Autonomous Agents and Multi-Agent Systems最新文献

{"title":"A normative approach for resilient multiagent systems","authors":"Geeta Mahala,&nbsp;Özgür Kafalı,&nbsp;Hoa Dam,&nbsp;Aditya Ghose,&nbsp;Munindar P. Singh","doi":"10.1007/s10458-023-09627-4","DOIUrl":"10.1007/s10458-023-09627-4","url":null,"abstract":"<div><p>We model a multiagent system (MAS) in socio-technical terms, combining a social layer consisting of norms with a technical layer consisting of actions that the agents execute. This approach emphasizes autonomy, and makes assumptions about both the social and technical layers explicit. Autonomy means that agents may violate norms. In our approach, agents are computational entities, with each representing a different stakeholder. We express stakeholder requirements of the form that a MAS is resilient in that it can recover (sufficiently) from a failure within a (sufficiently short) duration. We present <span>ReNo</span>, a framework that computes probabilistic and temporal guarantees on whether the underlying requirements are met or, if failed, recovered. <span>ReNo</span> supports the refinement of the specification of a socio-technical system through methodological guidelines to meet the stated requirements. An important contribution of <span>ReNo</span> is that it shows how the social and technical layers can be modeled jointly to enable the construction of resilient systems of autonomous agents. We demonstrate <span>ReNo</span> using a manufacturing scenario with competing public, industrial, and environmental requirements.</p></div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"37 2","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134796590","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ASN: action semantics network for multiagent reinforcement learning","authors":"Tianpei Yang,&nbsp;Weixun Wang,&nbsp;Jianye Hao,&nbsp;Matthew E. Taylor,&nbsp;Yong Liu,&nbsp;Xiaotian Hao,&nbsp;Yujing Hu,&nbsp;Yingfeng Chen,&nbsp;Changjie Fan,&nbsp;Chunxu Ren,&nbsp;Ye Huang,&nbsp;Jiangcheng Zhu,&nbsp;Yang Gao","doi":"10.1007/s10458-023-09628-3","DOIUrl":"10.1007/s10458-023-09628-3","url":null,"abstract":"<div><p>In multiagent systems (MASs), each agent makes individual decisions but all contribute globally to the system’s evolution. Learning in MASs is difficult since each agent’s selection of actions must take place in the presence of other co-learning agents. Moreover, the environmental stochasticity and uncertainties increase exponentially with the number of agents. Previous works borrow various multiagent coordination mechanisms for use in deep learning architectures to facilitate multiagent coordination. However, none of them explicitly consider that different actions can have different influence on other agents, which we call the action semantics. In this paper, we propose a novel network architecture, named Action Semantics Network (ASN), that explicitly represents such action semantics between agents. ASN characterizes different actions’ influence on other agents using neural networks based on the action semantics between them. ASN can be easily combined with existing deep reinforcement learning (DRL) algorithms to boost their performance. Experimental results on StarCraft II micromanagement and Neural MMO show that ASN significantly improves the performance of state-of-the-art DRL approaches, compared with several other network architectures. We also successfully deploy ASN to a popular online MMORPG game called Justice Online, which indicates a promising future for ASN to be applied in even more complex scenarios.</p></div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"37 2","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134796194","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A family of strategyproof mechanisms for activity scheduling","authors":"Xinping Xu,&nbsp;Jingwen Zhang,&nbsp;Minming Li,&nbsp;Lingjie Duan,&nbsp;Lihua Xie","doi":"10.1007/s10458-023-09624-7","DOIUrl":"10.1007/s10458-023-09624-7","url":null,"abstract":"<div><p>Recent years have seen various designs of strategyproof mechanisms in the facility location game and the obnoxious facility game, by considering the facility’s geo-location as a point in the spatial domain. In this paper, we extend this point to be a continuous interval, and study a novel activity scheduling game to schedule an activity in the normalized time domain [0, 1] based on all agents’ time reports for preferences/conflicts. The activity starts at time point <i>y</i> and lasts for a fixed time period of <i>d</i> with <span>(0le dle 1)</span>. Each agent <span>(iin N = {1, cdots , n})</span> wants his preferred time interval <span>([t_i,t_i+l_i])</span> to be close to or overlap with the activity interval <span>([y,y+d])</span>. Since agents are heterogeneous, we consider each agent <i>i</i> has weight <span>(alpha _i)</span> or <span>(beta _i)</span> when the activity is scheduled after or before his time interval, respectively. Thus each agent <i>i</i>’s cost is his weight (<span>(alpha _i)</span> or <span>(beta _i)</span>) multiplied by the time difference between his time interval <span>([t_i,t_i+l_i])</span> and the activity interval <span>([y,y+d].)