Early Detection Method for Money Fraudulent Activities on E-commerce Platform via Sentiment Analysis

Abstract

Shopee is one of the prominent online shopping platforms in Malaysia. Nonetheless, countless scam cases were reported on the platform, particularly on money fraudulent activities. This study constructed a quantitative model through linear programming that assessed sentiments based on customers’ reviews. Reviews from three selected Shopee products (‘M3 Smart Health Watch’, ‘Sony Headset Wired Gaming Headphone’, and ‘20000mAh Pineng 100% Original Powerbank’) were analysed using the proposed model. The data were converted into measurable metrics to enable quantitative fraud detection. The model enabled the early detection of possible money fraudulent activities on Shopee products based on customers’ reviews. Resultantly, ‘M3 Smart Health Watch’ is an authentic Shopee product. In contrast, ‘Sony Headset Wired Gaming Headphone’ and ‘20000mAh Pineng 100% Original Powerbank’ are money fraud products sold by scammers. The proposed model utilises free and readily available software, thus extending its usability to other small business owners.