Evaluation of Ratings and Review Sentiments Using SVM and Random Forest for Detecting Product Authenticity on Tokopediax

Abstrak

The digital era has made online shopping easier, but it has also presented significant challenges regarding product authenticity, especially for branded goods that are often counterfeited. This study aims to detect discrepancies between product ratings and reviews for Adidas products on Tokopedia. A total of 2,180 data points were obtained using web scraping techniques with Python. The data was processed through data pre-processing and feature extraction using TF-IDF. Machine learning algorithms Support Vector Machine (SVM) and Random Forest were applied with a training and testing data split of 80:20. The final evaluation results showed that the SVM model achieved an accuracy of 93.1% and an F1 score of 92.5%, while Random Forest achieved an accuracy of 83.7% and an F1 score of 82.7%. This indicates that sentiment analysis can be effectively used to detect discrepancies between ratings and reviews, which can support the development of product authenticity detection systems on e-commerce platforms.

Identifier

• OpenAlex: https://openalex.org/W7169860783
• DOI: https://doi.org/10.24167/proxies.v9i2.14046