A deceptive detection model based on topic, sentiment, and sentence structure information
- PDF / 1,390,883 Bytes
- 14 Pages / 595.224 x 790.955 pts Page_size
- 40 Downloads / 207 Views
A deceptive detection model based on topic, sentiment, and sentence structure information Xiaodong Du1 · Ruiqi Zhu2 · Fuqiang Zhao1 · Fangzhou Zhao1 · Ping Han3 · Zhengyu Zhu1
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Deceptive reviews on Web are a common phenomenon and how to detect them has a very important impact on products, services, and even business policies. In order to filter out deceptive reviews more accurately, a new model called Sentence Joint Topic Sentiment Model (SJTSM) is presented in this paper, which incorporates the sentence structure of reviews and the sentiment label information of words based on Latent Dirichlet Allocation (LDA) model to extract the review features. The proposed model employs Gibbs algorithm to estimate the maximum likelihood parameters and takes the vector of topicsentiment distribution as the review features. Then a voting system of multiple-classifier, which takes the extracted review feature vector as its input is designed to realize the classification of deceptive review detection. The comparative experiments on different public datasets with other existing methods based on LDA model show that the new classifying system based on SJTSM model can achieve more satisfying classification results on deceptive review detection. Keywords Deceptive detection · Topic model · Multiple-classifier · Feature extraction
1 Introduction In recent years, with the rapid development of Internet techniques, people are willing to express and share their opinions on online platforms. Online reviews about a product or service will often determine whether customers would like to buy it or not, and impact the business in the long term either positively or negatively. Therefore, some businesses attempt to fabricate fake or deceptive reviews to attract customers, which will seriously damage the rights of customers and market balance. Meanwhile, the negative effects of deceptive reviews are fatal to enterprises. Accordingly, how to automatically detect deceptive reviews is becoming Ping Han
[email protected] Zhengyu Zhu
zhu [email protected] 1
College of Computer Science, Chongqing University, Chongqing, 400044, China
2
School of Computing and Information Systems, The University of Melbourne, Melbourne, VIC 3000, Australia
3
School of Foreign Languages and Cultures, Chongqing University, Chongqing, 400044, China
a challenge in the research field. Many researchers have proposed different models to filter deceptive reviews. The behavior of deceptive reviews can be detected by distinguishing their characteristics [1, 2]. Horne et al. [3] found that the overall title structure and the use of proper nouns in the title are very significant in differentiating fake from real. Choi et al. [4] proposed a deceptive detection method that combined Term Frequency-Inverse Document Frequency (TF-IDF) technique and symmetrical conditional probability. Similarly, Ahmed et al. [5] proposed a deceptive detection model that used TF-IDF as feature extraction technique
Data Loading...
