Sequential data analysis is one of the important research tasks. Clustering is an important step in data analysis and it plays a vital role in data mining applications such as sequence clustering and web mining. Clustering of sequential data is the process of grouping the data based on the content and order of sequence similarity. Clustering of sequences requires computation of similarity between sequences. Rough clustering is one of the soft clustering methods which help to find overlapping clusters in real time applications. To overcome the limitation of rough set, an extension approach called covering based rough set is introduced where the partitions of a universe are replaced by covering in covering-based rough sets. This paper presents a novel soft sequential recognition approach to monitor user activity using wearable accelerometer device. © IAEME Publication.