Indexing Spatio-Temporal Pada Data Hotspot

Abstract

The spatial information of hotspot is always changing over time. The recording of these changes makes it necessary to add dimension of time (temporal) into database. The database which is capable to store spatial and temporal information is known as spatiotemporal database. This database is bigger than a normal database because it stores all versions of an object over time and not allowed data to be physically deleted. A large database size causes it’s search time increases. It makes an access structure is needed in spatiotemporal database. Indexing is one of the access structure applied in the spatiotemporal database that accelerates data searching process. Spatiotemporal indexing method consists of two parts namely spatial indexing which is based on geographical information, and temporal indexing which is based on data timing information. The process of hotspot data indexing will be done in this research by using R-Tree approach for spatial data and Change Temporal Index (CTI) for temporal data. R-Tree is a hierarchical, height-balanced external memory data structure for multidimensional object. The hierarchical of R-Tree based on list of administrative such as regency, provinces and island. Change Temporal Index makes indexed data based on temporal side like valid time. CTI keeps valid data in current database and previous data according to interval valid time in historical database. This research is expected to accelerate the retrieving process in spatiotemporal databases. Spatial indexing makes searching faster up to 70%, temporal indexing makes searching faster up to 35% and spatiotemporal indexing makes searching faster than without indexing up to 50%.