Deep Context Identification of Deceptive Reviews Using Word Vectors
This paper proposes deep context by word vectors for deceptive review identification. The basic idea is that since deceptive reviews and truthful reviews are composed by writers without and with real experience, respectively, there should be different con
- PDF / 449,233 Bytes
- 12 Pages / 439.37 x 666.142 pts Page_size
- 70 Downloads / 180 Views
Research Center on Big Data Sciences, Beijing University of Chemical Technology, Beijing 100029, People’s Republic of China {zhangwen,yipan_jiang}@mail.buct.edu.cn 2 School of Knowledge Science, Japan Advanced Institute of Science and Technology, 1-1 Ashahidai, Nomi, Ishikawa 923-1292, Japan [email protected]
Abstract. This paper proposes deep context by word vectors for deceptive review identification. The basic idea is that since deceptive reviews and truthful reviews are composed by writers without and with real experience, respectively, there should be different contexts of words used by them. Unlike previous work using the whole text collection to learn the word vectors, we produce two numerical vectors for each word by embedding contexts of words in deceptive and truthful reviews separately. Specifically, we propose a representation method called DCWord (Deep Context representation by Word vectors) to use average word vectors derived from deceptive and truthful contexts, respectively, to represent reviews for further classification. Then, we investigate three classifiers as support vector machine (SVM), simple logistic regression (LR) and back propagation neural network (BPNN) to identify the deceptive reviews. Experimental results on the Spam dataset demonstrate that by using the DCWord representation, SVM and LR have produced comparable performance and they outperform BPNN in deceptive review identification. The outcome of this study provides potential implications for online business intelligence in identifying deceptive reviews. Keywords: Online business intelligence Skip-gram model representation Deceptive review identification Deep learning
DCWord
1 Introduction With the prevalence of Web 2.0 and social networking, it is widely accepted that for whatever commercial products, the users’ opinions are indispensible and valuable for its success of winning good reputation in the market [1, 2]. Online reviews, which refer to users’ opinions on a given product which they have using experience in or have something to talk about, are massively emerging in the Internet. These reviews are used by potential customers in purchasing decision making or e-commerce merchants in online promotion campaign. Due to word-of-mouth effect, positive online reviews are helpful for good reputation of products and merchants while negative online reviews will damage its reputation. On the one side, some merchants endeavor to produce and collect positive online reviews for themselves meanwhile defame their competitors with negative online © Springer Nature Singapore Pte Ltd. 2016 J. Chen et al. (Eds.): KSS 2016, CCIS 660, pp. 213–224, 2016. DOI: 10.1007/978-981-10-2857-1_19
214
W. Zhang et al.
reviews, even by hiring “water army” to post online reviews [3]. On the other side, it is impossible to identify deceptive reviews and truthful reviews by human beings satisfactorily [4]. Even worse, anyone can post online reviews anonymously in the Internet with a little cost but may cause great commercial goodness for themselves or
Data Loading...
