The concept of an agent has become important in both Artificial Intelligence (AI) and mainstream computer science. Our aim in this paper is to point the reader at what we perceive to be the most important theoretical and practical issues associated with the design and construction of intelligent agents. For convenience, we divide these issues into three areas (though as the reader will see, the divisions are at times somewhat arbitrary). Agent theory is concerned with the question of what an agent is, and the use of mathematical formalisms for representing and reasoning about the properties of agents. Agent architectures can be thought of as software engineering models of agents;researchers in this area are primarily concerned with the problem of designing software or hardware systems that will satisfy the properties specified by agent theorists. Finally, agent languages are software systems for programming and experimenting with agents; these languages may embody principles proposed by theorists. The paper is not intended to serve as a tutorial introduction to all the issues mentioned; we hope instead simply to identify the most important issues, and point to work that elaborates on them. The article includes a short review of current and potential applications of agent technology.

/pdf/intelligent-agents-theory-and-practice-5dchj7p8pc.pdf

Intelligent Agents: Theory and Practice

Model-based Bayesian Reinforcement Learning (BRL) provides a principled solution to dealing with the exploration-exploitation trade-off, but such methods typically assume a fully observable environments. The few Bayesian RL methods that are applicable in partially observable domains, such as the Bayes-Adaptive POMDP (BA-POMDP), scale poorly. To address this issue, we introduce the Factored BA-POMDP model (FBA-POMDP), a framework that is able to learn a compact model of the dynamics by exploiting the underlying structure of a POMDP. The FBA-POMDP framework casts the problem as a planning task, for which we adapt the Monte-Carlo Tree Search planning algorithm and develop a belief tracking method to approximate the joint posterior over the state and model variables. Our empirical results show that this method outperforms a number of BRL baselines and is able to learn efficiently when the factorization is known, as well as learn both the factorization and the model parameters simultaneously.

/pdf/bayesian-reinforcement-learning-in-factored-pomdps-xpra8a8w97.pdf

Bayesian Reinforcement Learning in Factored POMDPs

In this provocative book, Barwise and Perry tackle the slippery subject of \"meaning, \" a subject that has long vexed linguists, language philosophers, and logicians.

/pdf/situations-and-attitudes-57f065zg2d.pdf

Situations and Attitudes.

The concept of an agent has become important in both Artificial Intelligence (AI) and mainstream computer science. In this article, we present a survey of what we perceive to be the most important theoretical and practical issues associated with the design and construction of intelligent agents. The article also includes a short review of current and potential applications of agent technology, and closes with a glossary of key terms, an annotated list of systems, and a detailed bibliography. Pointers to further reading are provided throughout.

/pdf/agent-theories-architectures-and-languages-a-survey-4m08hmierd.pdf

Agent theories, architectures, and languages: a survey

Methods and systems consistent with the present invention provide a data processing system having a business object model reflecting the data used during a business transaction. Consistent interfaces are generated from the business object model. These interfaces are suitable for use across industries, across businesses, and across different departments within a business during a business transaction.

Consistent set of interfaces derived from a business object model

Agents operate and exist in some environment, which typically is both computational and physical. The environment might be open or closed, and it might or might not contain other agents. Although there are situations where an agent can operate usefully by itself, the increasing interconnection and networking of computers is making such situations rare, and in the usual state of aaairs the agent interacts with other agents. Whereas the previous chapter deened the structure and characteristics of an individual agent, the focus of this chapter is on systems with multiple agents. At times, the number of agents may be too numerous to deal with them individually, and it is then more convenient to deal with them collectively, as a society of agents. In this chapter, we will learn how to analyze, describe, and design environments in which agents can operate eeectively and interact with each other productively. The environments will provide a computational infrastructure for such interactions to take place. The infrastructure will include protocols for agents to communicate and protocols for agents to interact. Communication protocols enable agents to exchange and understand messages. Interaction protocols enable agents to have conversations, which for our purposes are structured exchanges of messages. As a concrete example of these, a communication protocol might specify that the following types of messages can be exchanged between two agents: Propose a course of action Accept a course of action Reject a course of action Retract a course of action Disagree with a proposed course of action Counterpropose a course of action Based on these message types, the following conversation|an instance of an

Multiagent systems and societies of agents

The architecture of a special-purpose multiprocessor, hierarchically structured and functionally distributed, having ditributed cache memory for local processing and a common applictions task manager in each microcomputer. A group of identical microcomputers execute the total program in an intrinsically parallel mode within the frame times scheduled by a system state control microcomputer.

Multiple-microcomputer processing

This paper explores a linguistic approach to coordination modeling as a formal basis for supply-chain management (SCM) in manufacturing. The approach promotes the interchange of standard documents: enterprises need only describe their supply processes using OAG business object documents and UML interaction diagrams. Our methodology and tools analyze the documents and interactions in terms of four linguistic primitives and convert the diagrams into specifications and implementations of software agents. The agents then cooperate in automating the resultant supply chain. We evaluate our methodology in the context of several industrial scenarios. We conclude that supply-chain automation using software-agent technology is feasible.

/pdf/automating-supply-chain-management-3gahwpysx4.pdf

Automating supply-chain management

Chapter 9 – DAI for Document Retrieval: The MINDS Project

As you build a Web site, it is worthwhile asking, "Should I put my information where it belongs or where people are most likely to look for it?" Our recent research into improving searching through ontologies is providing some interesting results to answer this question. The techniques developed by our research bring organization to the information received and reconcile the semantics of each document. Our goal is to help users retrieve dynamically generated information that is tailored to their individual needs and preferences. We believe that it is easier for individuals or small groups to develop their own ontologies, regardless of whether global ones are available, and that these can be automatically and ex-post-facto related. We are working to determine the efficacy of local annotation for Web sources, as well as performing reconciliation that is qualified by measures of semantic distance. If successful, this research will enable software agents to resolve the semantic misconceptions that inhibit successful interoperation with other agents and that limit the effectiveness of searching distributed information sources.

/pdf/consensus-ontologies-reconciling-the-semantics-of-web-pages-4ut8w2flgm.pdf

Larry M. Stephens

Papers

Multiagent systems and societies of agents

Multiple-microcomputer processing

Automating supply-chain management

Chapter 9 – DAI for Document Retrieval: The MINDS Project

Consensus ontologies. Reconciling the semantics of Web pages and agents