Keyword-aware dominant route search for various user preferences

Abstract

Route search has been studied extensively. However existing solutions for route search are often insufficient in offering users the flexibility to specify their requirements. Recently, a new kind of keyword aware optimal route (KOR) query is proposed for finding an optimal route such that it covers a set of user-specified keywords, it satisfies a budget constraint (e.g., time), and the total popularity of the route is maximized. For example, a user may search for the route passing through cinema and bookstore within a travel time of 3 hours, which has the highest total rating score. KOR only returns one result regardless of user preferences: however some may care more about bookstore, while others think cinema more important. Apparently, it is not user friendly for users to specify their own preference explicitly. To meet the various user preferences, this paper presents a new route search query called Keyword-aware Dominant Routes (KDR) query, which returns all possible preferable routes, covering a set of keywords under a specified budget constraint. We devise a novel algorithm DR for efficiently finding the exact result of a KDR query and a heuristic algorithm FDR to approximately answer a KDR query. Experimental results show that our algorithms are efficient.