ICSME 2023
Sun 1 - Fri 6 October 2023 Bogotá, Colombia

Stack Overflow (SO) is a widely used question-and-answer (Q&A) forum dedicated to software development. It plays a supplementary role to official documentation (DOC for short) by offering practical examples and resolving uncertainties. However, the process of simultaneously consulting both the documentation and SO posts can be challenging and time-consuming due to their disconnected nature. In this study, we propose DOSA, a novel approach to automatically align SO and DOC, which inject domain-specific knowledge about the DOC structure into large language models (LLMs) through weak supervision and constrained decoding, thereby enhancing knowledge retrieval and streamlining task completion during the software development procedure. Our preliminary experiments find that DOSA outperforms various widely-used baselines, showing the promise of using generative retrieval models to perform low-resource software engineering tasks.

Fri 6 Oct

Displayed time zone: Bogota, Lima, Quito, Rio Branco change

10:30 - 12:00
10:30
16m
Talk
How do Developers Improve Code Readability? An Empirical Study of Pull Requests
Research Track
Carlos Eduardo Carvalho Dantas Federal University of Uberlândia, Adriano Mendonça Rocha Federal University of Uberlândia, Marcelo De Almeida Maia Federal University of Uberlandia
10:46
11m
Talk
Summarize Me: The Future of Issue Thread Interpretation
New Ideas and Emerging Results Track
Abhishek Kumar Indian Institute of Technology Kharagpur, Partha Pratim Das Indian Institute of Technology, Kharagpur, Partha Pratim Chakrabarti Indian Institute of Technology, Kharagpur
10:57
11m
Talk
Bugsplainer: Leveraging Code Structures to Explain Software Bugs with Neural Machine Translation
Tool Demo Track
Parvez Mahbub Dalhousie University, Ohiduzzaman Shuvo Dalhousie University, Masud Rahman Dalhousie University, Avinash Gopal
11:08
16m
Talk
Knowledge Graph based Explainable Question Retrieval for Programming Tasks
Research Track
Mingwei Liu Fudan University, Simin Yu Fudan University, Xin Peng Fudan University, Xueying Du Fudan University, Tianyong Yang Fudan University, Huanjun Xu Fudan University, Gaoyang Zhang Fudan University
Pre-print File Attached
11:24
11m
Talk
Investigating the Impact of Vocabulary Difficulty and Code Naturalness on Program Comprehension
Registered Reports Track
Bin Lin Radboud University, Gregorio Robles Universidad Rey Juan Carlos
11:35
11m
Talk
Aligning Documentation and Q&A Forum through Constrained Decoding with Weak Supervision
New Ideas and Emerging Results Track
Rohith Pudari University of Toronto, Shiyuan Zhou University of Toronto, Iftekhar Ahmed University of California at Irvine, Zhuyun Dai Google, Shurui Zhou University of Toronto
Pre-print
11:46
14m
Live Q&A
1:1 Q&A
Research Track