Analysis of the Lowest Product Prices in Online Stores Using Tsukamoto Fuzzy Logic with Two Sets in a Web-Based System

This study develops a web-based recommendation system to assist consumers in identifying the lowest product prices across Indonesian online stores by applying Tsukamoto fuzzy logic with two sets. The research is motivated by Indonesia’s rapidly expanding digital economy, projected to reach USD 100 billion by 2025, where competitive pricing plays a central role in consumer decision-making. The system was designed and implemented using PHP and MySQL, with input variables comprising product price (Rp0–Rp10,000,000) and store ratings (1–5 scale). Data were collected through literature review, observation, interviews, and website analysis, followed by system design, development, and evaluation. Testing on 50 electronic products produced an average recommended price of Rp2,625,000, with a 95% accuracy rate compared to manual searches and an average response time of 1.5 seconds. User evaluation, involving 30 respondents, indicated a 90% satisfaction rate, emphasizing ease of use and responsiveness. The Tsukamoto fuzzy logic approach demonstrated effectiveness in handling uncertainty between price and rating variables, generating precise recommendations through monotonic membership functions and α-predicate calculations. However, the system showed reduced accuracy for mid-range ratings, reflecting limitations of using only two fuzzy sets and aligning with previous findings comparing Tsukamoto and Sugeno methods. Additionally, frequent price fluctuations in online marketplaces necessitate periodic data updates. The study concludes that this system provides a practical and accurate tool to support online shopping decisions in Indonesia’s dynamic e-commerce landscape. Future improvements may include introducing a third fuzzy set, integrating real-time data retrieval, and enhancing consumer preference analysis through sentiment-based approaches.