Abstract
Authenticity and reliability of the information spread over the cyberspace is becoming increasingly important. This is especially important in e-commerce since potential customers check reviews and customer feedbacks online before making a purchasing decision. Although this information is easily accessible through related websites, lack of verification of the authenticity of these reviews raises concerns about their reliability. Besides, fraudulent users disseminate misinformation to deceive people into acting against their interest. So, detection of fake and unreliable reviews is a crucial problem that must be addressed by the security researchers. Here we propose a spam review detection framework that incorporates knowledge extracted from the textual content of the reviews with information obtained by exploiting the underlying reviewer-product network structure. In the proposed framework, first, feature vectors are learned for each review, reviewer and product by utilizing state-of-the-art algorithms developed for learning document and node embeddings, and then these are fed into a classifier to identify opinion spam. The effectiveness of our framework over existing techniques on detecting spam reviews is demonstrated in three different data sets containing online reviews. The experimental results obtained confirm that combining representations learned from reviewer-product network and textual review data significantly improves the detection of spam reviews.