Fake consumer review detection using deep neural networks integrating word embeddings and emotion mining
- PDF / 838,443 Bytes
- 16 Pages / 595.276 x 790.866 pts Page_size
- 54 Downloads / 228 Views
(0123456789().,-volV)(0123456789(). ,- volV)
S.I. : EMERGING APPLICATIONS OF DEEP LEARNING AND SPIKING ANN
Fake consumer review detection using deep neural networks integrating word embeddings and emotion mining Petr Hajek1
•
Aliaksandr Barushka1 • Michal Munk2
Received: 9 November 2019 / Accepted: 23 January 2020 Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract Fake consumer review detection has attracted much interest in recent years owing to the increasing number of Internet purchases. Existing approaches to detect fake consumer reviews use the review content, product and reviewer information and other features to detect fake reviews. However, as shown in recent studies, the semantic meaning of reviews might be particularly important for text classification. In addition, the emotions hidden in the reviews may represent another potential indicator of fake content. To improve the performance of fake review detection, here we propose two neural network models that integrate traditional bag-of-words as well as the word context and consumer emotions. Specifically, the models learn document-level representation by using three sets of features: (1) n-grams, (2) word embeddings and (3) various lexicon-based emotion indicators. Such a high-dimensional feature representation is used to classify fake reviews into four domains. To demonstrate the effectiveness of the presented detection systems, we compare their classification performance with several state-of-the-art methods for fake review detection. The proposed systems perform well on all datasets, irrespective of their sentiment polarity and product category. Keywords Neural network Deep learning Fake review Review spam Word embedding Emotion
1 Introduction Fake consumer reviews provide a fictitious and misleading opinion that does not reflect a consumer’s authentic product experience. They can be submitted and published on multiple online platforms such as shopping portal forums. Globally, the number of users of these platforms has steadily increased over recent years. For example, TripAdvisor, the world’s largest online travel platform, has over 455 million unique visitors in an average month and & Petr Hajek [email protected] Aliaksandr Barushka [email protected] Michal Munk [email protected] 1
Faculty of Economics and Administration, Institute of System Engineering and Informatics, University of Pardubice, Studentska´ 84, 532 10 Pardubice, Czech Republic
2
Department of Computer Science, Constantine the Philosopher University in Nitra, 949 74 Nitra, Slovakia
600 million consumer reviews of 7.5 million varieties of accommodation, restaurants and attractions [68]. Generally, online product reviews provide valuable information for consumers, who increasingly rely on them and consider them to be a trusted source of information [63]. Most marketplaces prioritize well-evaluated products (the socalled snowball effect), potentially rewarding businesses paying for fake reviews. Review volume and review valence have been
Data Loading...
