An Enhanced Tree Mining Algorithm for Finding Maximal Periodic Movements from Spatiotemporal Databases

Objectives: To find effective periodic patterns throughthe symbolic database representation of spatiotemporal data by using an efficient algorithm, ETMA (Enhanced Tree-based Mining Algorithm for large Databases). Methods/Statistical Analysis: There are distinct types of notions used to store and manage transactions data horizontally such as segment, sequence and individual symbols. The proposed algorithm can mine periodic patterns in full-series and subsection series also.The proposed algorithm based on prefix and suffix tree concurrently discover signal, series and section periodicity furthermore recognize and report only the efficient and non-redundant periods through the pruning techniques to abolish redundant (repetitive) periods. Findings: This algorithm finds interesting results with the help of a mixture of testing have been performed to validate the peroformance, strength, scalability, and correctness of the produced results in comparison with traditional algorithms. It is used to identify the three distinctcategories of maximal patterns effectively on various synthetic and reallife datasets. All tests are completedon distinct types of noisy such as insertion, deletion and replacement. ETMA reduces the running time and buildsservice of the proficient symbolic process. Moreover, ETMA simply report time-series instances dynamically, interms of symbol, sequence and segment approaches respectively. The length of the pattern, and proving efficiency of the pruning and searching strategies from synthetic and real datasets is a really open & challenging mining problem. Application/Improvements: This algorithm is 50% better than MAFIA and 20% better than STNR algorithms on Accident dataset and 50% better than STNR and 74% better than ECLAT algorithms on Mushroom dataset.