Since its inception in the 2000s, AI for Software Engineering (AI4SE) has grown rapidly. AI in its different forms, e.g., data mining, information retrieval, machine learning, natural language processing, etc., has been demonstrated to be able to produce good results for automating many tasks, including specification mining, bug and vulnerability discovery, bug localization, duplicate bug report identification, failure detection, program repair, technical question answering, code search, and many more. AI4SE has much potential to improve software engineers’ productivity and software quality. Due to its potential, it is currently one of the most popular research areas in the software engineering field. To advance AI4SE, this keynote puts forward Human-Centered AI4SE. Without considering humans, it is easy for AI-powered tools to hinder rather than help humans in their job or introduce unwanted and unacceptable side effects. Human-centered AI4SE puts humans (i.e., software practitioners) at the forefront of the design of AI4SE tools, with the goal of amplifying and augmenting software practitioners’ capabilities. I will describe some requirements of human-centered AI4SE. Specifically, among others, the need to (i) listen to humans, (ii) learn from (and like) humans, and (iii) synergize with humans. For each requirement, I will present a reflection on the progress the AI4SE area has made over the years, including work done by our research group in Singapore. At the end of this talk, I will describe the road ahead for the above-mentioned requirements toward making AI4SE tools trustworthy, which is an essential attribute to allow them to be widely used by practitioners.
{"title":"Human-Centered AI for Software Engineering: Requirements, Reflection, and Road Ahead","authors":"D. Lo","doi":"10.1145/3578527.3581767","DOIUrl":"https://doi.org/10.1145/3578527.3581767","url":null,"abstract":"Since its inception in the 2000s, AI for Software Engineering (AI4SE) has grown rapidly. AI in its different forms, e.g., data mining, information retrieval, machine learning, natural language processing, etc., has been demonstrated to be able to produce good results for automating many tasks, including specification mining, bug and vulnerability discovery, bug localization, duplicate bug report identification, failure detection, program repair, technical question answering, code search, and many more. AI4SE has much potential to improve software engineers’ productivity and software quality. Due to its potential, it is currently one of the most popular research areas in the software engineering field. To advance AI4SE, this keynote puts forward Human-Centered AI4SE. Without considering humans, it is easy for AI-powered tools to hinder rather than help humans in their job or introduce unwanted and unacceptable side effects. Human-centered AI4SE puts humans (i.e., software practitioners) at the forefront of the design of AI4SE tools, with the goal of amplifying and augmenting software practitioners’ capabilities. I will describe some requirements of human-centered AI4SE. Specifically, among others, the need to (i) listen to humans, (ii) learn from (and like) humans, and (iii) synergize with humans. For each requirement, I will present a reflection on the progress the AI4SE area has made over the years, including work done by our research group in Singapore. At the end of this talk, I will describe the road ahead for the above-mentioned requirements toward making AI4SE tools trustworthy, which is an essential attribute to allow them to be widely used by practitioners.","PeriodicalId":326318,"journal":{"name":"Proceedings of the 16th Innovations in Software Engineering Conference","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125594115","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Digital Twins are now mainstream technology in the engineering domain. Capabilities and underpinning concepts are well understood and augmented by proven theories from the physical sciences. Nonetheless the design of digital twins in engineering still remains essential a craft. As digital twin technology merges with more traditional computational modelling approaches such as that found in simulation, new application domains are emerging and public policy experts see significant potential in DT for understanding their complex system areas. Such domains have a significant sociotechnical component and as such a new type of digital twin is required, together with a means of specifying such a digital twin. This paper proposes a specification language/method for this purpose. Requirements elicitation for this language utilises a tabletop paper template that serves as a boundary object between domain experts and technical experts. The language is conformant with accepted practice in simulation methods and its semantics provides a route to implementation of a digital twin. We argue that the language is a contribution to a breadcrumb trail for future work in this emerging application area for digital twins.
