Cognitive Systems Research最新文献

英文中文

Complete tableau calculi for Regular MaxSAT and Regular MinSAT 完整的表演算为常规MaxSAT和常规MinSAT

IF 2.1 3区心理学 Q3 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Cognitive Systems Research

Pub Date : 2025-01-07 DOI: 10.1016/j.cogsys.2024.101319

Jordi Coll , Chu-Min Li , Felip Manyà , Elifnaz Yangin

The use of constraint models in symbolic AI has significantly increased during the last decades for their capability of certifying the existence of solutions as well as their optimality. In the latter case, approaches based on the Maximum and Minimum Satisfiability problems, or MaxSAT and MinSAT, have shown to provide state-of-the-art performances in solving many computationally challenging problems of social interest, including scheduling, timetabling and resource allocation. Indeed, the research on new approaches to MaxSAT and MinSAT is a trend still providing cutting-edge advances. In this work, we push in this direction by contributing new tableaux-based calculi for solving the MaxSAT and MinSAT problems of regular propositional logic, referred to as Regular MaxSAT and Regular MinSAT problems, respectively. For these problems, we consider as well the two extensions of the highest practical interest, namely the inclusion of weights to clauses, and the distinction between hard (mandatory) and soft (desirable) constraints. Hence, our methods handle any subclass of the most general variants: Weighted Partial Regular MaxSAT and Weighted Partial Regular MinSAT. We provide a detailed description of the methods and prove that the proposed calculi are sound and complete.

在过去的几十年里，符号人工智能中约束模型的使用显著增加，因为它们能够证明解决方案的存在性以及它们的最优性。在后一种情况下，基于最大和最小可满足性问题（MaxSAT和MinSAT）的方法在解决许多具有计算挑战性的社会问题（包括调度、时间表和资源分配）方面表现出了最先进的性能。事实上，对MaxSAT和MinSAT的新方法的研究仍然是一个前沿的发展趋势。在这项工作中，我们通过提供新的基于表的演算来解决规则命题逻辑的MaxSAT和MinSAT问题，分别称为规则MaxSAT和规则MinSAT问题，从而推动了这一方向。对于这些问题，我们也考虑了最高实际利益的两个扩展，即对条款的权重的包含，以及硬（强制）和软（理想）约束之间的区别。因此，我们的方法可以处理最一般变体的任何子类：加权偏正则MaxSAT和加权偏正则MinSAT。我们提供了详细的方法描述，并证明了所提出的演算是健全和完整的。

{"title":"Complete tableau calculi for Regular MaxSAT and Regular MinSAT","authors":"Jordi Coll , Chu-Min Li , Felip Manyà , Elifnaz Yangin","doi":"10.1016/j.cogsys.2024.101319","DOIUrl":"10.1016/j.cogsys.2024.101319","url":null,"abstract":"<div><div>The use of constraint models in symbolic AI has significantly increased during the last decades for their capability of certifying the existence of solutions as well as their optimality. In the latter case, approaches based on the Maximum and Minimum Satisfiability problems, or MaxSAT and MinSAT, have shown to provide state-of-the-art performances in solving many computationally challenging problems of social interest, including scheduling, timetabling and resource allocation. Indeed, the research on new approaches to MaxSAT and MinSAT is a trend still providing cutting-edge advances. In this work, we push in this direction by contributing new tableaux-based calculi for solving the MaxSAT and MinSAT problems of regular propositional logic, referred to as Regular MaxSAT and Regular MinSAT problems, respectively. For these problems, we consider as well the two extensions of the highest practical interest, namely the inclusion of weights to clauses, and the distinction between hard (mandatory) and soft (desirable) constraints. Hence, our methods handle any subclass of the most general variants: Weighted Partial Regular MaxSAT and Weighted Partial Regular MinSAT. We provide a detailed description of the methods and prove that the proposed calculi are sound and complete.</div></div>","PeriodicalId":55242,"journal":{"name":"Cognitive Systems Research","volume":"90 ","pages":"Article 101319"},"PeriodicalIF":2.1,"publicationDate":"2025-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143096915","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Computational model for affective processing based on Cognitive Sciences: An approach using deterministic finite automata’s and temporal heterogeneity 基于认知科学的情感处理计算模型：一种使用确定性有限自动机和时间异质性的方法

