Software maintenance is one of the most crucial aspects of software development. Software engineering researchers must develop practical solutions to handle the challenges presented in maintaining mature software systems. Research that addresses practical means of mitigating the risks involved when changing software, reducing the complexity of mature software systems, and eliminating the introduction of preventable bugs is paramount to today’s software engineering discipline. Giving software developers the information that they need to make quality decisions about changes that will negatively affect their software systems is a key aspect to mitigating those risks. This dissertation presents work performed to assist developers to collect and process data that plays a role in change decision-making during the maintenance phase. To address these problems, developers need a way to better understand the effects of a change prior to making the change. This research addresses the problems associated with increasing architectural complexity caused by software change using a twoold approach. The first approach is to characterize software changes to assess their architectural impact prior to their implementation. The second approach is to identify a set of architecture metrics that correlate to system quality and maintainability and to use these metrics to determine the level of difficulty involved in making a change. The two approaches have been combined and the results presented provide developers with a beneficial analysis framework that offers insight into the change process.

This paper explores an approach to extracting historical insights into a software keyword by data mining GitHub and Libraries.io. We test our method using the keyword GraphQL to see what insights we can gain. We managed to plot several timelines of how repositories and software libraries related to our keyword were created over time. We could also do a rudimentary analysis of how active said items were. We also extracted programing language data associated with each repository and library from GitHub and Libraries.io. With this data, we could, at worst, correlate which programming languages were associated with each item or, in the best case, predict what implementations of GraphQL they used. We found through our attempt many problems and caveats that needed to be dealt with but still concluded that extracting historical insights by data mining GitHub and Libraries.io is worthwhile.