Testing is critical to ensure the quality of widely-used web APIs. Automatic test data generation can help to reduce cost and improve overall effectiveness. This is commonly accomplished by using the powerful concept of search-based software testing (SBST). However, with web APIs growing larger and larger, SBST techniques face scalability challenges. This paper introduces a novel SBST based approach for generating API test data using deep reinforcement learning (DRL) as the search algorithm. By exploring the benefits of DRL in the context of scalable API test data generation, we show its potential as alternative to traditional search algorithms.
{"title":"Generating API Test Data Using Deep Reinforcement Learning","authors":"Steyn Huurman, Xiaoying Bai, Thomas Hirtz","doi":"10.1145/3387940.3392214","DOIUrl":"https://doi.org/10.1145/3387940.3392214","url":null,"abstract":"Testing is critical to ensure the quality of widely-used web APIs. Automatic test data generation can help to reduce cost and improve overall effectiveness. This is commonly accomplished by using the powerful concept of search-based software testing (SBST). However, with web APIs growing larger and larger, SBST techniques face scalability challenges. This paper introduces a novel SBST based approach for generating API test data using deep reinforcement learning (DRL) as the search algorithm. By exploring the benefits of DRL in the context of scalable API test data generation, we show its potential as alternative to traditional search algorithms.","PeriodicalId":309659,"journal":{"name":"Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114304300","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}
Despite the proliferation of software ecosystems (SECOs), growing a sustainable and healthy SECO remains a significant challenge. One approach to mitigate this challenge is the utilization of a mechanism that collects feedback from distributors (distros) and end-users of the SECO releases. This presentation aims at investigating the effectiveness of the feedback mechanism implemented by OpenStack to address the needs of end-users and distros. I mined the OpenStack repositories and mapped 20 distros' bug-related activities. Results suggest that OpenStack releases are actively maintained for 18 months before reaching end-of-life (EOL), which makes coordination with distros difficult because distros usually provide services to their end-users for a period between 36 - 60 months before reaching EOL. Also, bugs are fixed faster by the distros (7 - 76 days) than the OpenStack community (average of 4 months). However, only 22% of the bugs addressed by OpenStack distros are pushed back upstream.
{"title":"Cross-distribution Feedback in Software Ecosystems","authors":"A. Foundjem","doi":"10.1145/3387940.3392188","DOIUrl":"https://doi.org/10.1145/3387940.3392188","url":null,"abstract":"Despite the proliferation of software ecosystems (SECOs), growing a sustainable and healthy SECO remains a significant challenge. One approach to mitigate this challenge is the utilization of a mechanism that collects feedback from distributors (distros) and end-users of the SECO releases. This presentation aims at investigating the effectiveness of the feedback mechanism implemented by OpenStack to address the needs of end-users and distros. I mined the OpenStack repositories and mapped 20 distros' bug-related activities. Results suggest that OpenStack releases are actively maintained for 18 months before reaching end-of-life (EOL), which makes coordination with distros difficult because distros usually provide services to their end-users for a period between 36 - 60 months before reaching EOL. Also, bugs are fixed faster by the distros (7 - 76 days) than the OpenStack community (average of 4 months). However, only 22% of the bugs addressed by OpenStack distros are pushed back upstream.","PeriodicalId":309659,"journal":{"name":"Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116719186","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}
Software fault prediction is an important and beneficial practice for improving software quality and reliability. The ability to predict which components in a large software system are most likely to contain the largest numbers of faults in the next release helps to better manage projects, including early estimation of possible release delays, and affordably guide corrective actions to improve the quality of the software. However, developing robust fault prediction models is a challenging task and many techniques have been proposed in the literature. Traditional software fault prediction studies mainly focus on manually designing features (e.g. complexity metrics), which are input into machine learning classifiers to identify defective code. However, these features often fail to capture the semantic and structural information of programs. Such information is needed for building accurate fault prediction models. In this survey, we discuss various approaches in fault prediction, also explaining how in recent studies deep learning algorithms for fault prediction help to bridge the gap between programs' semantics and fault prediction features and make accurate predictions.
