Research On Discovery Of Component Usage Pattern Based Component Library Management System

With the rapid development of component-based software development technology, the component library becomes more complex, learning and retrieval of the components have higher demand.Sequential patterns discovery is a process of mining frequent longest sequence in a given time window on the sequential dataset. Given sequential database and the minimum supports, identify the longest sequence in the sequential database to meet the minimum threshold of supports, every one of the longest sequence is a sequential pattern. AprioriAll and GSP are recognized classic algorithms. Software component technology is key technology of current software engineering, software reuse, which relating to component representation, component classification, CLMS, retrieval and so on. How to improve the level of reusable components, for the use of personnel to provide more convenient, more personalized search component is a important task of component library management system.The process and algorithm of sequential patterns discovery of components usage, is given in this paper, and to further, the model of component library management system with sequential patterns is discussed.The process of PGSG splits into five stages, data migration, data preprocessing, algorithm computing, refinement and interpretation of results and saving the result. The algorithm has a faceted parameter, what can be targeted to a particular interest in the different facets of the concept of component usage in sequential patterns discovery. The component usage pattern based component library management system has one pattern library to store the found effective sequential patterns, one prgram&method library to store the algorithm procedures and application interface, with concurrency control. Due to the patterns library, the improvement of provide support to component retrieval in components library is notable. The achivement of this paper could improve component classification and intelligent retrieval, improve the quality and efficiency of learning and retrieval component.