TID based data and task parallelism for frequent data mining
Swati Rustogi, Manisha Sharma, Sudha Morwal · Journal of Information and Optimization Sciences · 2017
Data mining is used to extract hidden information from a large set of data. From the various data mining algorithms available, frequent data mining is one of the most relevant algorithms. This algorithm has been used in varied fields ranging from medicine, weather forecasting, education, insurance, etc. Frequent mining (FM) algorithms have evolved from single processor to multiple processors. Limitation of using multiprocessors is increased cost and decreased throughput. Also the FM algorithms have been based on single process. This is an impediment in case of large data. One solution is to use multiple processes in multi-core environment.In this paper we propose a FM algorithm based on task parallelism. The algorithm is compared with FM algorithm based on data parallelism and serial algorithm.