DeepSoft-C: Automatically Recommending Components for Issue Reports Using Deep Learning

Mahidol University focuses on creating quality research and innovation which is beneficial to economic, social and human development. The Faculty of Information and Communication Technology (ICT), has supported instructors and students to learn, explore and develop software using Artificial Intelligence (AI). The AI, human-like intelligence, can solve software problems, help analyze big data, and assist in developing deep learning.

Currently, it is common for organizations to develop large-scale software projects which generally contain complicated work processes and numerous components. In addition, problems and issues might occur during software development. In order to reduce the issues during the software development process, it is necessary to follow the issue solving process and create issue reports. Software developers then need to manage the details of the software development project which might cause them difficulties and delays in the development process.

DeepSoft-C is a research project which helps solve problems of the Issue Reports in software development, using Artificial Intelligence (AI) for analyzing, processing and categorizing the software components. This research is a collaboration of 3 universities from 2 countries:

  1. The Software Engineering Research Unit (SERU), the Faculty of Information and Communication Technology, Mahidol University, led by Dr. Morakot Choetkiertikul, and Dr. Chaiyong Ragkhitwetsagul, instructors of the Computer Science Academic Group
  2. The School of Computing and Information Technology, Faculty of Engineering and Information Sciences, University of Wollongong, Australia, led by Prof. Hoa Khanh Dam, and Prof. Aditya Ghose, Director of the Decision Systems Lab, University of Wollongong
  3. The School of Information Technology, Deakin University, Geelong, Australia, led by Prof. Dr. Truyen Tran, and Trang Pham, (Ph.D. student)

The DeepSoft-C Model will support software developers when the issues are reported. The AI of the DeepSoft-C will analyze the details of the Issue Report: details, problems, causes, and relations of the issues in order to recommend relevant software components. The DeepSoft-C Model helps software developers categorize the relevant software components precisely and rapidly in order to increase efficiency in managing software development projects. Those who are interested in the full article, please visit: https://doi.org/10.1007/s10664-020-09898-5.