Abstract: Big data is a term for information sets that are so extensive or complex that traditional data processing applications are inadequate. Map Reduce concept is used with an incremental and distributed inference method for large-scale ontologies which realizes high-performance analysis and runtime searching, but it does not find out the slow node at execution time. This paper proposes to improve the efficiency of the map reduce scheduling algorithms by using SAMR scheduling technique which uses the factual information and finds the slow node and launches multi tasks. In addition, the usage timing of each user is calculated. Finally, implement and test the effectiveness of the proposed approach on the Hadoop framework. The purpose of this paper is to speed up the query to the user.

Keywords: Big data, SAMR, Map Reduce, ontology reasoning.