Wed 6 Dec 2023 11:45 - 12:00 at Golden Gate C2 - Software Evolution II Chair(s): Csaba Nagy

Version incompatibility issues are rampant when reusing or reproducing deep learning models and applications. Existing techniques are limited to library dependency specifications declared in PyPI or open-source projects on GitHub. Therefore, they cannot account for latent errors due to undocumented version constraints or dependency to non-Python libraries such as CUDA and cuDNN. Meanwhile, Stack Overflow (SO) offers abundant and up-to-date discussions of version issues spanning across the deep learning stack, e.g., libraries, runtime, OS, and GPU.

We propose to extract the rich knowledge from SO and build a knowledge graph to support the detection of version compatibility issues. Specifically, we develop a novel indirect supervision method that uses a pre-trained Question-Answering (QA) model to extract logic facts from online discussions. Extracted logic facts are further consolidated into a probabilistic knowledge graph to resolve duplicates and conflicts in online discussions. Our evaluation results show that (1) our system can accurately extract version-related facts with 81% precision and 88% recall, and (2) our system can accurately identify 65% while two state-of-the-art approaches can only detect 18% and 6% on a benchmark of 10 popular DL projects.

Wed 6 Dec

Displayed time zone: Pacific Time (US & Canada) change

11:00 - 12:30
Software Evolution IIIndustry Papers / Research Papers / Journal First at Golden Gate C2
Chair(s): Csaba Nagy Software Institute - USI, Lugano
11:00
15m
Talk
Multilingual Code Co-Evolution Using Large Language Models
Research Papers
Jiyang Zhang University of Texas at Austin, Pengyu Nie University of Waterloo, Junyi Jessy Li University of Texas at Austin, USA, Milos Gligoric University of Texas at Austin
Media Attached
11:15
15m
Talk
A Characterization Study of Merge Conflicts in Java Projects
Journal First
Bowen Shen Virginia Tech, Muhammad Ali Gulzar Virginia Tech, USA, Fei He Tsinghua University, Na Meng Virginia Tech
Media Attached
11:30
15m
Talk
Modeling the Centrality of Developer Output with Software Supply Chains
Industry Papers
Audris Mockus The University of Tennessee, Peter C. Rigby Meta; Concordia University, Rui Abreu Meta, Parth Suresh Meta Platforms, Inc., Yifen Chen Meta Inc, Nachiappan Nagappan Meta
DOI Pre-print Media Attached
11:45
15m
Talk
Knowledge-based Version Incompatibility Detection for Deep Learning
Research Papers
Zhongkai Zhao Tongji University, Bonan Kou Purdue University, Mohamed Yilmaz Ibrahim Purdue University, Muhao Chen University of Southern California, Tianyi Zhang Purdue University
Media Attached
12:00
15m
Talk
On-Premise AIOps Infrastructure for a Software Editor SME: An Experience Report
Industry Papers
Anes Bendimerad Infologic, Youcef REMIL INSA Lyon, INFOLOGIC, Romain Mathonat Infologic, Mehdi Kaytoue Infologic
DOI Media Attached