| Component technology is the core technology to support software reuse,how to retrieve the optimal components from the massive component library to improve the system assembly rate is a key problem to be solved.The selection of components generally includes the retrieval of components and the selection of components.Component retrieval usually selects multiple component classes,and efficiency is the core issue of retrieval;The choice of component is to select the component from the result set of the component retrieval according to the user's requirement,how to choose a more credible component from the retrieved data set is another problem to be studied in this paper.The main work of this paper is as follows:(1)The problem of component retrieval efficiency.Firstly,the traditional component classification technology is analysized.Aiming at its limitation,the method of component mining classification is introduced in component retrieval.The ant colony classification algorithm is used to reuse the rules.Secondly,the algorithm is improved for the antecedent problem of ant colony algorithm,the pheromone concentration of ants is adjusted and the mutation operator is introduced.Finally,the experimental results show that the improved ant colony algorithm is more effective in checking the accuracy and recall of component query.(2)Select the credibility problem for the component.In this paper,we evaluate the components from multiple non-functional attributes,and use particle swarm optimization to select the components that meet the needs of users.According to the characteristics of "early maturing" in the process of selecting the components of the particle swarm,the crossover and mutation operators are introduced.Increasing the population diversity makes it possible to obtain global optimal solutions.Finally,the improved algorithm is better in the selection of components,and the selected components are more satisfying to the needs of the users. |
PDF Full Text Request