Grant: Discovery Grants Program - Individual
York University
$29,000.00
Knowledge Enhanced Software Library Documentation
FY 2024-25
Summary
Software is ubiquitous. We rely on software to drive the economy and to improve quality of life. Software errors can have significant impact in our daily life. In 2017, 606 software bugs cost around $1.7T and affected 3.7B people. Errors can be fatal, such as software bugs in self-driving cars and Boeing-737 Max flight tragedy resulting in passenger deaths. As such, we must educate software developers to produce reliable software efficiently. Modern day rapid software development is facilitated by the reuse of APIs (Application Programming Interfaces). Unfortunately, several research results show that official API documentation can be incomplete, obsolete, and incorrect. Developers rely on API usage knowledge shared in online forums like Stack Overflow (SO) to compensate for the shortcomings in official API documentation. The online Stack Exchange network that hosts SO had 9+ billion-page views from 100+ million users in 2018 alone. However, there are three major problems with the reuse of crowd-shared API knowledge: volume (i.e., big data), quality, and trust.
�Volume: The large volume of shared knowledge drives the needs to summarize it because developers struggle with the huge volume and scattered nature of the information. Among the numerous online platforms, SO alone has over 120M posts.
�Quality: The lack of quality of the shared code examples necessitates the needs for techniques to detect and warn about the issues. Studies find API misuse patterns and security vulnerabilities in C++ and Android examples, yet they were reused in popular mobile apps and GitHub repositories.
�Trust: Lack of trust and bias in a user's API review may negatively and wrongly influence the decision of others.
The long-term vision of this research program is to improve API documentation found in both formal (i.e., official) and informal resources (i.e., developer forums). I have two objectives (O):
[O1]. I aim to address trust, quality, and volume issues in online developer forums by designing automated techniques. The techniques can guide software developers during the reuse of API usage scenarios (e.g., by warning them of insecure code) and can be used to extract high quality data.
[O2]. I aim to improve API official documentation by automatically detecting and fixing problems in the documentation. The fixes can use high quality API knowledge extracted in O1.
The techniques will be deployed into Opiner, an online portal for API documentation. The program will train 11 HQPs (2 PhD, 4 MSc, 5 BSc), who will learn valuable Machine Learning, Natural Language Processing, and software development skills. The techniques can be used to produce high quality data from online forums. Canadian software companies can utilize the tools to improve their API documentation and to support developers' productivity, who spend 20% of their time searching for solutions on the web. The knowledge gained is transferable across domains under theme `Big Data Analytics'.
Program
Discovery Grants Program - Individual
Recipient
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
RGPIN-2021-02575
Name nom
Uddin, Gias
Department département
Electrical Engineering & Computer Science - Electrical Engineering & Computer Science
Organization id
38
Institution établissement
York University
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2021
Award amount
29000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1507
Committee name en
Computer Science
Committee name fr
Sciences informatiques
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
2706
Research subject group en
Information technology
Research subject group fr
Technologie de l'information
Research subject en
Software engineering
Research subject fr
Génie logiciel
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Knowledge Enhanced Software Library Documentation
Keywords
software documentation; documentation quality and problems; natural language processing; crowd-sourced knowledge; application programming interfaces (apis); information summarization; empirical software engineering; trust and bias in software review; machine learning; code example quality and security
Application summary
Software is ubiquitous. We rely on software to drive the economy and to improve quality of life. Software errors can have significant impact in our daily life. In 2017, 606 software bugs cost around $1.7T and affected 3.7B people. Errors can be fatal, such as software bugs in self-driving cars and Boeing-737 Max flight tragedy resulting in passenger deaths. As such, we must educate software developers to produce reliable software efficiently. Modern day rapid software development is facilitated by the reuse of APIs (Application Programming Interfaces). Unfortunately, several research results show that official API documentation can be incomplete, obsolete, and incorrect. Developers rely on API usage knowledge shared in online forums like Stack Overflow (SO) to compensate for the shortcomings in official API documentation. The online Stack Exchange network that hosts SO had 9+ billion-page views from 100+ million users in 2018 alone. However, there are three major problems with the reuse of crowd-shared API knowledge: volume (i.e., big data), quality, and trust.
�Volume: The large volume of shared knowledge drives the needs to summarize it because developers struggle with the huge volume and scattered nature of the information. Among the numerous online platforms, SO alone has over 120M posts.
�Quality: The lack of quality of the shared code examples necessitates the needs for techniques to detect and warn about the issues. Studies find API misuse patterns and security vulnerabilities in C++ and Android examples, yet they were reused in popular mobile apps and GitHub repositories.
�Trust: Lack of trust and bias in a user's API review may negatively and wrongly influence the decision of others.
The long-term vision of this research program is to improve API documentation found in both formal (i.e., official) and informal resources (i.e., developer forums). I have two objectives (O):
[O1]. I aim to address trust, quality, and volume issues in online developer forums by designing automated techniques. The techniques can guide software developers during the reuse of API usage scenarios (e.g., by warning them of insecure code) and can be used to extract high quality data.
[O2]. I aim to improve API official documentation by automatically detecting and fixing problems in the documentation. The fixes can use high quality API knowledge extracted in O1.
The techniques will be deployed into Opiner, an online portal for API documentation. The program will train 11 HQPs (2 PhD, 4 MSc, 5 BSc), who will learn valuable Machine Learning, Natural Language Processing, and software development skills. The techniques can be used to produce high quality data from online forums. Canadian software companies can utilize the tools to improve their API documentation and to support developers' productivity, who spend 20% of their time searching for solutions on the web. The knowledge gained is transferable across domains under theme `Big Data Analytics'.