Cooperative dialogues with the support of autonomous agents (original) (raw)

Graph-based representations of clarification strategies supporting automatic dialogue management

Italian Journal of Computational Linguistics

This paper aims at presenting a dialogue-oriented approach to the construction of a graph knowledge base (KB) supporting task-oriented human-machine interactions. In particular, we focus on different pragmatic scenarios, facing the Common Ground issue and arguing that knowledge bases (in the form of graphs) are needed to make a clarification and recover pieces of information when inconsistencies occur during the communicative exchange. The main contributions of this work are: 1) a flexible dialog system architecture designed to be plugged into existing service infrastructures, 2) a graph-based knowledge representation protocol to manage both dialog domain and dialog management, 3) a detailed investigation of clarification requests forms with respect to their functions. After a brief introduction (see Section 1), we present: the theoretical underpinnings of the paper and the background work (see Section 2) our system architecture (see Sections 3 and 4) and the clarification requests (CRs) issue (see Section 5); our CRs classification, and some examples in context (see Section 6).

An integrated structured analysis approach to intelligent agent communication

Proceedings IFIP World …, 1998

Intelligent multi-agent systems offer promising approaches for knowledge-intensive distributed applications. Now that such systems are becoming applied on a wider industrial scale, there is a practical need for structured analysis and design methods, similarly as exist for more conventional information and knowledge systems. This is still lacking for intelligent agent software. In this paper, we describe how the process of agent communication specification can be carried out through a structured analysis approach. The structured analysis approach we propose is an integrated extension of the CommonKADS methodology, a widely used standard for knowledge analysis and systems development. Our approach is based on and illustrated by a large-scale multi-agent application for distributed energy load management in industries and households, called HOMEBOTS, which is discussed as an extensive industrial case study. Task Model Task I/O info objects ..... Agent Model Agent capabilities ..... Transaction identifier/name I/O info objects agents involved communication plan constraints info exchange spec Information Exchange Specification Communication Plan Knowledge Model Task structure transfer functions ..... involved-in involved-in involved-in part-of part-of dialogue diagram transaction control communication type message content message control info form/medium 1. The overall communication plan, which governs the full dialogue between the agents.

Dialogues among Distributed Agents

1997

Abstract The paper presents a new language for implementing distributed and autonomous agents. Assuming that any autonomous agent may be modelled by a system engaged in a dialogue with other agents, we consider a previously described, theoretical dialogue architecture-called STROBE [5]-that reduces the computational primitives for modelling synchronous, two-partner dialogues to STReams, OBjects and Environments.

