Some researchers evaluate their fair Machine Learning (ML) algorithms by simulating data with a fair and biased version of its labels. The fair labels reflect what labels individuals deserve, while the biased labels reflect labels obtained through a biased decision process. Given such data, fair algorithms are evaluated by measuring how well they can predict the fair labels, after being trained on the biased ones. The big problem with these approaches is, that they are based on simulated data, which is unlikely to capture the full complexity and noise of real-life decision problems. In this paper, we show how we created a new, more realistic dataset with both fair and biased labels. For this purpose, we started with an existing dataset containing information about high school students and whether they passed an exam or not. Through a human experiment, where participants estimated the school performance given some description of these students, we collect a biased version of these labels. We show how this new dataset can be used to evaluate fair ML algorithms, and how some fairness interventions, that perform well in the traditional evaluation schemes, do not necessarily perform well with respect to the unbiased labels in our dataset, leading to new insights into the performance of debiasing techniques.
{"title":"Real-life Performance of Fairness Interventions - Introducing A New Benchmarking Dataset for Fair ML","authors":"Daphne Lenders, T. Calders","doi":"10.1145/3555776.3577634","DOIUrl":"https://doi.org/10.1145/3555776.3577634","url":null,"abstract":"Some researchers evaluate their fair Machine Learning (ML) algorithms by simulating data with a fair and biased version of its labels. The fair labels reflect what labels individuals deserve, while the biased labels reflect labels obtained through a biased decision process. Given such data, fair algorithms are evaluated by measuring how well they can predict the fair labels, after being trained on the biased ones. The big problem with these approaches is, that they are based on simulated data, which is unlikely to capture the full complexity and noise of real-life decision problems. In this paper, we show how we created a new, more realistic dataset with both fair and biased labels. For this purpose, we started with an existing dataset containing information about high school students and whether they passed an exam or not. Through a human experiment, where participants estimated the school performance given some description of these students, we collect a biased version of these labels. We show how this new dataset can be used to evaluate fair ML algorithms, and how some fairness interventions, that perform well in the traditional evaluation schemes, do not necessarily perform well with respect to the unbiased labels in our dataset, leading to new insights into the performance of debiasing techniques.","PeriodicalId":42971,"journal":{"name":"Applied Computing Review","volume":"89 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80350170","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}
Complex Event Processing (CEP) is a mature technology providing particularly efficient solutions for pattern detection in streaming settings. Nevertheless, even the most advanced CEP engines struggle to deal with cases when the number of pattern matches grows exponentially, e.g., when the queries involve Kleene operators to detect trends. In this work, we present an overview of state-of-the-art CEP engines used for pattern detection, focusing also on systems that discover demanding event trends. The main contribution lies in the comparison of existing CEP engine alternatives and the proposal of a novel hash-endowed automata-based lazy hybrid execution engine, called SASEXT, that undertakes the processing of pattern queries involving Kleene patterns. Our proposal is orders of magnitude faster than existing solutions.
{"title":"Exploring alternatives of Complex Event Processing execution engines in demanding cases","authors":"Styliani Kyrama, A. Gounaris","doi":"10.1145/3555776.3577734","DOIUrl":"https://doi.org/10.1145/3555776.3577734","url":null,"abstract":"Complex Event Processing (CEP) is a mature technology providing particularly efficient solutions for pattern detection in streaming settings. Nevertheless, even the most advanced CEP engines struggle to deal with cases when the number of pattern matches grows exponentially, e.g., when the queries involve Kleene operators to detect trends. In this work, we present an overview of state-of-the-art CEP engines used for pattern detection, focusing also on systems that discover demanding event trends. The main contribution lies in the comparison of existing CEP engine alternatives and the proposal of a novel hash-endowed automata-based lazy hybrid execution engine, called SASEXT, that undertakes the processing of pattern queries involving Kleene patterns. Our proposal is orders of magnitude faster than existing solutions.","PeriodicalId":42971,"journal":{"name":"Applied Computing Review","volume":"4 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78875002","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}
L. Bongiovanni, Luca Bruno, Fabrizio Dominici, Giuseppe Rizzo
Classification of documents according to a custom internal hierarchical taxonomy is a common problem for many organizations that deal with textual data. Approaches aimed to address this challenge are, for the vast majority, supervised methods, which have the advantage of producing good results on specific datasets, but the major drawbacks of requiring an entire corpus of annotated documents, and the resulting models are not directly applicable to a different taxonomy. In this paper, we aim to contribute to this important issue, by proposing a method to classify text according to a custom hierarchical taxonomy entirely without the need of labelled data. The idea is to first leverage the semantic information encoded into pre-trained Deep Language Models to assigned a prior relevance score for each label of the taxonomy using zero-shot, and secondly take advantage of the hierarchical structure to reinforce this prior belief. Experiments are conducted on three hierarchically annotated datasets: WebOfScience, DBpedia Extracts and Amazon Product Reviews, which are very diverse in the type of language adopted and have taxonomy depth of two and three levels. We first compare different zero-shot methods, and then we show that our hierarchy-aware approach substantially improves results across every dataset.
