Socio-technical analysis of evolving software library dependency networks
-
TypePost-doctorat
Description
Today\’s software systems heavily rely on open source software libraries. These libraries are accessible through a wide variety of so-called package distributions, and the development of these libraries is taking place through so-called social coding platforms such as GitHub, BitBucket and GitLab. These collections of software libraries form huge and complex software dependency networks that rapidly evolve over time. The collaborative software development activity through the social coding platforms adds many additional dimensions to the picture: a social dimension through the different project contributors and their interaction; a technical dimension through the tools they are using (bug and issue trackers, continuous integration systems, development bots, automated code review, quality and security analysis, and so on); a legal dimension related to software licensing issues, ethical considerations and data privacy issues. As such, developer communities are faced with many challenges in these rapidly evolving socio-technical software ecosystems.
The aim of this postdoctoral topic will be to conduct mixed-methods research, combining qualitative and quantitative empirical studies to understand the socio-technical issuess faced by these software ecosystems, and to provide recommendation, prediction and forecasting models and their associated tooling to alleviate these problems. The postdoctoral candidate should have a keen interest in open source software, good analytical and Python programming skills, be fluent in statistical analysis, data mining, and machine learning.
The exact topic of study will be determined in concertation between the promotor and the candidate. The proposed research fits in the Belgian interuniversity \ »excellence of science\ » research project SECO-ASSIST. Consult https://secoassist.github.io for examples of publications, presentations, videos, tools and datasets related to the proposed research topic.