Walking Route Recommender System Considering SAW Criteria

This paper proposes a walking route recommender system considering criteria for route safety, amenity and walkability. This paper refers to these criteria as SAW criteria. Walking is one of the easiest ways for health promotion, and a large number of people of all ages are enjoying it in various styles. As various people have different preference and health conditions, a walking rote recommender system has to provide them with a route considering various criteria, such as avoidance of steeps for elder people and existence of a coffee shop on route. However, existing route recommender systems usually employ only distance and necessary time between the current location and the specified location to make recommendations. This paper focuses on three criteria, including route safety, amenity and walkability, and proposes a method for recommending various routes considering these SAW criteria. In order to determine a route while considering such criteria, the proposed method combines A* algorithm and genetic algorithm. Another contribution of the system is to employ OSM (Open Street Map), which is converted into RDF (Resource Description Framework) and stored in SPARQL endpoint. Converting road information into RDF data makes it easy to extend database by incorporating various information about a road in future. This paper shows the proposed system is able to recommend reasonable and various routes through the simulations assuming users having various criteria and by subjective evaluation.