Since the walkthrough system may contain large-scale VRML models, the massive objects are always stored and scattered in the storage units of the walkthrough system. But this situation will increase the search time and reduce the system performance. However, this problem is never considered in the traditional walkthrough system. In this paper, we present an efficient clustering method to improve the efficiency of accessing objects. Meanwhile, the clustering methodology is particularly appropriate for the exploration of interrelationships among objects to reduce the access time. In other words, we introduce the relationships among transactions, views and objects. Based on these relationships, the clustering algorithm will determine how to cluster these objects efficiently. Besides, we suggest two clustering criteria - intra-pattern similarity table and inter-pattern frequency table. Our experimental evaluation on the walkthrough data set shows that our algorithm doesn't only significantly cut down the access time, but also enhance the performance of our system.