Efficient and Scalable Spatial Retrieval of Resident Involvement Information in City Events

Abstract

Information about resident involvement in reporting various city-related events (e.g., Potholes, traffic jams) via mobile apps is critical to key stakeholders for effective city management. While existing efforts have primarily focused on event data collection in space, they are not capable of performing efficient retrieval of information about resident involvement in event reporting across different spatial regions and at different spatial granularities. Hence, this work makes the following contributions. First, we present CRIS, a scalable system for efficient retrieval of resident involvement and event information across different spatial regions and at varying spatial granularities. Second, we propose the euR-tree, a novel R-tree-based index augmented with (a) a hash-based array for indexing events in space, and (b) fixed-length arrays for indexing resident involvement information in reporting events in space. The euR-tree is integrated into CRIS to realize efficient retrieval. Third, our performance study indicates that the euR-tree is indeed effective in performing such retrieval with reduced query response times and disk I/Os.