{"title":"Towards the Essence of Specifying Sociotechnical Digital Twins","authors":"B. Barn, Tony Clark, Souvik Barat, V. Kulkarni","doi":"10.1145/3578527.3578542","DOIUrl":"https://doi.org/10.1145/3578527.3578542","url":null,"abstract":"Digital Twins are now mainstream technology in the engineering domain. Capabilities and underpinning concepts are well understood and augmented by proven theories from the physical sciences. Nonetheless the design of digital twins in engineering still remains essential a craft. As digital twin technology merges with more traditional computational modelling approaches such as that found in simulation, new application domains are emerging and public policy experts see significant potential in DT for understanding their complex system areas. Such domains have a significant sociotechnical component and as such a new type of digital twin is required, together with a means of specifying such a digital twin. This paper proposes a specification language/method for this purpose. Requirements elicitation for this language utilises a tabletop paper template that serves as a boundary object between domain experts and technical experts. The language is conformant with accepted practice in simulation methods and its semantics provides a route to implementation of a digital twin. We argue that the language is a contribution to a breadcrumb trail for future work in this emerging application area for digital twins.","PeriodicalId":326318,"journal":{"name":"Proceedings of the 16th Innovations in Software Engineering Conference","volume":"150 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121261642","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Runtime enforcement is a mechanism that compels a (black-box) system to obey some expected properties. For that, it employs an enforcement monitor /enforcer which modifies an (untrusted) sequence of events into a sequence that complies with the property. In reality, we may have many critical (timed) properties to enforce. Furthermore, an ideal deployed system allows for system customization to meet the needs of the end-users. Thus, it is highly needed to build not a monolithic enforcer for all the properties, instead, there must be individual enforcers for each property and these individual enforcers should be composed accordingly. We investigate and provide a framework for composing enforcers of (timed) safety properties, formalized as timed automata, demonstrating that enforcement under this approach is not serially compositional in general. However, we identify and establish syntactic criteria on the automata, such that enforcers are serially compositional for any given set of safety timed automata satisfying these conditions. We show some examples of safety timed automata that satisfy those syntactic constraints, and via a prototype implementation, we evaluate the performance of our framework.
{"title":"Serial Compositional Runtime Enforcement of Safety Timed Properties","authors":"Saumya Shankar, Srinivas Pinisetty","doi":"10.1145/3578527.3578529","DOIUrl":"https://doi.org/10.1145/3578527.3578529","url":null,"abstract":"Runtime enforcement is a mechanism that compels a (black-box) system to obey some expected properties. For that, it employs an enforcement monitor /enforcer which modifies an (untrusted) sequence of events into a sequence that complies with the property. In reality, we may have many critical (timed) properties to enforce. Furthermore, an ideal deployed system allows for system customization to meet the needs of the end-users. Thus, it is highly needed to build not a monolithic enforcer for all the properties, instead, there must be individual enforcers for each property and these individual enforcers should be composed accordingly. We investigate and provide a framework for composing enforcers of (timed) safety properties, formalized as timed automata, demonstrating that enforcement under this approach is not serially compositional in general. However, we identify and establish syntactic criteria on the automata, such that enforcers are serially compositional for any given set of safety timed automata satisfying these conditions. We show some examples of safety timed automata that satisfy those syntactic constraints, and via a prototype implementation, we evaluate the performance of our framework.","PeriodicalId":326318,"journal":{"name":"Proceedings of the 16th Innovations in Software Engineering Conference","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133250040","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The past decade has seen unprecedented growth in Software Engineering— developers spend enormous time and effort to create new products. With such enormous growth comes the responsibility of producing and maintaining quality and robust software. However, developing such software is non-trivial— 50% of software developers’ valuable time is wasted on finding and fixing bugs, costing the global economy around USD$1.1 trillion. Today, I will discuss how AI can help in different stages of the software development life cycle for developing quality products. In particular, I will talk about Programming Language Processing (PLP), an emerging research field that can model different aspects of code (source, binary, execution, etc.) to automate diverse Software Engineering tasks, including code generation, bug finding, security analysis, etc.
{"title":"Programming Language Processing : How AI can Revolutionize Software Development?","authors":"Baishakhi Ray","doi":"10.1145/3578527.3581766","DOIUrl":"https://doi.org/10.1145/3578527.3581766","url":null,"abstract":"The past decade has seen unprecedented growth in Software Engineering— developers spend enormous time and effort to create new products. With such enormous growth comes the responsibility of producing and maintaining quality and robust software. However, developing such software is non-trivial— 50% of software developers’ valuable time is wasted on finding and fixing bugs, costing the global economy around USD$1.1 trillion. Today, I will discuss how AI can help in different stages of the software development life cycle for developing quality products. In particular, I will talk about Programming Language Processing (PLP), an emerging research field that can model different aspects of code (source, binary, execution, etc.) to automate diverse Software Engineering tasks, including code generation, bug finding, security analysis, etc.","PeriodicalId":326318,"journal":{"name":"Proceedings of the 16th Innovations in Software Engineering Conference","volume":"193 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134019708","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Recent years have witnessed tremendous progress in NLP-based code comprehension via deep neural networks (DNN) learning, especially Large Language Models (LLMs). While the original application of LLMs is focused on code generation, there have been attempts to extend the application to more specialized tasks, like code similarity, author attribution, code repairs, and so on. As data plays an important role in the success of any machine learning approach, researchers have also proposed several benchmarks which are coupled with a specific task at hand. It is well known in the machine learning (ML) community that the presence of biases in the dataset affects the quality of the ML algorithm in a real-world scenario. This paper evaluates several existing datasets from DNN’s application perspective. We specifically focus on training datasets of C/C++ language code. Our choice of language stems from the fact that while LLM-based techniques have been applied and evaluated on programming languages like Python, JavaScript, and Ruby, there is not much LLM research for C/C++. As a result, datasets generated synthetically or from real-world codes are in individual research work. Consequently, in the absence of a uniform dataset, such works are hard to compare with each other. In this work, we aim to achieve two main objectives– 1. propose code-centric features that are relevant to security program analysis tasks like vulnerability detection; 2. a thorough (qualitative and quantitative) examination of the existing code datasets that demonstrate the main characteristics of the individual datasets to have a clear comparison. Our evaluation finds exciting facts about existing datasets highlighting gaps that need to be addressed.
