Overview

Recently, deep learning (DL) and machine learning (ML) methods have been massively and successfully applied in various software engineering (SE) and program languages (PL) tasks. The results are promising and exciting, and lead to further opportunities of exploring the amenability of DL and ML to different SE and PL tasks. Notably, the choice of the representations on which DL and ML methods are applied critically impacts the performance of the DL and ML methods. The rapidly developing field of representation learning (RL) in artificial intelligence is concerned with questions surrounding how we can best learn meaningful and useful representations of data. A broad view of the RL in SE & PL can include the topics, e.g., deep learning, feature learning, compositional modeling, structured prediction, and reinforcement learning.

The workshop will advance the pace of research in the unique intersection of representation learning and SE & PL, which will, in the long term, lead to more effective solutions to common software engineering tasks such as coding, maintenance, testing, and porting.

Due to the unforeseen effects of the Covid 19 pandemic, ESEC/FSE 2020 and RL+SE&PL 2020 will be held virtually. More information to come on the ESEC/FSE 2020 website

Call For Participation

The main goal of the 1st RL+SE&PL workshop is to accelerate the exposure of the SE and PL community to the emerging and fast growing AI research in representation learning that plays as the key center in AI, leading to research results, techniques and perspectives that challenge the status quo in the discipline.

TOPICS OF INTEREST Every type of software artifact (e.g., Code, Documentation, Software requirements, commits, logs) can have different characteristics, features and structures that require more effort. These observations indicate further opportunities in a wide set of topics, but not limited to,
  • Learning representations for code clone detection
  • Learning to model code for vulnerability detection and fixing
  • API knowledge graph representation learning
  • Representation learning for fault localization and testing
  • Learning alignments between natural language (e.g., comments) and code, Cross-language porting
  • Recovering useful facts (for question answering, etc)
  • Detecting buggy or anomalous code
  • Learning representations for automatic code modification: for defect repair, reformatting, and refactoring

We aim to build a vibrant forum for forward-looking, innovative research in learning representations in software engineering and programming languages. We welcome the following types of submissions:
1. Abstracts/Posters/Presentations (up to 2 pages, including references)
2. Short papers (up to 4 pages, including references), e.g., tools, demos, position statements, technical reports.
3. Technical papers (up to 10 pages, including references)

Submissions will be reviewed primarily for relevance, will not appear in ACM Digital Library, and may be published subsequently elsewhere.

All of the accepted submissions (all 3 types) will be part of the ESEC/FSE proceedings under the copyright of the ACM digital library.

All types of the accepted submissions will be invited for presentation.

Submissions will be handled via EasyChair: https://easychair.org/conferences/?conf=rlsepl2020

Important Dates

Dates:
  • July 26th, 2020 Extended to August 21th, 2020: Workshop papers submission
  • August 26th, 2020 Extended to September 7th, 2020: Workshop papers notification
  • September 18th, 2020: Workshop papers camera-ready (hard)
  • November 9th, 2020: RL+SE&PL Day

Program

Time is based on Central Daylight Time (Dallas time)
Nov. 09th, Monday:
  • 9:00 a.m. - 9:10 a.m.       Opening

  • 9:10 a.m. - 9:55 a.m.       Keynote Speaker: Miltos Allamanis, Microsoft Research in Cambridge, UK
    Title: Graph Neural Networks in Software Engineering Research
    Abstract: Many software engineering artifacts can be efficiently represented as graph structures. During the last few years, machine learning research has popularized Graph Neural Networks (GNN), neural networks that accepts graphs as inputs. Since then, GNNs have been commonly used to learn representation of source code for a variety of tasks. In this talk, I will give a brief overview of graph neural networks and then discuss software engineering research based on GNNs and graph representations of source code. Finally, I will conclude with open problems, practical considerations, and future directions.

  • 9:55 a.m. - 10:00 a.m.     Q&A for Keynote

  • 10:00 a.m. - 10:10 a.m.    Short Break

  • 10:10 a.m. - 10:30 a.m.   A Differential Evolution-Based Approach for Effort-Aware Just-in-Time Software Defect Prediction
    Authors: Xingguang Yang(East China University of Science and Technology), Huiqun Yu(East China University of Science and Technology), Guisheng Fan(East China University of Science and Technology), KangYang(East China University of Science and Technology)

  • 10:30 a.m. - 10:50 a.m.   A Multi-Task Representation Learning Approach for Source Code
    Authors: Deze Wang(National University of Defense Technology), Wei Dong(National University of Defense Technology), Shanshan Li(National University of Defense Technology)

  • 10:50 a.m. - 11:10 a.m.   When Representation Learning Meets Software Analysis
    Authors: Ming Fan(Xi'an Jiaotong University), Ang Jia(Xi'an Jiaotong University), Jingwen Liu(Xi'an Jiaotong University), Ting Liu(Xi'an Jiaotong University), Wei Chen(State Grid Shaanxi Electric Power Research Institute)

  • 11:10 a.m. - 11:30 a.m.   Statistical Machine Translation Outperforms Neural Machine Translation in Software Engineering: Why and How
    Authors: Hung Phan(Iowa State University), Ali Jannesari(Iowa State University)

  • 11:30 a.m. - 11:50 a.m.   Towards Demystifying Dimensions of Source Code Embeddings
    Authors: Md Rafiqul Islam Rabin(University of Houston), Arjun Mukherjee(University of Houston), Omprakash Gnawali(University of Houston), Mohammad Amin Alipour(University of Houston)

  • 11:50 a.m. - 12:10 p.m.   Boosting Component-Based Synthesis with Control Structure Recommendation
    Authors: Binbin Liu(National University of Defense Technology), Wei Dong(National University of Defense Technology), Yating Zhang(National University of Defense Technology), Daiyan Wang(National University of Defense Technology), Jiaxin Liu(National University of Defense Technology)

  • 12:10 p.m. - 12:20 p.m.    Closing: Representation Learning for Software Engineering and Programming Languages
    Presenters: Shaohua Wang(New Jersey Institute of Technology), Tien N. Nguyen(The University of Texas at Dallas)

Organizers

Program Committee

Program Chairs:
Dr. Shaohua WangNew Jersey Institute of Technology, USA
Dr. Tien N. NguyenThe University of Texas at Dallas, USA
Program Committee:
Dr. Hoa Khanh DamUniversity of Wollongong, Australia
Dr. Jiaping GuiNEC America, USA
Dr. Heng LiPolytechnique Montreal, Canada
Dr. Tien N. NguyenUniversity of Texas at Dallas, USA
Dr. Michael PradelUniversity of Stuttgart, Germany
Dr. Weiyi ShangConcordia University, Canada
Dr. Yuan TianQueen's University, Canada
Dr. Shaohua WangNew Jersey Institute of Technology, USA

Contact

For questions or comments about the workshop, please contact program chair Shaohua Wang or web chair Jiahao Fan