</span> The social cost is the summation of all agents’ costs. In this game, agents’ preferred time intervals <span>([t_i,t_i+l_i])</span>’s are private information and they may misreport such information to the social planner. Our objective is to choose the activity starting time <i>y</i> so that the mechanisms are strategyproof (i.e., all agents should be truthful to report <span>(t_i)</span>’s and <span>(l_i)</span>’s) and perform well with respect to minimizing the social cost. We design a mechanism outputting an optimal solution and prove that it is group strategyproof. For the objective of minimizing the maximum cost among agents, we design another strategyproof mechanism with the approximation ratio <span>(1+min {alpha /beta ,beta /alpha })</span> when <span>(alpha _i=alpha , beta _i = beta)</span> for <span>(iin N,)</span> and prove it is the best strategyproof mechanism. In the obnoxious activity scheduling game, each agent prefers his conflicting time interval <span>([t_i,t_i+l_i])</span> to be far away from the activity interval <span>([y,y+d])</span>. We design deterministic and randomized group strategyproof mechanisms, and compare their provable approximation ratios to the lower bounds. Finally, we consider the cost/utility of each agent as a 0-1 indicator function and find group strategyproof mechanisms for minimizing the social cost and maximizing the social utility.</p></div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"37 2","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50500785","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Connected coordinated motion planning with bounded stretch","authors":"Sándor P. Fekete,&nbsp;Phillip Keldenich,&nbsp;Ramin Kosfeld,&nbsp;Christian Rieck,&nbsp;Christian Scheffer","doi":"10.1007/s10458-023-09626-5","DOIUrl":"10.1007/s10458-023-09626-5","url":null,"abstract":"<div><p>We consider the problem of connected coordinated motion planning for a large collective of simple, identical robots: From a given start grid configuration of robots, we need to reach a desired target configuration via a sequence of parallel, collision-free robot motions, such that the set of robots induces a connected grid graph at all integer times. The objective is to minimize the <i>makespan</i> of the motion schedule, i.e., to reach the new configuration in a minimum amount of time. We show that this problem is <span>NP</span>-complete, even for deciding whether a makespan of 2 can be achieved, while it is possible to check in polynomial time whether a makespan of 1 can be achieved. On the algorithmic side, we establish simultaneous constant-factor approximation for two fundamental parameters, by achieving <i>constant stretch</i> for <i>constant scale</i>. Scaled shapes (which arise by increasing all dimensions of a given object by the same multiplicative factor) have been considered in previous seminal work on self-assembly, often with unbounded or logarithmic scale factors; we provide methods for a generalized scale factor, bounded by a constant. Moreover, our algorithm achieves a <i>constant stretch factor</i>: If mapping the start configuration to the target configuration requires a maximum Manhattan distance of <i>d</i>, then the total duration of our overall schedule is <span>(mathcal {O}(d))</span>, which is optimal up to constant factors.</p></div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"37 2","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10458-023-09626-5.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50491884","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Price of anarchy of traffic assignment with exponential cost functions","authors":"Jianglin Qiao,&nbsp;Dave de Jonge,&nbsp;Dongmo Zhang,&nbsp;Simeon Simoff,&nbsp;Carles Sierra,&nbsp;Bo Du","doi":"10.1007/s10458-023-09625-6","DOIUrl":"10.1007/s10458-023-09625-6","url":null,"abstract":"<div><p>The rapid evolution of technology in connected automated and autonomous vehicles offers immense potential for revolutionizing future intelligent traffic control and management. This potential is exemplified by the diverse range of control paradigms, ranging from self-routing to centralized control. However, the selection among these paradigms is beyond technical consideration but a delicate balance between autonomous decision-making and holistic system optimization. A pivotal quantitative parameter in navigating this balance is the concept of the “price of anarchy” (PoA) inherent in autonomous decision frameworks. This paper analyses the price of anarchy for road networks with traffic of CAV. We model a traffic network as a routing game in which vehicles are selfish agents who choose routes to travel autonomously to minimize travel delays caused by road congestion. Unlike existing research in which the latency function of road congestion was based on polynomial functions like the well-known BPR