Mining frequent spatial-textual sequence patterns

Abstract

Penetration of GPS-enabled devices has resulted in the generation of a lot of Spatial-Textual data, which can be mined or analyzed to improve various location-based services. One such kind of data is Spatial-Textual sequential data (Activity-Trajectory data), i.e. a sequence of locations visited by a user with each location having a set of activities performed by the user is a Spatial-Textual sequence. Mining such data for knowledge discovery is a cumbersome task due to the complexity of the data type and its representation. In this paper, we propose a mining framework along with algorithms for mining Spatial-Textual sequence data to find out frequent Spatial-Textual sequence patterns. We study the use of existing sequence mining algorithms in the context of Spatial-Textual sequence data and propose efficient algorithms which outperform existing algorithms in terms of computation time, as we observed by extensive experimentation. We also design an external memory algorithm to mine large-size data which cannot be accommodated in main memory. The external memory algorithm uses spatial dimension to partition the data into a set of chunks to minimize the number of false positives and has been shown to outperform the naïve external-memory algorithm that uses random partitioning.