{"title":"Deep Learning for Software Defect Prediction: A Survey","authors":"Safa Omri, C. Sinz","doi":"10.1145/3387940.3391463","DOIUrl":"https://doi.org/10.1145/3387940.3391463","url":null,"abstract":"Software fault prediction is an important and beneficial practice for improving software quality and reliability. The ability to predict which components in a large software system are most likely to contain the largest numbers of faults in the next release helps to better manage projects, including early estimation of possible release delays, and affordably guide corrective actions to improve the quality of the software. However, developing robust fault prediction models is a challenging task and many techniques have been proposed in the literature. Traditional software fault prediction studies mainly focus on manually designing features (e.g. complexity metrics), which are input into machine learning classifiers to identify defective code. However, these features often fail to capture the semantic and structural information of programs. Such information is needed for building accurate fault prediction models. In this survey, we discuss various approaches in fault prediction, also explaining how in recent studies deep learning algorithms for fault prediction help to bridge the gap between programs' semantics and fault prediction features and make accurate predictions.","PeriodicalId":309659,"journal":{"name":"Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115841855","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}
While Search-Based Software Testing (SBST) has improved significantly in the last decade we propose that more flexible, probabilistic models can be leveraged to improve it further. Rather than searching for an individual, or even sets of, test case(s) or datum(s) that fulfil specific needs the goal can be to learn a generative model tuned to output a useful family of values. Such generative models can naturally be decomposed into a structured generator and a probabilistic model that determines how to make non-deterministic choices during generation. While the former constrains the generation process to produce valid values the latter allows learning and tuning to specific goals. SBST techniques differ in their level of integration of the two but, regardless of how close it is, we argue that the flexibility and power of the probabilistic model will be a main determinant of success. In this short paper, we present how some existing SBST techniques can be viewed from this perspective and then propose additional techniques for flexible generative modelling the community should consider. In particular, Probabilistic Programming languages (PPLs) and Genetic Programming (GP) should be investigated since they allow for very flexible probabilistic modelling. Benefits could range from utilising the multiple program executions that SBST techniques typically require to allowing the encoding of high-level test strategies.
{"title":"Flexible Probabilistic Modeling for Search Based Test Data Generation","authors":"R. Feldt, S. Yoo","doi":"10.1145/3387940.3392215","DOIUrl":"https://doi.org/10.1145/3387940.3392215","url":null,"abstract":"While Search-Based Software Testing (SBST) has improved significantly in the last decade we propose that more flexible, probabilistic models can be leveraged to improve it further. Rather than searching for an individual, or even sets of, test case(s) or datum(s) that fulfil specific needs the goal can be to learn a generative model tuned to output a useful family of values. Such generative models can naturally be decomposed into a structured generator and a probabilistic model that determines how to make non-deterministic choices during generation. While the former constrains the generation process to produce valid values the latter allows learning and tuning to specific goals. SBST techniques differ in their level of integration of the two but, regardless of how close it is, we argue that the flexibility and power of the probabilistic model will be a main determinant of success. In this short paper, we present how some existing SBST techniques can be viewed from this perspective and then propose additional techniques for flexible generative modelling the community should consider. In particular, Probabilistic Programming languages (PPLs) and Genetic Programming (GP) should be investigated since they allow for very flexible probabilistic modelling. Benefits could range from utilising the multiple program executions that SBST techniques typically require to allowing the encoding of high-level test strategies.","PeriodicalId":309659,"journal":{"name":"Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115284707","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}
Cyrine Zid, D. Humeniuk, Foutse Khomh, G. Antoniol
Testing heterogeneous IoT applications such as a home automation systems integrating a variety of devices poses serious challenges. Oftentimes requirements are vaguely defined. Consumer grade cyber-physical devices and software may not meet the reliability and quality standard needed. Plus, system behavior may partially depend on various environmental conditions. For example, WI-FI congestion may cause packet delay; meanwhile cold weather may cause an unexpected drop of inside temperature. We surmise that generating and executing failure exposing scenarios is especially challenging. Modeling phenomenons such as network traffic or weather conditions is complex. One possible solution is to rely on machine learning models approximating the reality. These models, integrated in a system model, can be used to define surrogate models and fitness functions to steer the search in the direction of failure inducing scenarios. However, these models also should be validated. Therefore, there should be a double loop co-evolution between machine learned surrogate models functions and fitness functions. Overall, we argue that in such complex cyber-physical systems, co-evolution and multi-hybrid approaches are needed.