IF 2.1 3区心理学 Q3 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Cognitive Systems Research

Pub Date : 2025-01-07 DOI: 10.1016/j.cogsys.2025.101322

Carlos Zárate, Félix Ramos, Alan Christian López Fraga

Cognitive architectures represent an alternative in the quest to develop general purpose artificial intelligence, for which cognitive sciences are studied. In this work we will focus on modeling affective processing, an important component to enable basic emotional capabilities. This component was developed with the aim of generating affective responses in the presence of stimuli, necessary to feed a basic emotion model already proposed within our research group. For the proposal we used a layer-based model with Deterministic Finite Automata’s (DFA) to process stimuli along the time, which works as structures to store and represent stimulus–response associations. This approach provides an independent component, contrary to the proposals commonly seen in the state of the art, where this process is often embedded in the feelings and emotions calculations. This model was tested to respond to the sounds consonance, showing that is capable to provide and reinforce responses for specific stimuli features. The results obtained show that the model is capable of making associations between the encoded stimuli and the expected responses, taking advantage of the fact that it is not necessary to be trained to identify stimulus patterns but only to learn to respond to them.

认知架构代表了开发通用人工智能的另一种选择，认知科学正是为此而研究的。在这项工作中，我们将重点关注情感处理的建模，这是实现基本情感能力的重要组成部分。该组件的开发目的是在刺激存在的情况下产生情感反应，这对于我们研究小组已经提出的基本情感模型是必要的。对于这个提议，我们使用了一个基于层的模型和确定性有限自动机（DFA）来处理刺激，该模型作为存储和表示刺激-反应关联的结构。这种方法提供了一个独立的组件，这与目前技术中常见的建议相反，在这些建议中，这个过程通常嵌入在感觉和情绪计算中。实验结果表明，该模型能够提供和加强对特定刺激特征的反应。结果表明，该模型能够在编码刺激和预期反应之间建立联系，利用了不需要训练就能识别刺激模式，只需要学习对刺激模式做出反应的事实。

{"title":"Computational model for affective processing based on Cognitive Sciences: An approach using deterministic finite automata’s and temporal heterogeneity","authors":"Carlos Zárate, Félix Ramos, Alan Christian López Fraga","doi":"10.1016/j.cogsys.2025.101322","DOIUrl":"10.1016/j.cogsys.2025.101322","url":null,"abstract":"<div><div>Cognitive architectures represent an alternative in the quest to develop general purpose artificial intelligence, for which cognitive sciences are studied. In this work we will focus on modeling affective processing, an important component to enable basic emotional capabilities. This component was developed with the aim of generating affective responses in the presence of stimuli, necessary to feed a basic emotion model already proposed within our research group. For the proposal we used a layer-based model with Deterministic Finite Automata’s (DFA) to process stimuli along the time, which works as structures to store and represent stimulus–response associations. This approach provides an independent component, contrary to the proposals commonly seen in the state of the art, where this process is often embedded in the feelings and emotions calculations. This model was tested to respond to the sounds consonance, showing that is capable to provide and reinforce responses for specific stimuli features. The results obtained show that the model is capable of making associations between the encoded stimuli and the expected responses, taking advantage of the fact that it is not necessary to be trained to identify stimulus patterns but only to learn to respond to them.</div></div>","PeriodicalId":55242,"journal":{"name":"Cognitive Systems Research","volume":"90 ","pages":"Article 101322"},"PeriodicalIF":2.1,"publicationDate":"2025-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143096914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Constructivist procedural learning for grounded cognitive agents 基于认知主体的建构主义程序学习

