An N-Way Model Merging Approach Based on Artificial Bee Colony Algorithm

Abstract

In N-way model merging, model matching plays an important role. However, the N-way model matching problem has been recognized as NP-hard. Aim: To search the optimal or near-optimal matching solution efficiently, this paper proposes an N-way model matching algorithm based on the Artificial Bee Colony (ABC) algorithm. Method: This algorithm combines global heuristic search and local search to deal with the complexity of N-way model matching. We evaluated the proposed N-way model merging approach through case studies and we evaluated the proposed ABCMatch algorithm by comparing it with Genetic Algorithm (GA) and Elephant Herding Optimization (EHO). Results: The experimental results show that ABCMatch can obtain more accurate model matching solutions in a shorter time, and the average model matching accuracy of ABCMatch is 2.7725% higher than GA and 1.8804% higher than EHO. Conclusion: Results demonstrate that our method provides an effective way for software engineers to merge UML models in collaborative modeling scenarios.