Scalable Spatio-Temporal Reasoning of Sequential Events using Spark Framework

Spatio-temporal Big Data analytics has empowered the computational efficiency of software systems. Knowledge discovery using Spatio-temporal Big Data analytics is highly demanded in geographic applications. Remote sensing technology has made ease the availability of spatial and temporal facts of earth resources. We propose scalable reasoning on formal representation over distributed architecture considering the spatial and temporal facts of geographic feature. Scalability in data analytics towards reasoning critical geographic events is essential for computation, as spatio-temporal facts scale over period of time. This necessitates inferring geospatial events and processes causing geospatial dynamism on a distributed frame-work. Thus, scalable MapReduce reasoning on formal representation of knowledge base is proposed. Logical event based queries are evaluated on distributed knowledge base. The comparative evaluation on execution time of the system is done by evaluating event based queries on local data store and distributed data store. In future, the proposed distributed reasoning is to be experimented on cluster with varying number of nodes to support computation in large scale.