{"title":"A Code Centric Evaluation of C/C++ Vulnerability Datasets for Deep Learning Based Vulnerability Detection Techniques","authors":"Ridhi Jain, Nicole Gervasoni, Mthandazo Ndhlovu, Sanjay Rawat","doi":"10.1145/3578527.3578530","DOIUrl":"https://doi.org/10.1145/3578527.3578530","url":null,"abstract":"Recent years have witnessed tremendous progress in NLP-based code comprehension via deep neural networks (DNN) learning, especially Large Language Models (LLMs). While the original application of LLMs is focused on code generation, there have been attempts to extend the application to more specialized tasks, like code similarity, author attribution, code repairs, and so on. As data plays an important role in the success of any machine learning approach, researchers have also proposed several benchmarks which are coupled with a specific task at hand. It is well known in the machine learning (ML) community that the presence of biases in the dataset affects the quality of the ML algorithm in a real-world scenario. This paper evaluates several existing datasets from DNN’s application perspective. We specifically focus on training datasets of C/C++ language code. Our choice of language stems from the fact that while LLM-based techniques have been applied and evaluated on programming languages like Python, JavaScript, and Ruby, there is not much LLM research for C/C++. As a result, datasets generated synthetically or from real-world codes are in individual research work. Consequently, in the absence of a uniform dataset, such works are hard to compare with each other. In this work, we aim to achieve two main objectives– 1. propose code-centric features that are relevant to security program analysis tasks like vulnerability detection; 2. a thorough (qualitative and quantitative) examination of the existing code datasets that demonstrate the main characteristics of the individual datasets to have a clear comparison. Our evaluation finds exciting facts about existing datasets highlighting gaps that need to be addressed.","PeriodicalId":326318,"journal":{"name":"Proceedings of the 16th Innovations in Software Engineering Conference","volume":"92 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128096955","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The video game development industry, also known as the interactive entertainment business, is involved in building, marketing, advertising, and monetizing video games. Over the last few years, this industry has come into the mainstream from initially being in the focused market. The growing video gamer demographic has increased by video game development methods and techniques. A postmortem of a video game is a close examination of the video game’s development phase and an analysis of what went right or wrong with the video game. Unfortunately, since there is not much understanding regarding the challenges encountered by programmers, there is a lack of trustworthiness primarily because postmortems lack a proper structure and are informally written. In this work, with the help of word embeddings and ensemble machine learning classifiers, a systematic analysis is performed on various technical and non-technical issues faced by the video game industry. It is believed that automation and machine learning classifiers could aid game developers in identifying what problem they are facing, given the quote (description), and thus be able to figure out a solution quickly. Frequently committed mistakes could be identified and avoided, and this work could act as a starting point to further consider software development and video game development in the same context.
