Implementation of Rabin Karp Algorithm in E-Commerce Search Box Feature (Case Study: Sinar Baja Store)

Abstract

Inside any e-commerce platform, search features are a key factor in an online business. In order to look for the desired item name, users need to type a pattern into the search feature. Inside the typing process users may make some mistypes. Based on the problem, the writer tried to create a search box as the search feature inside an e-commerce system with a capability to look for item names, even if there are some accidental mistypes inside the user's typing process. The main purpose is to reduce the possibility of empty results after a searching process due to the user’s accidental mistypes. This research utilizes Rabin Karp Algorithm in order to look for the item names. Result of this research is that the Rabin Karp Algorithm can be implemented in a website-based e-commerce, but this method is slower than SQL as a method to search item names inside the same system. It happens because Rabin Karp requires a longer process than SQL to get the similarity percentage from each comparison. This research shows that k-gram value inside the algorithm affects the overall results with a condition where greater the value.