function, we focus on routing games where an exponential function can specify the latency of road traffic. We first calculate a tight upper bound for the price of anarchy for this class of games and then compare this result with the tight upper bound of the PoA for routing games with the BPR latency function. The comparison shows that as long as the traffic volume is lower than the road capacity, the tight upper bound of the PoA of the games with the exponential function is lower than the corresponding value with the BPR function. Finally, numerical results based on real-world traffic data demonstrate that the exponential function can approximate road latency as close as the BPR function with even tighter exponential parameters, which results in a relatively lower upper bound.</p></div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"37 2","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50447098","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Generating and choosing organisations for multi-agent systems","authors":"Cleber J. Amaral,&nbsp;Jomi F. Hübner,&nbsp;Stephen Cranefield","doi":"10.1007/s10458-023-09623-8","DOIUrl":"10.1007/s10458-023-09623-8","url":null,"abstract":"<div><p>The design of organisations is a complex and laborious task. It is the subject of recent studies, which define models to automatically perform this task. However, existing models constrain the space of possible solutions by requiring a priori definitions of organisational roles and usually are not suitable for planning resource use. This paper presents GoOrg, a model that uses as input a set of goals and a set of available agents to generate different arrangements of organisational structures made up of synthesised organisational positions. The most distinguishing characteristics of GoOrg are the use of organisational positions instead of roles and that positions are automatically synthesised rather than required as a priori defined inputs. These characteristics facilitate the parametrisation, the use for resource planning and the chance of finding feasible solutions. This paper also introduces two model extensions, which define processes and constraints that illustrate how GoOrg suits different domains. Among aspects that surround an organisation design, this paper discusses models’ input, agents’ abstractions and resource planning.</p></div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"37 2","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10458-023-09623-8.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50511749","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Effect of asynchronous execution and imperfect communication on max-sum belief propagation","authors":"Roie Zivan,&nbsp;Ben Rachmut,&nbsp;Omer Perry,&nbsp;William Yeoh","doi":"10.1007/s10458-023-09621-w","DOIUrl":"10.1007/s10458-023-09621-w","url":null,"abstract":"<div><p>Max-sum is a version of belief propagation that was adapted for solving distributed constraint optimization problems. It has been studied theoretically and empirically, extended to versions that improve solution quality and converge rapidly, and is applicable to multiple distributed applications. The algorithm was presented both as synchronous and asynchronous algorithms. However, neither the differences in the performance of the two execution versions nor the implications of imperfect communication (i.e., massage delay and message loss) on the two versions have been investigated to the best of our knowledge. We contribute to the body of knowledge on Max-sum by: (1) Establishing the theoretical differences between the two execution versions of the algorithm, focusing on the construction of beliefs; (2) Empirically evaluating the differences between the solutions generated by the two versions of the algorithm, with and without message delay or loss; and (3) Establishing both theoretically and empirically the positive effect of damping on reducing the differences between the two versions. Our results indicate that, in contrast to recent published results indicating that message latency has a drastic (positive) effect on the performance of distributed local search algorithms, the effect of imperfect communication on Damped Max-sum (DMS) is minor. The version of Max-sum that includes both damping and splitting of function nodes converges to high quality solutions very fast, even when a large percentage of the messages sent by agents do not arrive at their destinations. Moreover, the quality of solutions in the different versions of DMS is dependent of the number of messages that were received by the agents, regardless of the amount of time they were delayed or if these messages are only a portion of the total number of messages that was sent by the agents.