CG-KQML+: an agent communication language and its use in a multi-agent system

2001

This paper deals with communication protocols between agents and between agents and users [3]. It presents a new communication model which is based on a careful analysis of speech act theory and on two fundamental principles applied to communication: a) communication is considered as a negotiation process and, b) communication results in an exchange of mental states. Using this model of communication and the conceptual graph formalism for the representational level, we propose a new agent communication language, called CG-KQML+ which is an extension of the KQML language. The paper also shows the use of CG-KQML+ in a MAS called POSTAGE which aims at helping users in their correspondence task. In POSTAGE, software agents manage administrative correspondence on behalf of and in cooperation with their users. Users and agents have interactions which respect administrative correspondence rules. A POSTAGE agent is responsible for sending the generated message to the addressee's POSTAGE agent. The paper presents the second version of POSTAGE which is implemented using the Prolog+CG language. This paper deals with communication protocols between agents and between agents and users [ 3]. It presents a new communication model which is based on a deep analysis of speech act theory [ 22] [28] and on two fundamental principles: a) communication is considered as a negotiation process [14, 18], b) communication results in an exchange of mental states [7, 24]. Thus, we consider agents' communication as exchanges of mental states (goals, beliefs, etc.) and exchanges of what we call communicational states (CS). Communication is considered as a negotiation game where agents negotiate about proposed CSs. An agent proposes a CS and other agents react to the proposal by accepting, rejecting the proposed CS or even asking for further information. Such an action establishes a relationship between the CS and the agent that is called an agent's positioning. Using this model of communication and the conceptual graph formalism for the representational level, we developed a new agent communication language, called CG-KQML+ which is an extension of the KQML language [12]. CG-KQML+ overcomes some limitations of KQML: KQML performatives are limited to the assertive and directive categories, inappropriate choice of performatives, different interpretations of KQML performatives. The paper also shows the use of CG-KQML+ in a MAS called POSTAGE (POSTman AGEnt) [2]. The aim of this MAS is to help users to achieve correspondence tasks. In POSTAGE, software agents manage administrative correspondence on behalf of and in cooperation with their users. Users and agents interact respecting administrative correspondence rules. A POSTAGE agent is responsible for sending the generated message to the addressee's POSTAGE agent. A first version of POSTAGE has been implemented using ECLIPSE [11] and Delphi [9]. Since that time and by using the Conceptual Graph formalism more fully, we enhanced our standardization work as well as our formulation of POSTAGE. Now, a new version of POSTAGE has been implemented with Prolog+CG language [ 15]. Being a CG-based extension of Prolog, Prolog+CG provides the abstraction level needed to easily implement a CG-based application. Indeed, our new version of POSTAGE is more concise and readable. Moreover, the integration of Java and Prolog+CG [16] enabled us to develop the front/end interface using Java and the kernel of the system using Prolog+CG. Section 2 presents our agent communication model. Section 3 presents CG-KQML+. Section 4 presents the POSTAGE multi-agent system. Section 5 discusses some future works and concludes the paper. 2 The communication model When interacting, agents can engage in two kinds of communication: agent/user communication and inter-agent communication (Figure 1). Agents communicate with users in order to characterize their needs and to provide them with answers or solutions. Agents communicate with each other in order to exchange various kinds of information. When communicating with other agents, an agent uses a specific Agent Communication Language (ACL). An agent's architecture contains a communication process which handles communication activities as well as other processes used to perform various tasks such as planning, decision making or negotiation. In this paper, we focus on the communication activity.

Engineering issues in inter-agent dialogues

This paper presents a logical framework for modeling of complex dialogues between intelligent and autonomous agents. Our overall approach builds on the assumption that an agent is composed of a set of modules, each of them conveying the appropriate knowledge to carry out a certain dialogue type, such as deliberation, negotiation, persuasion, etc. Much attention has been paid in keeping our framework as operational as possible, in that the architecture of agents and their conversational protocol are thoroughly interrelated. Due to the proposed knowledge structure, set of application-independent rules (called dialogue policies) and the combination of backward and forward reasoning, the framework can generate automatic dialogues between agents.

Modelling and visualizing agent conversations

Proceedings of the fifth …, 2001

The notion of conversations, and their associated conversation protocols, as abstract representations to be used to characterise extended agent message exchanges has been discussed in the agent community for several years now. However, there is still no consensus on an appropriate formal specification for these abstractions. This paper describes a proposed formalism based on Petri Nets that is to be used for the modelling of complex, concurrent conversations between agents in a multi-agent system. The approach described in this paper can be used both to define simple conversation protocols and to define protocols for more complex conversations composed of a number of simpler conversations. With this method it is possible (a) to capture the concurrent characteristics of a conversation, (b) to capture the state of a complex conversation during runtime, and (c) to reuse a given conversation structure for processing multiple concurrent messages. It is also possible to monitor a conversation and to visualise, both online and off-line, the progress of a conversation. A prototype implementation of a system supporting predefined conversation protocols as Petri Net models developed for the New Zealand Distributed Information Systems research platform is described.

Cluster-Based Graphs for Conceiving Dialog Systems

2017

We describe an unsupervised modeling of the structure of task-oriented dialogs. The dialogs are related to a given domain (for instance, train booking). This modeling is used as a basis for conceiving a dialog system architecture. The modeling is represented by a graph. It displays the main stages of the dialogs and the transitions between them. Our approach consists in applying coclustering to a representative dialog corpus. Thus we obtain main topics that appear in the corpus. We then compute the topics transitions within each dialog. The resulting graph describes the main topics in the corpus, and their overall sequential organization.

A formal framework for inter-agent dialogues

2001

Abstract We present a logic-based formalism for modeling of dialogues between intelligent and autonomous software agents, building on a theory of abstract dialogue games which we present. The formalism enables representation of complex dialogues as sequences of moves in a combination of dialogue games, and allows dialogues to be embedded inside one another. The formalism can be readily operationalized and its modular nature enables different types of dialogues to be represented.