RECOMMENDATION SYSTEM WITH CONTENT-BASED FILTERING METHOD FOR CULINARY TOURISM IN MANGAN APPLICATION

Abstract

Culinary tourism has become one of the current trends in the culinary world. The amount of information circulating does not necessarily make culinary tourists easier to determine the desired menu choices. The use of search engines alone is still not enough, so we need a recommendation system that can provide advice according to user needs.

Content-based filtering method is able to produce recommendations that are user independence, making it suitable for use in developing culinary information providers such as the case studies of MANGAN applications where the number of users is still small and culinary data will always increase. This method recommends several objects based on the similarity of the selected object to the recommended object. Object similarity is calculated using the cosine similarity function based on profile items formed from a restaurant's content features.

Tests carried out on the results of recommendations with three different thresholds, obtained an average precision value of  0,8915 and an average accuracy value of 0,5118. The low value of accuracy is due to systematic errors. The results of this study are the content-based filtering method can be used to assist users in choosing restaurants based on the similarity of the item profile of a restaurant.