{"title":"Zero-Shot Taxonomy Mapping for Document Classification","authors":"L. Bongiovanni, Luca Bruno, Fabrizio Dominici, Giuseppe Rizzo","doi":"10.1145/3555776.3577653","DOIUrl":"https://doi.org/10.1145/3555776.3577653","url":null,"abstract":"Classification of documents according to a custom internal hierarchical taxonomy is a common problem for many organizations that deal with textual data. Approaches aimed to address this challenge are, for the vast majority, supervised methods, which have the advantage of producing good results on specific datasets, but the major drawbacks of requiring an entire corpus of annotated documents, and the resulting models are not directly applicable to a different taxonomy. In this paper, we aim to contribute to this important issue, by proposing a method to classify text according to a custom hierarchical taxonomy entirely without the need of labelled data. The idea is to first leverage the semantic information encoded into pre-trained Deep Language Models to assigned a prior relevance score for each label of the taxonomy using zero-shot, and secondly take advantage of the hierarchical structure to reinforce this prior belief. Experiments are conducted on three hierarchically annotated datasets: WebOfScience, DBpedia Extracts and Amazon Product Reviews, which are very diverse in the type of language adopted and have taxonomy depth of two and three levels. We first compare different zero-shot methods, and then we show that our hierarchy-aware approach substantially improves results across every dataset.","PeriodicalId":42971,"journal":{"name":"Applied Computing Review","volume":"86 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79385181","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}
Kristof Jannes, Vincent Reniers, Wouter Lenaerts, B. Lagaisse, W. Joosen
Distributed Ledger Technology (DLTs) or blockchains have been steadily emerging and providing innovation in the past decade for several use cases, ranging from financial networks, to notarization, or trustworthy execution via smart contracts. DLTs are enticing due to their properties of decentralization, non-repudiation, and auditability (transparency). These properties are of high potential to access control systems that can be implemented on-chain, and are executed without infringement and full transparency. While it remains uncertain which use cases will truly turn out to be viable, many use cases such as financial transactions can benefit from integrating certain restrictions via access control on the blockchain. In addition, smart contracts may in the future present security risks that are currently yet unknown. As a solution, access control policies can provide flexibility in the execution flow when adopted by smart contracts. In this paper, we present our DEDACS architecture which provides decentralized and dynamic access control for smart contracts in a policy-based manner. Our access control is expressive as it features policies, and dynamic as the environment or users can be changed, or alternative policies can be assigned to smart contracts. DEDACS ensures that our access control preserves the desired properties of decentralization and transparency, while aiming to keep the costs involved as minimal as possible. We have evaluated DEDACS in the context of a Uniswap token-exchange platform, in which we evaluated the costs related to (i) the introduced overhead at deployment time and (ii) the operational overhead cost. DEDACS introduces a relative overhead of on average 52% at deployment time, and an operational overhead between 52% and 80% depending on the chosen policy and its complexity.
