Cognitive Agents Learning by Communicating
نویسندگان
چکیده
Cognitive Agent communication is a research field in full development. We propose here an extension and an implementation of the STROBE model, which regards the Agents as Scheme interpreters. These Agents are able to interpret messages in a dedicated environment including an interpreter that learns from the current conversation. These interpreters evolve dynamically, progressively with the conversations, and thus represent evolving meta level Agent knowledge. We illustrate this theoretical model by a “teacher-student” dialogue experimentation, where an Agent learns a new performative at the completion of the conversation. Details of the implementation are not provided here, but are available. RÉSUMÉ. La communication entre Agents cognitifs est un domaine de recherche en pleine effervescence. Nous proposons ici un modèle, basé sur le modèle STROBE, qui considère les Agents comme des interpréteurs Scheme. Ces Agents sont capables d'interpréter des messages dans un environnement donné incluant un interpréteur qui apprend de la conversation. Ces interpréteurs peuvent en outre évoluer dynamiquement au fur et à mesure des conversations et ils représentent la connaissance de ces Agents au niveau méta. Nous illustrons ce modèle théorique par une expérimentation de dialogue de type « professeur-élève », où un Agent apprend un nouveau performatif à l'issue de la conversation. Les détails de l'implémentation ne sont pas fournis ici, mais sont disponibles.
منابع مشابه
Hierarchical Functional Concepts for Knowledge Transfer among Reinforcement Learning Agents
This article introduces the notions of functional space and concept as a way of knowledge representation and abstraction for Reinforcement Learning agents. These definitions are used as a tool of knowledge transfer among agents. The agents are assumed to be heterogeneous; they have different state spaces but share a same dynamic, reward and action space. In other words, the agents are assumed t...
متن کاملDeep Communicating Agents for Abstractive Summarization
We present deep communicating agents in an encoder-decoder architecture to address the challenges of representing a long document for abstractive summarization. With deep communicating agents, the task of encoding a long text is divided across multiple collaborating agents, each in charge of a subsection of the input text. These encoders are connected to a single decoder, trained end-to-end usi...
متن کاملSimulating Language Learning in Community of Agents Using Self-Organizing Maps
In this article, we present a model of a cognitive system, or an agent, with the following properties: it can perceive its environment, it can move in its environment, it can perform some simple actions, and it can send and receive messages. The main components of its internal structure include a working memory, a semantic memory, and a decision making mechanism. In our implemented simulation m...
متن کاملApprentissage issu de la communication pour des agents cognitifs
Cognitive Agent communication is a field of research in full development. We propose here a model, based on the STROBE model, which regards the Agents as Scheme interpreters. These Agents are able to interpret messages of a conversation in both a given environment and an interpreter dedicated to the current conversation. These interpreters can moreover evolve dynamically progressively with the ...
متن کاملThe Emergence of Grammar in Communicating Autonomous Robotic Agents
Over the past five years, the topic of the origins of language is gaining prominence as one of the big unresolved questions of cognitive science. Artificial Intelligence can make a major contribution to this problem by working out precise, testable models using grounded robotic agents which interact with a real world environment and communicate among themselves or with humans about this environ...
متن کاملBio-linguistic transition and Baldwin effect in an evolutionary naming-game model
We examine an evolutionary naming-game model where communicating agents are equipped with an evolutionarily selected learning ability. Such a coupling of biological and linguistic ingredients results in an abrupt transition: upon a small change of a model control parameter a poorly communicating group of linguistically unskilled agents transforms into almost perfectly communicating group with l...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2003