Publication Type
Conference Proceeding Article
Version
publishedVersion
Publication Date
5-2022
Abstract
Shell commands are widely used for accomplishing tasks, such as network management and file manipulation, in Unix and Linux platforms. There are a large number of shell commands available. For example, 50,000+ commands are documented in the Ubuntu Manual Pages (MPs). Quite often, programmers feel frustrated when searching and orchestrating appropriate shell commands to accomplish specific tasks. To address the challenge, the shell programming community calls for easy-to-use tutorials for shell commands. However, existing tutorials (e.g., TLDR) only cover a limited number of frequently used commands for shell beginners and provide limited support for users to search for commands by a task. We propose an approach, i.e., ShellFusion, to automatically generate comprehensive answers (including relevant shell commands, scripts, and explanations) for shell programming tasks. Our approach integrates knowledge mined from Q&A posts in Stack Exchange, Ubuntu MPs, and TLDR tutorials. For a query that describes a shell programming task, ShellFusion recommends a list of relevant shell commands. Specifically, ShellFusion retrieves the top-n Q&A posts with questions similar to the query and detects shell commands with options (e.g., ls -t) from the accepted answers of the retrieved posts. Next, ShellFusion filters out irrelevant commands with descriptions in MP and TLDR that share little semantics with the query, and further ranks the candidate commands based on their similarities with the query and the retrieved posts. To help users understand how to achieve the task using a recommended command, ShellFusion generates a comprehensive answer for each command by synthesizing knowledge from Q&A posts, MPs, and TLDR. Our evaluation of 434 shell programming tasks shows that ShellFusion significantly outperforms Magnum (the state-of-the-art natural language-to-Bash command approach) by at least 179.6% in terms of MRR@K and MAP@K. A user study conducted with 20 shell programmers further shows that ShellFusion can help users address programming tasks more efficiently and accurately, compared with Magnum and DeepAns (a recent answer recommendation baseline).
Keywords
Shell programming, Answer generation, Knowledge fusion
Discipline
Computer Engineering
Research Areas
Data Science and Engineering
Publication
Proceedings of the 44th International Conference on Software Engineering, Pittsburgh, USA, 2022 May 25 - 27
ISBN
9781450392211
Identifier
10.1145/3510003.3510131
Publisher
IEEE Computer Society
City or Country
Pittsburgh, PA
Citation
ZHANG, Neng; LIU, Chao; XIA, Xin; TREUDE, Christoph; ZOU, Ying; LO, David; and ZHENG, Zibin.
ShellFusion: answer generation for shell programming tasks via knowledge fusion. (2022). Proceedings of the 44th International Conference on Software Engineering, Pittsburgh, USA, 2022 May 25 - 27.
Available at: https://ink.library.smu.edu.sg/sis_research/7709
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Additional URL
https://doi.org/10.1145/3510003.3510131