{"title":"Automatic Identification of Video Game Development Problems using Word Embedding and Ensemble Classifiers","authors":"Anirudh, L. Kumar, N. B. Murthy, A. Krishna","doi":"10.1145/3578527.3578543","DOIUrl":"https://doi.org/10.1145/3578527.3578543","url":null,"abstract":"The video game development industry, also known as the interactive entertainment business, is involved in building, marketing, advertising, and monetizing video games. Over the last few years, this industry has come into the mainstream from initially being in the focused market. The growing video gamer demographic has increased by video game development methods and techniques. A postmortem of a video game is a close examination of the video game’s development phase and an analysis of what went right or wrong with the video game. Unfortunately, since there is not much understanding regarding the challenges encountered by programmers, there is a lack of trustworthiness primarily because postmortems lack a proper structure and are informally written. In this work, with the help of word embeddings and ensemble machine learning classifiers, a systematic analysis is performed on various technical and non-technical issues faced by the video game industry. It is believed that automation and machine learning classifiers could aid game developers in identifying what problem they are facing, given the quote (description), and thus be able to figure out a solution quickly. Frequently committed mistakes could be identified and avoided, and this work could act as a starting point to further consider software development and video game development in the same context.","PeriodicalId":326318,"journal":{"name":"Proceedings of the 16th Innovations in Software Engineering Conference","volume":"129 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121157158","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Unsafe programming languages like C/C++ lack efficient memory management module. A programmer is privileged to do explicit allocation and deallocation of heap memory blocks in C/C++ programming languages. Memory errors related to heap memory are difficult to capture using software testing in such programming languages. Program analysis can improve the reliability of software by identifying and repairing bugs related to dynamic memory management. An efficient static analyzer needs to look at the control flow graph of the whole program binary. Capturing context, path, and flow sensitivity is challenging using abstract interpretation and can lead to over approximated results. We propose Heap Object Flow Graph (HOFG), a program representation that eases program analysis to detect memory errors using static program analysis. An efficient program analysis on HOFG called HOFG-Analyser is defined, to capture memory leaks. We were able to detect memory leaks on different benchmarks written in C/C++ programming language with sizes up to 521K Lines of Code (LoC) with minimal false positive rates. Our experimental evaluation show that HOFG-Analyser is efficient and effective than an existing static analysis tool, INFER. In many cases, more number of leaks and less false positive rate is achieved by the HOFG-Analyser.
{"title":"Memory leak detection using Heap Object Flow Graph (HOFG)","authors":"Haritha Madhav C, Unnikrishnan Cheramangalath","doi":"10.1145/3578527.3578528","DOIUrl":"https://doi.org/10.1145/3578527.3578528","url":null,"abstract":"Unsafe programming languages like C/C++ lack efficient memory management module. A programmer is privileged to do explicit allocation and deallocation of heap memory blocks in C/C++ programming languages. Memory errors related to heap memory are difficult to capture using software testing in such programming languages. Program analysis can improve the reliability of software by identifying and repairing bugs related to dynamic memory management. An efficient static analyzer needs to look at the control flow graph of the whole program binary. Capturing context, path, and flow sensitivity is challenging using abstract interpretation and can lead to over approximated results. We propose Heap Object Flow Graph (HOFG), a program representation that eases program analysis to detect memory errors using static program analysis. An efficient program analysis on HOFG called HOFG-Analyser is defined, to capture memory leaks. We were able to detect memory leaks on different benchmarks written in C/C++ programming language with sizes up to 521K Lines of Code (LoC) with minimal false positive rates. Our experimental evaluation show that HOFG-Analyser is efficient and effective than an existing static analysis tool, INFER. In many cases, more number of leaks and less false positive rate is achieved by the HOFG-Analyser.","PeriodicalId":326318,"journal":{"name":"Proceedings of the 16th Innovations in Software Engineering Conference","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133524427","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
At the current speed of technological advancement, particularly with the pandemic driving the shift from physical to virtual, everyone must have access to the digital world. Guidelines for accessible technology exist to assist software engineers in creating websites accessible to disabled populations but are extensive and, therefore, difficult to follow completely. Additionally, no standard guidelines are available to address a wider range of human diversity beyond disabilities. In this paper, we present a browser extension tool, Inclusivity Checker, that can assist in building more inclusive websites by performing tests on checkpoints categorized into various user groups based on disability type and diversity parameters. These checkpoints are based on existing guidelines and present developers with errors, warnings, and tips. To gauge the efficiency of the developed tool, we evaluated the results manually. Furthermore, we collected the developers’ feedback on the tool to test its usability.
