A text categorisation tool for open source communities based on semantic analysis

Abstract

Open source software (OSS) projects are supported by communities interacting through software repositories and mailing lists. Thousands of contributors participate in the development of the projects although they rarely meet each other. The result is a huge archived repository with thousands of questions, answers and contributions usually difficult to explore. We propose a tool based on semantic analysis for both performing an automatic knowledge discovery and a categorisation of the content of mailing lists repositories. Semantic analysis is a practical method for extracting and inferring relations of words in passages of discourse, producing measures of relations among words or passages that are well correlated with semantic similarity. The objective of this article is two-fold: (1) to develop a text categorisation tool based on indexing terms and semantic annotation, and (2) to apply the developed tool to extract the main dimensions related to knowledge sharing activities in virtual communities. Debian Linux ports to embedded processors are used as a case study to accomplish the proposed double objective.

Keywords:

• semantic analysis
• text categorisation
• open source
• virtual communities

Acknowledgements

This work has been supported by the Spanish Ministry of Education and Science (Research Project with reference DPI2007-60128) and the Consejería de Innovación, Ciencia y Empresa (Research Project with reference P07-TIC-02621).