{"title":"Double Cycle Hybrid Testing of Hybrid Distributed IoT System","authors":"Cyrine Zid, D. Humeniuk, Foutse Khomh, G. Antoniol","doi":"10.1145/3387940.3392218","DOIUrl":"https://doi.org/10.1145/3387940.3392218","url":null,"abstract":"Testing heterogeneous IoT applications such as a home automation systems integrating a variety of devices poses serious challenges. Oftentimes requirements are vaguely defined. Consumer grade cyber-physical devices and software may not meet the reliability and quality standard needed. Plus, system behavior may partially depend on various environmental conditions. For example, WI-FI congestion may cause packet delay; meanwhile cold weather may cause an unexpected drop of inside temperature. We surmise that generating and executing failure exposing scenarios is especially challenging. Modeling phenomenons such as network traffic or weather conditions is complex. One possible solution is to rely on machine learning models approximating the reality. These models, integrated in a system model, can be used to define surrogate models and fitness functions to steer the search in the direction of failure inducing scenarios. However, these models also should be validated. Therefore, there should be a double loop co-evolution between machine learned surrogate models functions and fitness functions. Overall, we argue that in such complex cyber-physical systems, co-evolution and multi-hybrid approaches are needed.","PeriodicalId":309659,"journal":{"name":"Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121914500","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}
M. D. Stefano, Fabiano Pecorelli, D. Tamburri, Fabio Palomba, A. D. Lucia
Software engineering projects are now more than ever a community effort. In the recent past, researchers have shown that their success may not only depend on source code quality, but also on other aspects like the balance of distance, culture, global engineering practices, and more. In such a scenario, understanding the characteristics of the community around a project and foresee possible problems may be the key to develop successful systems. In this paper, we focus on this research problem and propose an exploratory study on the relation between community patterns, i.e., recurrent mixes of organizational or social structure types, and smells, i.e., sub-optimal patterns across the organizational structure of a software development community that may be precursors of some sort of social debt. We exploit association rule mining to discover frequent relations between them. Our findings show that different organizational patterns are connected to different forms of socio-technical problems, possibly suggesting that practitioners should put in place specific preventive actions aimed at avoiding the emergence of community smells depending on the organization of the project.
{"title":"Splicing Community Patterns and Smells: A Preliminary Study","authors":"M. D. Stefano, Fabiano Pecorelli, D. Tamburri, Fabio Palomba, A. D. Lucia","doi":"10.1145/3387940.3392204","DOIUrl":"https://doi.org/10.1145/3387940.3392204","url":null,"abstract":"Software engineering projects are now more than ever a community effort. In the recent past, researchers have shown that their success may not only depend on source code quality, but also on other aspects like the balance of distance, culture, global engineering practices, and more. In such a scenario, understanding the characteristics of the community around a project and foresee possible problems may be the key to develop successful systems. In this paper, we focus on this research problem and propose an exploratory study on the relation between community patterns, i.e., recurrent mixes of organizational or social structure types, and smells, i.e., sub-optimal patterns across the organizational structure of a software development community that may be precursors of some sort of social debt. We exploit association rule mining to discover frequent relations between them. Our findings show that different organizational patterns are connected to different forms of socio-technical problems, possibly suggesting that practitioners should put in place specific preventive actions aimed at avoiding the emergence of community smells depending on the organization of the project.","PeriodicalId":309659,"journal":{"name":"Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122109093","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}
Genetic improvement (GI) uses automated search to find improved versions of existing software. If originally GI directly evolved populations of software, most GI work nowadays use a solution representation based on a list of mutations. This representation however has some limitations, notably in how genetic material can be re-combined. We introduce a novel stack-based representation and discuss its possible benefits.
{"title":"Stack-Based Genetic Improvement","authors":"Aymeric Blot, J. Petke","doi":"10.1145/3387940.3392174","DOIUrl":"https://doi.org/10.1145/3387940.3392174","url":null,"abstract":"Genetic improvement (GI) uses automated search to find improved versions of existing software. If originally GI directly evolved populations of software, most GI work nowadays use a solution representation based on a list of mutations. This representation however has some limitations, notably in how genetic material can be re-combined. We introduce a novel stack-based representation and discuss its possible benefits.","PeriodicalId":309659,"journal":{"name":"Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122806462","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}
Jürgen Musil, Angelika Musil, Danny Weyns, S. Biffl
The technology-driven transformation process continues to spawn novel, growth-oriented digital application domains and platforms. The user base of these society-level software systems consists of a larger proportion of the community and that involve a large set of stakeholder groups. In case of an incident there is a public demand from a variety of stakeholders for multilateral intervention in order to correct the behavior of the software system. For software engineering as a technical discipline that has been fostered and matured in corporate and organizational context, this is a major challenge because it has to deal with a multitude of multidisciplinary stakeholders and their concerns. In order to stimulate further discussions, we discuss software governance on societal level and identify future research challenges of this increasingly relevant topic.
