Parallel Processing for Fuzzy Queries in Human Resources Websites

Abstract

In this paper, we use Parallel FuzzyCLIPS to parallelize the execution of FQHR websites for two kinds of task partitioning in both grid and cluster environments. First, a new architecture of FQHR websites is proposed to parallelize the execution of fuzzy queries on grid and cluster systems. Second, our approach supports two kinds of task partitioning: data level and module level. The data level partitions the facts and allocates them to multiple processors for parallel execution, while the module level partitions the rules and allocates them to different processors. Third, a performance evaluation model is presented to analyze the proposed approach. Finally, we implement a parallelized FQHR website to test the speedups by experiments and to verify the results of performance analysis.