Modelo para a construção de sistemas tutores multiagentes

• Ano de defesa: 2012
• Autor(a) principal: Silva, Marlos Tacio
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Alagoas; Brasil; Programa de Pós-Graduação em Modelagem Computacional de Conhecimento; UFAL
• Palavras-chave em Português: Mathema; Petri, Redes de; Ontologia; Petri, Networks of; Ontology; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://www.repositorio.ufal.br/handle/riufal/1821

This work is situated in an interdisciplinary research program on computational modeling of knowledge, focusing on one challenge in the field of Intelligent Tutoring Systems with respect to manage the complexity involved in effectively building such systems. In this context, three aspects related to the mentioned challenge were addressed: (1) lack of concrete guidelines to be used by the involved actors (i.e., authors and developers) in the process of building this environments; (2) conceptual lack between the author’s knowledge and the available tools for that end; and (3) lack of a flexible and adequate software architecture for building intelligent software entities. Thus, this work aims to present a systematic approach with models to help the construction of Multiagent Tutoring Systems based on Mathema’s architecture. From the author point of view, this systematic aims to modeling a given domain via a graph structure. Based on this structure we derive a Petri net, to check both structural and behavioral properties and a knowledge base, which will be operated by a pedagogical planner. From the viewpoint of the developer, this systematic aims to use the graph structure defined by the author to identify a set of tutor agents and, thereafter, builds these agents based on flexible software architecture. For the empirical evaluation, we develop a case study consisting in structuring a course of Computer Science. In addition, more specific studies were developed, one in the context of Computational Logic and another in the context of Machine Learning. These studies demonstrate the feasibility of using the proposal, obtaining satisfactory results in the solutions presented to answer the research questions addressed.