{"title":"Society-Level Software Governance: A Challenging Scenario","authors":"Jürgen Musil, Angelika Musil, Danny Weyns, S. Biffl","doi":"10.1145/3387940.3392269","DOIUrl":"https://doi.org/10.1145/3387940.3392269","url":null,"abstract":"The technology-driven transformation process continues to spawn novel, growth-oriented digital application domains and platforms. The user base of these society-level software systems consists of a larger proportion of the community and that involve a large set of stakeholder groups. In case of an incident there is a public demand from a variety of stakeholders for multilateral intervention in order to correct the behavior of the software system. For software engineering as a technical discipline that has been fostered and matured in corporate and organizational context, this is a major challenge because it has to deal with a multitude of multidisciplinary stakeholders and their concerns. In order to stimulate further discussions, we discuss software governance on societal level and identify future research challenges of this increasingly relevant topic.","PeriodicalId":309659,"journal":{"name":"Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125158003","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}
Microservice architecture (MSA) has become a popular architectural style. The main advantages of MSA include modularization and scalability. However, the development and maintenance of Microservice-based systems are more complex than traditional monolithic architecture. This research plans to develop a novel Chatbot system, referred to as MSABot (Microservice Architecture Bot), to assist in the development and operation of Microservice-based systems by using Chatbots. MSABot integrates a variety of tools to allow users to understand the current status of Microservice development and operation, and to push the information of system errors or risks to users. For the operators who take over the maintenance of Microservices, MSABot also allows them to quickly understand the overall service architecture and the operation status of each service. Besides, we invited multiple users who are familiar with the technology of Microservice or ChapOps to evaluate MSABot. The results of the survey show that more than 90% of the respondents believe that MSABot can adequately support the development and maintenance of Microservice-based systems.
{"title":"MSABot","authors":"Chun-Ting Lin, Shang-Pin Ma, Yu-Wen Huang","doi":"10.1145/3387940.3391501","DOIUrl":"https://doi.org/10.1145/3387940.3391501","url":null,"abstract":"Microservice architecture (MSA) has become a popular architectural style. The main advantages of MSA include modularization and scalability. However, the development and maintenance of Microservice-based systems are more complex than traditional monolithic architecture. This research plans to develop a novel Chatbot system, referred to as MSABot (Microservice Architecture Bot), to assist in the development and operation of Microservice-based systems by using Chatbots. MSABot integrates a variety of tools to allow users to understand the current status of Microservice development and operation, and to push the information of system errors or risks to users. For the operators who take over the maintenance of Microservices, MSABot also allows them to quickly understand the overall service architecture and the operation status of each service. Besides, we invited multiple users who are familiar with the technology of Microservice or ChapOps to evaluate MSABot. The results of the survey show that more than 90% of the respondents believe that MSABot can adequately support the development and maintenance of Microservice-based systems.","PeriodicalId":309659,"journal":{"name":"Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125556997","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}
A novel approach is proposed for constructing models of anomaly detectors using supervised learning from the traces of normal and abnormal operations of an Industrial Control System (ICS). Such detectors are of value in detecting process anomalies in complex critical infrastructure such as power generation and water treatment systems. The traces are obtained by systematically "fuzzing", i.e., manipulating the sensor readings and actuator actions in accordance with the boundaries/partitions that define the system's state. The proposed approach is tested in a Secure Water Treatment (SWaT) testbed -- a replica of a real-world water purification plant, located at the Singapore University of Technology and Design. Multiple supervised classifiers are trained using the traces obtained from SWaT. The efficacy of the proposed approach is demonstrated through empirical evaluation of the supervised classifiers under various performance metrics. Lastly, it is shown that the supervised approach results in significantly lower false positive rates as compared to the unsupervised ones.
{"title":"Domain-Based Fuzzing for Supervised Learning of Anomaly Detection in Cyber-Physical Systems","authors":"Herman Wijaya, M. Aniche, A. Mathur","doi":"10.1145/3387940.3391486","DOIUrl":"https://doi.org/10.1145/3387940.3391486","url":null,"abstract":"A novel approach is proposed for constructing models of anomaly detectors using supervised learning from the traces of normal and abnormal operations of an Industrial Control System (ICS). Such detectors are of value in detecting process anomalies in complex critical infrastructure such as power generation and water treatment systems. The traces are obtained by systematically \"fuzzing\", i.e., manipulating the sensor readings and actuator actions in accordance with the boundaries/partitions that define the system's state. The proposed approach is tested in a Secure Water Treatment (SWaT) testbed -- a replica of a real-world water purification plant, located at the Singapore University of Technology and Design. Multiple supervised classifiers are trained using the traces obtained from SWaT. The efficacy of the proposed approach is demonstrated through empirical evaluation of the supervised classifiers under various performance metrics. Lastly, it is shown that the supervised approach results in significantly lower false positive rates as compared to the unsupervised ones.","PeriodicalId":309659,"journal":{"name":"Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-06-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125826528","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：