016/03
ER  -