Grant: Alliance Grants
University of Calgary
$58,000.00
Human Centric Language Models for Issue Triaging Assistance in Java Runtime Systems
FY 2023-24
Summary
The Java runtime systems are crucial to support the development and deployment of Java applications across different computing platforms and architectures. The systems are often developed and maintained in close collaboration between big software companies (e.g., IBM, Microsoft, etc.) and the OSS (Open-Source Software) developers. Due to the needs of developing the runtime systems for various computing platforms and operating systems, the testing and maintenance of the systems can become a difficult task. Issues/bugs arise across multiple builds of the systems every day and developers need to be assigned quickly but accurately to fix those bugs. Our discussions with the IBM runtime system DevOps team finds that issue triaging can be non-trivial and time-consuming due to the following reasons: Time to Find (TTF) a bug and Time to Assign (TTA) a bug. While significant research efforts were devoted to address both challenges for software systems in general, the DevOps team finds that the current techniques can be less effective to help triaging for runtime systems due to the diversity and volume of test cases across multiple platforms and operating systems and the idiosyncrasies of builds that need to be executed daily. Large Language Models (LLMs) show promises towards supporting test case management and bug fixing. However, we are not aware of the effectiveness of the LLMs for Java runtime systems. We aim to modernize the issue triaging process for JRSs by developing a recommendation engine based on LLMs. This research will positively impact the Canadian software companies (including IBM Canada) with tools and techniques for software test analysis. The companies will be in better position to address to client needs quickly and more efficiently, which will then help them with better product and client retention management.
Program
Alliance Grants
Recipient
Location
Alberta, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
586446-2023
Name nom
Uddin, Gias G
Department département
Schulich School of Engineering
Organization id
11
Institution établissement
University of Calgary
Province en
Alberta
Province fr
Alberta
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2023
Competition year année de concours
2023
Award amount
58000
Program id
ALLRP
Program name en
Alliance Grants
Program name fr
Subventions Alliance
Group en
RESEARCH PARTNERSHIPS
Group fr
PARTENARIATS DE RECHERCHE
Committee code
96
Committee name en
RPP Internal Decision Cttee
Committee name fr
Cté de décision interne - PPR
Area of application code
801
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Computer software
Area of application fr
Logiciel
Research subject code
2804
Research subject group en
Artificial intelligence
Research subject group fr
Expert systems
Research subject en
Intelligence artificielle
Research subject fr
Systèmes experts
Application title
Human Centric Language Models for Issue Triaging Assistance in Java Runtime Systems
Keyword
Java Runtime Systems; Large Language Models; Machine Learning; Software Engineering; Software Testing
Application summary
The Java runtime systems are crucial to support the development and deployment of Java applications across different computing platforms and architectures. The systems are often developed and maintained in close collaboration between big software companies (e.g., IBM, Microsoft, etc.) and the OSS (Open-Source Software) developers. Due to the needs of developing the runtime systems for various computing platforms and operating systems, the testing and maintenance of the systems can become a difficult task. Issues/bugs arise across multiple builds of the systems every day and developers need to be assigned quickly but accurately to fix those bugs. Our discussions with the IBM runtime system DevOps team finds that issue triaging can be non-trivial and time-consuming due to the following reasons: Time to Find (TTF) a bug and Time to Assign (TTA) a bug. While significant research efforts were devoted to address both challenges for software systems in general, the DevOps team finds that the current techniques can be less effective to help triaging for runtime systems due to the diversity and volume of test cases across multiple platforms and operating systems and the idiosyncrasies of builds that need to be executed daily. Large Language Models (LLMs) show promises towards supporting test case management and bug fixing. However, we are not aware of the effectiveness of the LLMs for Java runtime systems. We aim to modernize the issue triaging process for JRSs by developing a recommendation engine based on LLMs. This research will positively impact the Canadian software companies (including IBM Canada) with tools and techniques for software test analysis. The companies will be in better position to address to client needs quickly and more efficiently, which will then help them with better product and client retention management.