Bloom filter-based framework for cache management in large cloud metadata databases

Abstract

As the data in cloud computing environment grows exponentially over the past few years, retrieving required data in shorter time becomes tedious. This paper proposes a probabilistic framework for efficient retrieval of data from huge datasets using combined approach of clustering and frequent pattern analysis using maximum frequent transaction (MFT) set algorithm based on similarity of transactions provided by a novel data structure called Bloomier matrix filter (BMF). In the proposed model clustering the metadata file is done on two levels. The first level of cluster is a base cluster which is created in an offline mode, while uploading the data based on keyword using tf-idf and second level of cluster is a derived cluster which is created in an online mode, while downloading the data. Frequent transactions are generated based on the run time statistics of the transaction provided by the Bloomier matrix filter analysis. Based on the run time statistics of the BMF the dynamic cluster is derived. We have implemented the model in a cloud environment and the experimental results shows that our approach is more efficient than the existing search technology and increases throughput by handling more number of queries efficiently with reduced latency.