Moshi Wei

Moshi Wei

Software Developer, Researcher

Bio

Hello, I am Moshi, a Ph.D. student in Software Engineering from York University. My research focuses on addressing the usability and reliability issues of deep-learning APIs, such as recommending and detecting misuse of deep-learning APIs. During my Ph.D. study, I have published several papers on deep-learning API recommendation and misuse detection in top-tier software engineering conferences (ICSE, ASE) and journals. Besides my research experience, I also have experience working in industry companies such as Huawei Canada and Ericsson. I am currently looking for professorial positions in Canada to continue my research on API usability and reliability.

Highlights

  • Publications in top venues: 3 ICSE, 1 FSE, 1 ISSTA
  • Research Interests: software engineering, machine learning, software reliability, software defect prediction, and software testing. Focusing on using deep learning, and language model techniques to improve system usability and reliability.

Education

Ph.D., Software Engineering
York University, Toronto, Ontario, Canada
Thesis: Enhancing Usability and Reliability in Deep Learning APIs: Misuse Detection, Patching, and Recommendations
Supervisor: Dr. Song Wang
Sep. 2020 – May. 2024

M.Sc., Software Engineering
University of Waterloo, Waterloo, Ontario, Canada
Supervisor: Dr. Lin Tan
Dec. 2017 – May. 2019

B.Sc., Computer Science
University of Ottawa, Ottawa, Ontario, Canada
Sep. 2012 – May. 2016

Publication

  • P1 Moshi Wei, Nima Shiri Harzevili, Yuekai Huang, Jinqiu Yang, Junjie Wang, and Song Wang.
    Demystifying and Detecting Misuses of Deep Learning APIs.
    44th International Conference on Software Engineering (ICSE 2024) 13 pages, CORE A*.

  • P2 Jiho Shin, Moshi Wei, Junjie Wang, Lin Shi, and Song Wang.
    The Good, the Bad, and the Missing: Neural Code Generation for Machine Learning Tasks.
    ACM Transactions on Software Engineering and Methodology (TOSEM’23). 20 pages, CORE A*.

  • P3 Omar Alhussein, Moshi Wei, Arashmid Akhavain.
    Dynamic Encoding and Decoding of Information for Split Learning in Mobile-Edge Computing: Leveraging Information Bottleneck Theory.
    IEEE Global Communications Conference (Globecom’23).

  • P4 Moshi Wei, Yuchao Huang, Jinqiu Yang, Junjie Wang, and Song Wang.
    CoCoFuzzing: Testing Neural Code Models with Coverage-Guided Fuzzing.
    IEEE Transactions on Reliability (TR’22). 14 pages, CORE A.(Cited 10+ Times)

  • P5 Moshi Wei, Yuchao Huang, Junjie Wang, Jiho Shin, Shiri harzevili Nima, and Song Wang.
    API Recommendation for Machine Learning Libraries: How Far Are We?.
    The 15th Joint Meeting of the European Software Engineering Conference and the ACM SIGSOFT Symposium on the Foundations of Software Engineering (ESEC/FSE 2022). 12 pages, CORE A*.

  • P6 Moshi Wei, Shiri harzevili Nima, Yuchao Huang, Junjie Wang, and Song Wang.
    CLEAR: Contrastive Learning for API Recommendation.
    44th International Conference on Software Engineering (ICSE2022) (ICSE 2022). 12 pages, CORE A*. (Cited 23+ Times)

  • P7 Yuchao Huang, Moshi Wei, Song Wang, Junjie Wang, Qing Wang.
    Yet Another Combination of IR-and Neural-based Comment Generation.
    Information and Software Technology 2022 (IST 2022). 11 pages, CORE A.

  • P8 Song Wang, Nishtha Shrestha, Abarna Kucheri Subburaman, Junjie Wang,Moshi Wei, and Nachiappan Nagappan.
    Automatic Unit Test Generation for Machine Learning Libraries: How Far Are We?.
    43rd International Conference on Software Engineering (ICSE 2021). 12 pages, CORE A*. (Cited 24+ Times)

  • P9 Thibaud Lutellier, Hung Viet Pham, Lawrence Pang, Yitong Li, Moshi Wei, Lin Tan.
    Coconut: combining context-aware neural translation models using ensemble for program repair.
    Proceedings of the 29th ACM SIGSOFT international symposium on software testing and analysis (ISSTA 2020). 12 pages, CORE A*. (Cited 236+ Times)

Industrial Experience

Research Intern - International Strategy - Research Nov. 2023 – Now\

Huawei Technologies Canada, Toronto, Ontario, CA\

  • Responsible for advanced Software AI Agent research.

Research Intern - Advanced Wireless Capability - Jan. 2023 – May. 2023\

Huawei Technologies Canada, Ottawa, Ontario, CA\

  • Developed deep learning model for 6G signal transmission prototype.\
  • Explored the impact of the Large Language Model on enterprise-level network configuration software.

Research Intern - R&D - Nov. 2021 – Jan. 2022\

Ericsson Software, Toronto, Ontario, CA\

  • Developed a prototype for a low-code application development platform for network service providers.

Machine Learning Engineer - Business Intelligence - May. 2019 – Jul. 2020\

Achievers, Toronto, Ontario, CA\

  • Designed and implemented a recommendation system for user-to-user recommendation.\
  • Maintained a data warehouse for a professional development network application with employee data from 300+ public companies.

Research Experience

Graduate Research Assistant at York University - Sep. 2020 – Now

  • Conducted research on the API recommendation model which became the State-of-the-art tool in 2022
  • Conducted research on applying the Large Language Model for deep learning API misuse detection and repair
  • Conducted research on software defect prediction and test case generation
  • Mentored students in the research lab
  • Published research findings in top software engineering conferences and journals
  • Peer-reviewed papers for software engineering conferences and journals

    Research Assistant at University of Waterloo - Dec. 2017 – May. 2019

  • Conducted research on automated bug repair
  • Published research findings in top software engineering conferences and journals
  • Peer-reviewed papers for software engineering conferences and journals