Abstract—Even though many incidents about fake online consumer reviews have been reported, very few studies have been conducted to date to examine the credibility of online consumer comments. One of the reasons is the lack of an effective computational method to deal with the huge number of online comments which are not embedded with explicit features for a spam detection system to separate the untruthful comments (i.e., spam) from the legitimate ones (i.e., ham). To improve the hygiene and the usefulness of online comments, there is a pressing need to develop a robust methodology for an objective and systematic assessment of the quality of online comments. The main contribution of this paper is the design, development, and evaluation of a novel machine learning based methodology for the assessment of the credibility of online comments. Our preliminary experiments show that the proposed quality assessment methodology is more effective than other baseline methods such as a peer-review based quality assessment method.

Index Terms—opinion credibility, opinion analysis, spam detection, SVM, machine learning

Cite: Otto K. M. Cheng and Raymond Y. K. Lau, "Toward a Machine Learning Based Approach for Assessing the Credibility of Online Comments," Journal of Industrial and Intelligent Information, Vol. 2, No. 3, pp. 175-178, September 2014. doi: 10.12720/jiii.2.3.175-178