{"title":"DEDACS: Decentralized and dynamic access control for smart contracts in a policy-based manner","authors":"Kristof Jannes, Vincent Reniers, Wouter Lenaerts, B. Lagaisse, W. Joosen","doi":"10.1145/3555776.3577676","DOIUrl":"https://doi.org/10.1145/3555776.3577676","url":null,"abstract":"Distributed Ledger Technology (DLTs) or blockchains have been steadily emerging and providing innovation in the past decade for several use cases, ranging from financial networks, to notarization, or trustworthy execution via smart contracts. DLTs are enticing due to their properties of decentralization, non-repudiation, and auditability (transparency). These properties are of high potential to access control systems that can be implemented on-chain, and are executed without infringement and full transparency. While it remains uncertain which use cases will truly turn out to be viable, many use cases such as financial transactions can benefit from integrating certain restrictions via access control on the blockchain. In addition, smart contracts may in the future present security risks that are currently yet unknown. As a solution, access control policies can provide flexibility in the execution flow when adopted by smart contracts. In this paper, we present our DEDACS architecture which provides decentralized and dynamic access control for smart contracts in a policy-based manner. Our access control is expressive as it features policies, and dynamic as the environment or users can be changed, or alternative policies can be assigned to smart contracts. DEDACS ensures that our access control preserves the desired properties of decentralization and transparency, while aiming to keep the costs involved as minimal as possible. We have evaluated DEDACS in the context of a Uniswap token-exchange platform, in which we evaluated the costs related to (i) the introduced overhead at deployment time and (ii) the operational overhead cost. DEDACS introduces a relative overhead of on average 52% at deployment time, and an operational overhead between 52% and 80% depending on the chosen policy and its complexity.","PeriodicalId":42971,"journal":{"name":"Applied Computing Review","volume":"31 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75192893","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}
Combining the challenges of streaming data and multi-label learning, the task of mining a drifting, multi-label data stream requires methods that can accurately predict labelsets, adapt to various types of concept drift and run fast enough to process each data point before the next arrives. To achieve greater accuracy, many multi-label algorithms use computationally expensive techniques, such as multiple adaptive windows, with little concern for runtime and memory complexity. We present Aging and Rejuvenating kNN (ARkNN) which uses simple resources and efficient strategies to weight instances based on age, predictive performance, and similarity to the incoming data. We break down ARkNN into its component strategies to show the impact of each and experimentally compare ARkNN to seven state-of-the-art methods for learning from multi-label data streams. We demonstrate that it is possible to achieve competitive performance in multi-label classification on streams without sacrificing runtime and memory use, and without using complex and computationally expensive dual memory strategies.
{"title":"Aging and rejuvenating strategies for fading windows in multi-label classification on data streams","authors":"M. Roseberry, S. Džeroski, A. Bifet, Alberto Cano","doi":"10.1145/3555776.3577625","DOIUrl":"https://doi.org/10.1145/3555776.3577625","url":null,"abstract":"Combining the challenges of streaming data and multi-label learning, the task of mining a drifting, multi-label data stream requires methods that can accurately predict labelsets, adapt to various types of concept drift and run fast enough to process each data point before the next arrives. To achieve greater accuracy, many multi-label algorithms use computationally expensive techniques, such as multiple adaptive windows, with little concern for runtime and memory complexity. We present Aging and Rejuvenating kNN (ARkNN) which uses simple resources and efficient strategies to weight instances based on age, predictive performance, and similarity to the incoming data. We break down ARkNN into its component strategies to show the impact of each and experimentally compare ARkNN to seven state-of-the-art methods for learning from multi-label data streams. We demonstrate that it is possible to achieve competitive performance in multi-label classification on streams without sacrificing runtime and memory use, and without using complex and computationally expensive dual memory strategies.","PeriodicalId":42971,"journal":{"name":"Applied Computing Review","volume":"57 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76923236","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}
Hugo Sousa, Arian Pasquali, Alípio Jorge, Catarina Sousa Santos, M'ario Amorim Lopes
Textual health records of cancer patients are usually protracted and highly unstructured, making it very time-consuming for health professionals to get a complete overview of the patient's therapeutic course. As such limitations can lead to suboptimal and/or inefficient treatment procedures, healthcare providers would greatly benefit from a system that effectively summarizes the information of those records. With the advent of deep neural models, this objective has been partially attained for English clinical texts, however, the research community still lacks an effective solution for languages with limited resources. In this paper, we present the approach we developed to extract procedures, drugs, and diseases from oncology health records written in European Portuguese. This project was conducted in collaboration with the Portuguese Institute for Oncology which, besides holding over 10 years of duly protected medical records, also provided oncologist expertise throughout the development of the project. Since there is no annotated corpus for biomedical entity extraction in Portuguese, we also present the strategy we followed in annotating the corpus for the development of the models. The final models, which combined a neural architecture with entity linking, achieved F1 scores of 88.6, 95.0, and 55.8 per cent in the mention extraction of procedures, drugs, and diseases, respectively.