</p></div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"37 2","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"50482745","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Fairness criteria for allocating indivisible chores: connections and efficiencies","authors":"Ankang Sun,&nbsp;Bo Chen,&nbsp;Xuan Vinh Doan","doi":"10.1007/s10458-023-09618-5","DOIUrl":"10.1007/s10458-023-09618-5","url":null,"abstract":"<div><p>We study several fairness notions in allocating indivisible <i>chores</i> (i.e., items with disutilities) to agents who have additive and submodular cost functions. The fairness criteria we are concerned with are envy-free up to any item, envy-free up to one item, maximin share (MMS), and pairwise maximin share (PMMS), which are proposed as relaxations of envy-freeness in the setting of additive cost functions. For allocations under each fairness criterion, we establish their approximation guarantee for other fairness criteria. Under the additive setting, our results show strong connections between these fairness criteria and, at the same time, reveal intrinsic differences between goods allocation and chores allocation. However, such strong relationships cannot be inherited by the submodular setting, under which PMMS and MMS are no longer relaxations of envy-freeness and, even worse, few non-trivial guarantees exist. We also investigate efficiency loss under these fairness constraints and establish their prices of fairness.</p></div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"37 2","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10458-023-09618-5.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46467925","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multi-scenario approach to continuously learn and understand norm violations","authors":"Thiago Freitas dos Santos,&nbsp;Nardine Osman,&nbsp;Marco Schorlemmer","doi":"10.1007/s10458-023-09619-4","DOIUrl":"10.1007/s10458-023-09619-4","url":null,"abstract":"<div><p>Using norms to guide and coordinate interactions has gained tremendous attention in the multiagent community. However, new challenges arise as the interest moves towards dynamic socio-technical systems, where human and software agents interact, and interactions are required to adapt to changing human needs. For instance, different agents (human or software) might not have the same understanding of what it means to violate a norm (e.g., what characterizes hate speech), or their understanding of a norm might change over time (e.g., what constitutes an acceptable response time). The challenge is to address these issues by learning to detect norm violations from the limited interaction data and to explain the reasons for such violations. To do that, we propose a framework that combines Machine Learning (ML) models and incremental learning techniques. Our proposal is equipped to solve tasks in both tabular and text classification scenarios. Incremental learning is used to continuously update the base ML models as interactions unfold, ensemble learning is used to handle the imbalance class distribution of the interaction stream, Pre-trained Language Model (PLM) is used to learn from text sentences, and Integrated Gradients (IG) is the interpretability algorithm. We evaluate the proposed approach in the use case of Wikipedia article edits, where interactions revolve around editing articles, and the norm in question is prohibiting vandalism. Results show that the proposed framework can learn to detect norm violation in a setting with data imbalance and concept drift.</p></div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"37 2","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10458-023-09619-4.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"41757932","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Hardness of candidate nomination","authors":"Katarína Cechlárová,&nbsp;Julien Lesca,&nbsp;Diana Trellová,&nbsp;Martina Hančová,&nbsp;Jozef Hanč","doi":"10.1007/s10458-023-09622-9","DOIUrl":"10.1007/s10458-023-09622-9","url":null,"abstract":"<div><p>We consider elections where the set of candidates is split into parties and each party can nominate just one candidate. We study the computational complexity of two problems. The <span>Possible President</span> problem asks whether a given party candidate can become the unique winner of the election for some nominations from other parties. The <span>Necessary President</span> is the problem to decide whether a given candidate will be the unique winner of the election for any possible nominations from other parties. We consider several different voting rules and show that for all of them the <span>Possible President</span> problem is NP-complete, even if the size of each party is at most two; for some voting rules we prove that the <span>Necessary President</span> is coNP-complete. Further, we formulate integer programs to solve the <span>Possible President</span> and <span>Necessary President</span> problems and test them on real and artificial data.</p></div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"37 2","pages":""},"PeriodicalIF":1.9,"publicationDate":"2023-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10458-023-09622-9.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48956407","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}