Affinity-based Fragmentation for Sensor Data
Kalgi Gandhi, Minal Bhise · 2019
Affinity-based Fragmentation AFA method for fragmenting rapidly growing RDF data is a workload aware method that keeps frequently queried data in a fragment. It is demonstrated using sensor Linked Open Dataset LOD. Performance parameters are Query Execution Time QET and Data Loading Time DLT. QET averaged over all the query types is 47% faster for fragmented data. QET for frequently occurring query types, linear and star, increases linearly while for range and snowflake, it scales exponentially. Results reported here verify that AFA executes in constant time for scaled data hence will help in building faster analytics for weather forecast.