Fake Reviews Detection

The Study in the paper discusses the importance of accurately identifying fake online reviews, which can influence customer decisions and cause financial losses.

It suggests a method for identifying outlier reviews that involves examining records related to the products under consideration, rather than just looking at the reviews and comments themselves.

The method involves extracting review records to a temporal feature vector and using an isolation forest algorithm to identify outlier product reviews. The performance of this method will be evaluated using a dataset from Amazon China, and will be compared to other temporal outlier detection techniques.

The impact of different parameters on review record identification was investigated.