{"title":"A Biomedical Entity Extraction Pipeline for Oncology Health Records in Portuguese","authors":"Hugo Sousa, Arian Pasquali, Alípio Jorge, Catarina Sousa Santos, M'ario Amorim Lopes","doi":"10.1145/3555776.3578577","DOIUrl":"https://doi.org/10.1145/3555776.3578577","url":null,"abstract":"Textual health records of cancer patients are usually protracted and highly unstructured, making it very time-consuming for health professionals to get a complete overview of the patient's therapeutic course. As such limitations can lead to suboptimal and/or inefficient treatment procedures, healthcare providers would greatly benefit from a system that effectively summarizes the information of those records. With the advent of deep neural models, this objective has been partially attained for English clinical texts, however, the research community still lacks an effective solution for languages with limited resources. In this paper, we present the approach we developed to extract procedures, drugs, and diseases from oncology health records written in European Portuguese. This project was conducted in collaboration with the Portuguese Institute for Oncology which, besides holding over 10 years of duly protected medical records, also provided oncologist expertise throughout the development of the project. Since there is no annotated corpus for biomedical entity extraction in Portuguese, we also present the strategy we followed in annotating the corpus for the development of the models. The final models, which combined a neural architecture with entity linking, achieved F1 scores of 88.6, 95.0, and 55.8 per cent in the mention extraction of procedures, drugs, and diseases, respectively.","PeriodicalId":42971,"journal":{"name":"Applied Computing Review","volume":"52 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75786805","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}
Stefano Bistarelli, A. Bracciali, R. Klomp, Ivan Mercanti
The Bitcoin language SCRIPT has undergone several technically non-trivial updates, still striving from security and minimal risk exposure. Up-to-date, formal verification is of strong interest for script programs that validate the correctness of the Bitcoin decentralised ledger, and allow more and more sophisticated protocols and decentralised applications to be implemented on top of Bitcoin transactions. We propose ScriFy, a comprehensive framework for the verification of the current SCRIPT language: a symbolic semantics and execution model, a model checker, and a modular (dockered), open-source verifier. Given the SCRIPT code that locks a Bitcoin transaction, ScriFy returns the minimal information needed to successfully execute it and authorise the transaction. Distinguishably, ScriFy features both recently added SCRIPT operators and an enhanced analysis, which considers prior information in the ledger. The framework is proved correct and validated through significant examples.