{"title":"Inclusivity Checker: A Testing Tool to Detect Inclusivity Bugs in Websites","authors":"Kushagra Pathak, Parita Patel, Mital Kamani, Saurabh Tiwari","doi":"10.1145/3578527.3578547","DOIUrl":"https://doi.org/10.1145/3578527.3578547","url":null,"abstract":"At the current speed of technological advancement, particularly with the pandemic driving the shift from physical to virtual, everyone must have access to the digital world. Guidelines for accessible technology exist to assist software engineers in creating websites accessible to disabled populations but are extensive and, therefore, difficult to follow completely. Additionally, no standard guidelines are available to address a wider range of human diversity beyond disabilities. In this paper, we present a browser extension tool, Inclusivity Checker, that can assist in building more inclusive websites by performing tests on checkpoints categorized into various user groups based on disability type and diversity parameters. These checkpoints are based on existing guidelines and present developers with errors, warnings, and tips. To gauge the efficiency of the developed tool, we evaluated the results manually. Furthermore, we collected the developers’ feedback on the tool to test its usability.","PeriodicalId":326318,"journal":{"name":"Proceedings of the 16th Innovations in Software Engineering Conference","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134444858","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Cloud is a distributed ecosystem and differs significantly from on-premise software development platform. Cloud application development is built upon a service-based architecture, application programming interface driven communications, container-based infrastructure and a bias for DevOps process such as continuous improvement, agile development, continuous delivery and collaborative development among developers, quality assurance teams, security professionals, IT operations and line-of-business stakeholders. Therefore, while building applications on the cloud there needs to be a novel attitude to requirements gathering, software design, development, deployment, debugging, maintenance and testing. The main objective of the workshop is to discuss how Cloud Software Engineering differs from traditional software engineering and the challenges that arise and create a community around the relevant work areas.
{"title":"Cloud Software Engineering","authors":"Rupashree Rangaiyengar","doi":"10.1145/3578527.3581746","DOIUrl":"https://doi.org/10.1145/3578527.3581746","url":null,"abstract":"Cloud is a distributed ecosystem and differs significantly from on-premise software development platform. Cloud application development is built upon a service-based architecture, application programming interface driven communications, container-based infrastructure and a bias for DevOps process such as continuous improvement, agile development, continuous delivery and collaborative development among developers, quality assurance teams, security professionals, IT operations and line-of-business stakeholders. Therefore, while building applications on the cloud there needs to be a novel attitude to requirements gathering, software design, development, deployment, debugging, maintenance and testing. The main objective of the workshop is to discuss how Cloud Software Engineering differs from traditional software engineering and the challenges that arise and create a community around the relevant work areas.","PeriodicalId":326318,"journal":{"name":"Proceedings of the 16th Innovations in Software Engineering Conference","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132963966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sharath H. Padmanabha, Fahad Shaikh, Mayank Bansal, Debanjan Chatterjee, Preeti Singh, Amey Karkare, Purushottam Kar
We report the design, development and deployment of PRIORITY, an intelligent portal aimed at reducing the workload of instructors, tutors and teaching assistants in large programming courses of creating lab, assignment and exam problems every week. PRIORITY offers a scalable, user friendly and indexed repository of problems that can be queried to retrieve problems related to a particular programming concept, say for loops. PRIORITY accomplishes this by casting problem retrieval as a multi-label learning problem and using solving it using novel feature selection and AI-techniques. We also report the results of an A/B test and user survey, both conducted while PRIORITY was being used to offer a CS1 course taught at IIT Kanpur with over 500 students. PRIORITY has been in deployment at IIT Kanpur for almost 2 years now and our experience thus far suggests that it not only presents a valuable tool for course administrators, but also opens up several intriguing problems at the intersection of programming instruction, pedagogy, machine learning, semi-supervised learning and information retrieval. Code for PRIORITY is available at https://github.com/purushottamkar/priority/
{"title":"PRIORITY: An Intelligent Problem Indicator Repository","authors":"Sharath H. Padmanabha, Fahad Shaikh, Mayank Bansal, Debanjan Chatterjee, Preeti Singh, Amey Karkare, Purushottam Kar","doi":"10.1145/3578527.3578533","DOIUrl":"https://doi.org/10.1145/3578527.3578533","url":null,"abstract":"We report the design, development and deployment of PRIORITY, an intelligent portal aimed at reducing the workload of instructors, tutors and teaching assistants in large programming courses of creating lab, assignment and exam problems every week. PRIORITY offers a scalable, user friendly and indexed repository of problems that can be queried to retrieve problems related to a particular programming concept, say for loops. PRIORITY accomplishes this by casting problem retrieval as a multi-label learning problem and using solving it using novel feature selection and AI-techniques. We also report the results of an A/B test and user survey, both conducted while PRIORITY was being used to offer a CS1 course taught at IIT Kanpur with over 500 students. PRIORITY has been in deployment at IIT Kanpur for almost 2 years now and our experience thus far suggests that it not only presents a valuable tool for course administrators, but also opens up several intriguing problems at the intersection of programming instruction, pedagogy, machine learning, semi-supervised learning and information retrieval. Code for PRIORITY is available at https://github.com/purushottamkar/priority/","PeriodicalId":326318,"journal":{"name":"Proceedings of the 16th Innovations in Software Engineering Conference","volume":"35 9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123517817","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：