Design and analysis of similarity measure for discovering similarity profiled temporal association patterns

A wide variety of real time applications generate temporaldata. Determining and unearthing similar temporal association patterns is complex and challenging task when considering time stamped temporal databases. Previous works have considered only existing distance measure and did not address retrieval and discovery of similar temporal association patterns using new distance measures. In this paper, we design a new similarity measure which suits the temporal context thatcan be applied to obtain all valid STAP (similar temporal association patterns). Our approach considers approximating the association pattern support bounds and applying the proposed similarity measure. The results show the proposed approach has better computational complexity compared to other approaches and improvedtime efficiency.