{"title":"Towards automated verification of Bitcoin-based decentralised applications","authors":"Stefano Bistarelli, A. Bracciali, R. Klomp, Ivan Mercanti","doi":"10.1145/3555776.3578996","DOIUrl":"https://doi.org/10.1145/3555776.3578996","url":null,"abstract":"The Bitcoin language SCRIPT has undergone several technically non-trivial updates, still striving from security and minimal risk exposure. Up-to-date, formal verification is of strong interest for script programs that validate the correctness of the Bitcoin decentralised ledger, and allow more and more sophisticated protocols and decentralised applications to be implemented on top of Bitcoin transactions. We propose ScriFy, a comprehensive framework for the verification of the current SCRIPT language: a symbolic semantics and execution model, a model checker, and a modular (dockered), open-source verifier. Given the SCRIPT code that locks a Bitcoin transaction, ScriFy returns the minimal information needed to successfully execute it and authorise the transaction. Distinguishably, ScriFy features both recently added SCRIPT operators and an enhanced analysis, which considers prior information in the ledger. The framework is proved correct and validated through significant examples.","PeriodicalId":42971,"journal":{"name":"Applied Computing Review","volume":"199 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73956331","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}
Low-power and Lossy Networks (LLN) are utilised for numerous Internet of Things (IoT) applications. IEEE has specified the Time-slotted Channel Hopping (TSCH) Media Access Control (MAC) to target the needs of Industrial IoT. TSCH supports deterministic communications over unreliable wireless environments and balances energy, bandwidth and latency. Furthermore, the Minimal 6TiSCH configuration defined Routing Protocol for Low power and Lossy networks (RPL) with the Objective Function 0 (OF0). Inherent factors from RPL operation, such as joining procedure, parent switching, and trickle timer fluctuations, may introduce overhead and overload the network with control messages. The application and RPL control data may lead to an unpredicted networking bottleneck, potentially causing network instability. Hence, a stable RPL operation contributes to a healthy TSCH operation. In this paper, we explore TSCH MAC and RPL metrics to identify factors that lead to performance degradation and specify indicators to anticipate network disorders towards increasing Industrial IoT reliability. A TSCH Schedule Function might employ the identified aspects to foresee disturbances, proactively allocate the proper amount of cells, and avoid networking congestion.
{"title":"Towards the support of Industrial IoT applications with TSCH","authors":"Ivanilson F. Vieira Júnior, M. Curado, J. Granjal","doi":"10.1145/3555776.3577752","DOIUrl":"https://doi.org/10.1145/3555776.3577752","url":null,"abstract":"Low-power and Lossy Networks (LLN) are utilised for numerous Internet of Things (IoT) applications. IEEE has specified the Time-slotted Channel Hopping (TSCH) Media Access Control (MAC) to target the needs of Industrial IoT. TSCH supports deterministic communications over unreliable wireless environments and balances energy, bandwidth and latency. Furthermore, the Minimal 6TiSCH configuration defined Routing Protocol for Low power and Lossy networks (RPL) with the Objective Function 0 (OF0). Inherent factors from RPL operation, such as joining procedure, parent switching, and trickle timer fluctuations, may introduce overhead and overload the network with control messages. The application and RPL control data may lead to an unpredicted networking bottleneck, potentially causing network instability. Hence, a stable RPL operation contributes to a healthy TSCH operation. In this paper, we explore TSCH MAC and RPL metrics to identify factors that lead to performance degradation and specify indicators to anticipate network disorders towards increasing Industrial IoT reliability. A TSCH Schedule Function might employ the identified aspects to foresee disturbances, proactively allocate the proper amount of cells, and avoid networking congestion.","PeriodicalId":42971,"journal":{"name":"Applied Computing Review","volume":"19 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87524582","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}
Ademar França de Sousa Neto, F. Ramos, D. Albuquerque, Emanuel Dantas, M. Perkusich, H. Almeida, A. Perkusich
Agile Software Development (ASD) implicitly manages risks through, for example, its short development cycles (i.e., iterations). The absence of explicit risk management activities in ASD might be problematic since this approach cannot handle all types of risks, might cause risks (e.g., technical debt), and does not promote knowledge reuse throughout an organization. Thus, there is a need to bring discipline to agile risk management. This study focuses on bringing such discipline to organizations that conduct multiple projects to develop software products using ASD, specifically, the Scrum framework, which is the most popular way of adopting ASD. For this purpose, we developed a novel solution that was articulated in partnership with an industry partner. It is a process to complement the Scrum framework to use a recommender system that recommends risks and response plans for a target project, given the risks registered for similar projects in an organization's risk memory (i.e., database). We evaluated the feasibility of the proposed recommender system solution using pre-collected datasets from 17 projects from our industry partner. Since we used the KNN algorithm, we focused on finding the best configuration of k (i.e., the number of neighbors) and the similarity measure. As a result, the configuration with the best results had k = 6 (i.e., six neighbors) and used the Manhattan similarity measure, achieving precision = 45%; recall = 90%; and F1-score = 58%. The results show that the proposed recommender system can assist Scrum Teams in identifying risks and response plans, and it is promising to aid decision-making in Scrum-based projects. Thus, we concluded that our proposed recommender system-based risk management process is promising for helping Scrum Teams address risks more efficiently.
