Journal
IEEE ACCESS
Volume 10, Issue -, Pages 25555-25564Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/ACCESS.2022.3152806
Keywords
Business; Feature extraction; Electronic commerce; Machine learning; Sentiment analysis; Decision making; Support vector machines; Fake reviews; K-Nearest Neighbor (KNN); machine learning; natural language processing; sentiment analysis; support vector machine (SVM)
Categories
Funding
- Deanship of Scienti~c Research (DSR), King Abdulaziz University, Jeddah [D-213-611-1443]
Ask authors/readers for more resources
The outbreak of Covid-19 has led to an increase in global online shopping, highlighting the significant impact of online reviews on businesses. In our research, we proposed a fake review detection model using text classification and machine learning techniques, which outperformed other state-of-the-art techniques with high accuracy.
The outbreak of Covid-19 and the enforcement of lockdown, social distancing, and other precautionary measures lead to a global increase in online shopping. The increasing significance of online shopping and extensive use of e-commerce has increased competition between companies for online selling. Highlights that online reviews play a significant role in boosting a business or slandering it. Product review is an essential factor in customers' decision-making, leading to an intense topic known as fraudulent or fake reviews detection. Given these reviews' power over a business, the treacherous acts of giving false reviews for personal gains have increased with time. In our research, we proposed a fake review detection model by using Text Classification and techniques related to Machine Learning. We used classifiers such as Support Vector Machine, K-Nearest Neighbor, and logistic regression (SKL), using a bigram model that detects fraudulent reviews based on the number of pronouns, verbs, and sentiments. Our proposed methodology for detecting fake online reviews outperforms on the yelp dataset and the TripAdvisor dataset compared to other state-of-the-art techniques with 95% and 89.03% accuracy.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available