A Faster Approach for Establishing Trip Similarity

Abstract

Location data traces sourced from devices like GPS, Mobile Signals etc. have important uses in travel pattern applications. Location data traces are arranged as trips and pattern mining algorithms are applied to extract underlying patterns. Many of these algorithms/techniques are dependent on similarity of these trips made by users for example vehicle turn prediction, similarity in route traveling, frequent patterns in traveling the route, user travel route prediction etc. This paper presents an efficient way to establish the similarity between the trips using clustering technique. Proposed technique for trips clustering is efficient in data storage as well as computational time.