{"title":"Towards a Recommender System-based Process for Managing Risks in Scrum Projects","authors":"Ademar França de Sousa Neto, F. Ramos, D. Albuquerque, Emanuel Dantas, M. Perkusich, H. Almeida, A. Perkusich","doi":"10.1145/3555776.3577748","DOIUrl":"https://doi.org/10.1145/3555776.3577748","url":null,"abstract":"Agile Software Development (ASD) implicitly manages risks through, for example, its short development cycles (i.e., iterations). The absence of explicit risk management activities in ASD might be problematic since this approach cannot handle all types of risks, might cause risks (e.g., technical debt), and does not promote knowledge reuse throughout an organization. Thus, there is a need to bring discipline to agile risk management. This study focuses on bringing such discipline to organizations that conduct multiple projects to develop software products using ASD, specifically, the Scrum framework, which is the most popular way of adopting ASD. For this purpose, we developed a novel solution that was articulated in partnership with an industry partner. It is a process to complement the Scrum framework to use a recommender system that recommends risks and response plans for a target project, given the risks registered for similar projects in an organization's risk memory (i.e., database). We evaluated the feasibility of the proposed recommender system solution using pre-collected datasets from 17 projects from our industry partner. Since we used the KNN algorithm, we focused on finding the best configuration of k (i.e., the number of neighbors) and the similarity measure. As a result, the configuration with the best results had k = 6 (i.e., six neighbors) and used the Manhattan similarity measure, achieving precision = 45%; recall = 90%; and F1-score = 58%. The results show that the proposed recommender system can assist Scrum Teams in identifying risks and response plans, and it is promising to aid decision-making in Scrum-based projects. Thus, we concluded that our proposed recommender system-based risk management process is promising for helping Scrum Teams address risks more efficiently.","PeriodicalId":42971,"journal":{"name":"Applied Computing Review","volume":"30 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73470263","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}
Building conversational agents to help humans in domain-specific tasks is challenging since the agent needs to understand the natural language and act over it while accessing domain expert knowledge. Modern natural language processing techniques led to an expansion of conversational agents, with recent pretrained language models achieving increasingly accurate language recognition results using ever-larger open datasets. However, the black-box nature of such pretrained language models obscures the agent's reasoning and its motivations when responding, leading to unexplained dialogues. We develop a belief-desire-intention (BDI) agent as a task-oriented dialogue system to introduce mental attitudes similar to humans describing their behavior during a dialogue. We compare the resulting model with a pipeline dialogue model by leveraging existing components from dialogue systems and developing the agent's intention selection as a dialogue policy. We show that combining traditional agent modelling approaches, such as BDI, with more recent learning techniques can result in efficient and scrutable dialogue systems.
{"title":"Modeling a Conversational Agent using BDI Framework","authors":"Alexandre Yukio Ichida, Felipe Meneguzzi","doi":"10.1145/3555776.3577657","DOIUrl":"https://doi.org/10.1145/3555776.3577657","url":null,"abstract":"Building conversational agents to help humans in domain-specific tasks is challenging since the agent needs to understand the natural language and act over it while accessing domain expert knowledge. Modern natural language processing techniques led to an expansion of conversational agents, with recent pretrained language models achieving increasingly accurate language recognition results using ever-larger open datasets. However, the black-box nature of such pretrained language models obscures the agent's reasoning and its motivations when responding, leading to unexplained dialogues. We develop a belief-desire-intention (BDI) agent as a task-oriented dialogue system to introduce mental attitudes similar to humans describing their behavior during a dialogue. We compare the resulting model with a pipeline dialogue model by leveraging existing components from dialogue systems and developing the agent's intention selection as a dialogue policy. We show that combining traditional agent modelling approaches, such as BDI, with more recent learning techniques can result in efficient and scrutable dialogue systems.","PeriodicalId":42971,"journal":{"name":"Applied Computing Review","volume":"8 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74685130","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：