A web based product price forecast using fuzzy CBR

A healthy and sustainable development of aquaculture requires an accurate forecast of aquatic product prices. This paper presents an experiment with a web-based application to perform product price prediction. It integrates three key technologies: automatic data extraction algorithms to retrieve aquatic product prices from various related websites, attribute-oriented induction algorithms to capture varied time granularity data and Case-Based Reasoning (CBR) as its reasoning logic to deal with a focus on fuzzy matching. Test results of the experiment have shown that the application can successfully extract the web-based data as designed and effectively analyze the data leading to a reliable forecast of aquatic product prices.