Daniel Rousseau

A social-psychological model for synthetic actors

1. ABSTRACT WC provide synthetic actors that portray fictive characters by improvising their behavior in a muhimedia environment. Actors are either autonomous or avatars directed by users. Their improvisation is based on the directions they receive and the context, Directions can take different forms: high-level scenarios, user commands, and personality changes in the character portrayed. In this paper, WC look at this last form of direction. We propose a sociaLpsychological model, in which we can define personality traits that depend on the values of moods and attitudes. We show how synthetic actors can exploit such a mode1 to produce performances theatrically interesting, believable, and diverse. The Cybercaf6 is used to test those features.

Towards AI Formalisms for Legal Evidence

The discipline or artificial intelligence for law, for all its accomplishments, e.g., in modelling argumentation has generally not dealt with legal evidence. For the latter domain formalization was developped, instead, in forensic statitics, as well as in jury modelling. We claim that the time is ripe for endeavors aiming at constructing integrated evidence support systems for investigative and legal purposes, combining several paradigms. We introduce our own dramatis personae approach, within AI models of agents and planning.

Speech acts in a connected discourse: a computational representation based on conceptual graph theory

Abstract The purpose of this research is to study how to represent speech acts within texts. We consider that conversations are temporal sequences of connected illocutionary acts which result in discourses. Using Searle and Vanderveken illocutionary logic, we propose an extension of Sowas conceptual graph theory that enables us to represent the conceptual structure of conversations and to model the speech acts that are the basic units of communication in conversations. This approach enables us to represent explicitly several time coordinate systems which are underlying the use of temporal knowledge in discourses: the locutors’ temporal perspective; the localization of temporal situations (processes, events, states etc.); the utterance perspectives of agents who utter sentences. Within an utterance perspective, speakers use sentences with the intention to perform illocutionary acts whose propositional content is represented by conceptual graphs. Our approach also enables us to model several linguistic phe...

Knowledge acquisition from prescriptive texts

There is a growing interest for the application of artificial intelligence in law. Research activities have investigated different areas : formulating legislation with the aid of logical models, legal reasoning, case-based reasoning, developing expert systems applied to the juridical or administrative domains. In project A.C.A.T. (Acquisition des connaissances et analyse de textes), we explore the possibility of creating knowledge bases by exploiting information contained in texts which are used in organizations. Our research focuses on a particular category of prescriptive texts : regulations from the Government of Québec. In order to verify these hypothesis we are developing a knowledge-acquisition system which will enable human specialists to transform a prescriptive text into the form of a knowledge base which can be exploited by an inference engine. We introduce a model which enables us to identify three layers in prescriptive texts : the macrostructure, the microstructure and the dominial component. We describe the general architecture of the knowledge acquisition system which enables us to create “deontic” knowledge bases. We present the main knowledge structures used by the knowledge acquisition sub-system : the text grammars of macrostructure and microstructure.

Interpreting communicative acts and building a conversation model

Researchers and industry are actively developing Software Agents (SAs), autonomous software that will assist users in achieving various tasks, collaborate with them, or even act on their behalf. To explore new interaction modes for SAs which need to be more sophisticated than simple exchanges of messages, we have analysed human conversations and elaborated an interaction approach for SAs based on a conversation model. Using this approach, we have developed a multi-agent system that simulates conversations involving SAs. We assume that SAs perform communicative acts to negotiate about mental states, such as beliefs and goals, turn-taking and special conversational sequences. We also assume that SAs respect communication protocols when they negotiate. In this paper, we describe the conceptual structure of communicative acts, the knowledge structures used to model a conversation, and the communication protocols. We show how an inference engine using ‘conversation-managing rules’ can be integrated in a conversational agent responsible for interpreting communicative acts, and we discuss the different kinds of rules that we propose. The prototype PSICO was implemented to simulate conversations on a computer platform.

SACD: a system for acquiring knowledge from regulatory texts

Abstract Huge amounts of knowledge are available in human organizations in the form of texts. However, there are still few research projects dealing with the problem of extracting knowledge from texts in order to build knowledge bases. We present here a knowledge acquisition system, SACD, which identifies the logical structure of prescriptive texts (regulations, norms), from which it generates a knowledge base composed of a set of so called “eontic rules”. We are currently working on French versions of regulatory texts. In this paper we present the elements of a text grammar which is used by SACD to analyze the logical content of prescriptive texts: modalities, conditions, cross-references etc. Since cross-references are the main elements influencing the interpretation of the text logical content, we show how SACD process them in order to generate deontic rules.