Research On The Evolution Of Open Source Software Ecosystem In GitHub

In recent years,open source has become the core means of high-efficiency and high-quality software development.Open source software has become a strategic resource demand in various industries,and the form of software development has been further expanded from the original small-scale development to a large-scale and standardized development model.Open source software ecosystem also came into being.Open source software ecosystem refers to a collection of open source software projects that are interrelated and co evolved in a common development platform or environment.Its main characteristics are complex relationship and dynamic evolution.The characteristics of its dynamic evolution have accelerated the horizontal integration and vertical division of labor of the software industry,and profoundly changed the previous ecological structure of the software industry.Since it was proposed in 2003,open source software ecosystem and its evolution have attracted extensive attention from academia and industry.Insight into the evolution law of open source software ecosystem plays a good guiding role in studying the social relations,resource flow and predicting the future trend of software ecology.However,up to now,there is no general method to quantify the evolution of open source software ecosystem.Therefore,this paper has carried out the following work:(1)Based on the Social Network theory,the open source software project network is modeled,a community detection method(ABIS)based on the interaction strength between software project repositories is proposed,and the open source software ecosystem on Git Hub is mined in a panoramic way,and more than 400 software ecosystems in six stages are obtained.(2)A state matching method based on intersection union ratio(IOU)is proposed to identify the software ecosystem living in adjacent stages.According to this method,three states of software ecosystem are defined: rebirth,development and extinction; At the same time,the factor analysis model is used to summarize the five main evolution factors of open source software ecosystem: scale factor,structure factor,activity factor,attention factor and benefit factor.(3)The Grey Wolf Optimization(GWO)algorithm is introduced into the acquisition of evolution rules of open source software ecosystem,and the appropriate optimization operator is designed.By solving the optimization problem,the eigenvector expression of evolution factors is obtained,and then the evolution rules are obtained.The relationship between evolution state and evolution factors is quantitatively analyzed,and a prototype of intelligent acquisition tool for evolution rules of open source software ecosystem is designed and implemented.(4)Experiments are carried out in GH Archive data set and Net Logo simulation environment to verify the practicability and effectiveness of the proposed method.