Distributed query optimization and indexing schemas in heterogeneous cloud server

Abstract

Distributed query processing is the procedure of answering queries in a distributed environment where data is managed at multiple sites in a computer system Query processing involves the transformation of a high level query into query execution plan and also the execution of this plan The goal of this transformation is to produce a plan which is equivalent to the original query and is also efficient Optimizing the query processing time in terms of its processing time and computational cost was the motivating factor to take up this research An effective distributed query processing using the method of canonicalization is proposed in this thesis Initially the big data is distributed across the Hadoop platform Then the data is preprocessed and the noise is removed from it With the pre-processed data, the metadata is created and the data is indexed by using NoSQL The user query is compared with the created metadata and it is parsed and verified Then the user query is mapped with the database Finally the semantic query matching is employed to check the given query with the database to enhance the search accuracy The Citation Parser CP method is used to convert the query string into simple symbols thus reduces the query processing time and enhances the system performance For enhancing the satisfaction level the proposed Optimal Query Anatomization uses the similarity and filtering approaches for storing processing and managing huge amounts of data Initially preprocessing is done to remove the noise and the inconsistencies newline