IF 2.1 3区心理学 Q3 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE

Cognitive Systems Research

Pub Date : 2025-01-05 DOI: 10.1016/j.cogsys.2025.101321

Sean Kugele

Constructivism is a learning theory based on the idea that individuals actively build their understanding of the world through their interactions with their environment. Learning is a dynamic process where new knowledge builds on prior knowledge, and a learner’s mental models are continually refined by their experiences. Building on this theoretical framework and Drescher’s seminal contributions to constructivist AI, this paper explores constructivism within the context of LIDA (Learning Intelligent Decision Agent), a biologically inspired cognitive architecture. Specifically, I develop a modified version of Drescher’s schema mechanism, which I use to implement LIDA’s Procedural Memory and Action Selection modules. I demonstrate that an agent based on this implementation can construct an accurate internal model of its environmental interactions and use that model to select goal-directed behaviors. This work significantly advances LIDA’s computational capabilities by implementing grounded instructionist procedural learning, hierarchical action plans, and the selection of exploratory behaviors. These computational enhancements will enable the creation of more sophisticated LIDA-based agents that can operate in more complex environments where the hand-coding of procedural knowledge is infeasible. An alternate way to view this work is as an enhancement to Drescher’s schema mechanism, which is a purely symbolic and ungrounded cognitive system. LIDA’s sensory and perceptual systems provide a means by which the schema mechanism’s representations can be grounded. This, in itself, is an important contribution of this paper.

建构主义是一种学习理论，其基础是个体通过与环境的互动积极地建立对世界的理解。学习是一个动态的过程，新知识建立在已有知识的基础上，学习者的思维模式会随着他们的经历而不断完善。基于这一理论框架和Drescher对建构主义人工智能的开创性贡献，本文探讨了LIDA（学习智能决策代理）背景下的建构主义，LIDA是一种受生物学启发的认知架构。具体来说，我开发了一个修改版本的Drescher的模式机制，我用它来实现LIDA的程序记忆和动作选择模块。我证明了基于这种实现的智能体可以构建其环境相互作用的精确内部模型，并使用该模型选择目标导向的行为。这项工作通过实施基础教学程序学习、分层行动计划和探索行为的选择，显著提高了LIDA的计算能力。这些计算能力的增强将能够创建更复杂的基于lida的代理，这些代理可以在更复杂的环境中运行，在这些环境中，程序知识的手工编码是不可行的。另一种看待这项工作的方式是作为对Drescher的图式机制的增强，这是一个纯粹的象征性和无根据的认知系统。LIDA的感觉和知觉系统提供了一种方法，通过这种方法，图式机制的表征可以建立起来。这本身就是本文的一个重要贡献。

{"title":"Constructivist procedural learning for grounded cognitive agents","authors":"Sean Kugele","doi":"10.1016/j.cogsys.2025.101321","DOIUrl":"10.1016/j.cogsys.2025.101321","url":null,"abstract":"<div><div>Constructivism is a learning theory based on the idea that individuals actively build their understanding of the world through their interactions with their environment. Learning is a dynamic process where new knowledge builds on prior knowledge, and a learner’s mental models are continually refined by their experiences. Building on this theoretical framework and Drescher’s seminal contributions to constructivist AI, this paper explores constructivism within the context of LIDA (Learning Intelligent Decision Agent), a biologically inspired cognitive architecture. Specifically, I develop a modified version of Drescher’s schema mechanism, which I use to implement LIDA’s Procedural Memory and Action Selection modules. I demonstrate that an agent based on this implementation can construct an accurate internal model of its environmental interactions and use that model to select goal-directed behaviors. This work significantly advances LIDA’s computational capabilities by implementing grounded instructionist procedural learning, hierarchical action plans, and the selection of exploratory behaviors. These computational enhancements will enable the creation of more sophisticated LIDA-based agents that can operate in more complex environments where the hand-coding of procedural knowledge is infeasible. An alternate way to view this work is as an enhancement to Drescher’s schema mechanism, which is a purely symbolic and ungrounded cognitive system. LIDA’s sensory and perceptual systems provide a means by which the schema mechanism’s representations can be grounded. This, in itself, is an important contribution of this paper.</div></div>","PeriodicalId":55242,"journal":{"name":"Cognitive Systems Research","volume":"90 ","pages":"Article 101321"},"PeriodicalIF":2.1,"publicationDate":"2025-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143135647","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0