Pre-conference talk for ZHANG Ting, YANG Zhou and Ratnadira WIDYASARI

Talk #1: Multi-Modal API Recommendation
by ZHANG Ting, PhD Candidate
for 30th IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER 2023)

Several approaches have been developed to automatically recommend APIs based on either a natural language query or source code context. However, none of these API recommendation approaches have utilized these two sources of information at the same time (i.e., leveraging natural language query and source code context together). In this work, we propose an approach named MulaRec, which leverages the information from natural language query (annotation) and source code context. The results confirm that our approach outperforms state-of-the-art API recommendation approaches which only leverage a single type of information as the input. Our work also demonstrates that multi-modal information can boost the performance of API recommendation approaches by 20%-50% better in terms of BLEU-score than the baselines.

Talk #2: Exploring and Repairing Gender Fairness Violations in Word Embedding-based Sentiment Analysis Model through Adversarial Patches 
by YANG Zhou, PhD Candidate
for 30th IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER 2023)

With the advancement of sentiment analysis (SA) models and their incorporation into our daily lives, fairness testing on these models is crucial, since unfair decisions can cause discrimination to a large population. Our work conducts a comprehensive empirical study to reveal the extent of fairness violations, specifically gender fairness, exhibited by popular word embedding-based SA models. We define fairness violation as the behavior in which an SA model predicts variants created from a text, which merely differ in gender classes, to have different sentiments. Realizing the importance of addressing such significant violations, we introduce adversarial patches (AP) as a way of patch generation in an automated program repair (APR) system to fix them. By adopting adversarial fine-tuning, our proposed AP reduces fairness violations by at least 25%.

Talk #3: Topic Recommendation for GitHub Repositories: How Far Can Extreme Multi-Label Learning Go?
by Ratnadira WIDYASARI, PhD Candidate
for 30th IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER 2023)

GitHub is one of the most popular platforms for version control and collaboration. In GitHub, developers are able to assign related topics to their repositories, which is helpful for finding similar repositories. However, the number of topics is large and this makes it challenging to assign a relevant set of topics to a repository. In this study, we try to address the problem of identifying the topics from a GitHub repository by treating it as an extreme multi-label learning (XML) problem. We collect data of 21K GitHub repositories containing 37K labels of topics. We evaluate multiple XML techniques, where we get the best results from ZestXML which is a combination of zero-shot and XML. The results show that ZestXML improves the baseline in terms of the average F1-score by 17.35%.

ZHANG, Ting is a Ph.D. candidate at SMU SCIS, supervised by Prof. David Lo and Prof. Lingxiao Jiang. Her research focuses on automatic software bug management, from detecting duplicate bug reports to repairing API misuse bugs.