Banca de QUALIFICAÇÃO: RICARDO YUTAKA ICHIURA

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
DISCENTE : RICARDO YUTAKA ICHIURA
DATA : 16/08/2019
HORA: 14:00
LOCAL: na sala 407, 4º andar, Bloco B, Campus SA da Fundação Universidade Federal do ABC, localizada na Avenida dos Estados, 5001, Santa Terezinha, Santo André, SP
TÍTULO:

Multi-Agent Systems for Pervasive Learning Environments

PÁGINAS: 40
GRANDE ÁREA: Engenharias
ÁREA: Engenharia Elétrica
RESUMO:

Learning in a variety of environments, away from traditional classes, has become more and more 
frequent in today's world. The big propeller for this rupture in the classical practices of knowledge 
transmission is the advances in Information and Communication Technologies - known as such as ICT. 

A learning style that presents itself with great possibilities of success is Pervasive Learning, where learning 
is done with the help of sensor-monitored environments, which help interact with the learner in their learning 
process. Sensors scattered around the environment can monitor the learner's presence, the route made 
by him through the various workstations, his interaction with other apprentices, compliance with safety 
procedures, time spent in each activity and other activities more. At the same time, they can interact 
according to the level of learning and knowledge of each one, presenting consultation material and exercises 
and exclusive and targeted activities individually, according to their degree of development. 

Each outcome of the monitored activities serves to of the learner and, consequently, to update the 
knowledge model of each allowing a central system to modify teaching strategies. for each of them 
individually. The Bloom Taxonomy and the Model of Felder-Silverman were the main models used to classify
 the expected results in the teaching / learning process and to define the each of the parties involved in the 
process. All this structure of sensing and data acquisition, with subsequent conversion of results and 
management of these, integrate a larger system, commonly called the Intelligent Tutor System (STI). 
This composition of hardware and software is intended to manage the process of teaching / learning, 
and consists of four basic models: Knowledge - where all the knowledge about learning in question, 
Student Model - where data from acquired knowledge and learning style of each student or learner, 
Pedagogical Model - where are the driving lines for the and the Interface Model - where the interaction 
between the learner and the STI occurs.

It is in the Interface Model that all interactions with the learner occur within the concept of pervasive 
environments monitored by sensors. The data collected by the various types of sensors are transferred to
 the Tutor System Intelligent through this interface, as well as the response actions to some of these 
collected stimuli are also transmitted through it. The actions of data collection by sensors, responses to
 stimuli and monitoring of the environment are simple actions, from the point of view of design, but 
complex if analyzed by the optics of the diversity of detected stimuli and collected data. A response 
action can be a simple closure of computer file, but also can be the shutdown of the supply of electrical
 energy of a given equipment or recalculation, and subsequent rewriting, of the status level of a particular
 aspect evaluated in an apprentice. This complexity of input and output information can be better managed 
with the use of autonomous agents of a Multi-Agent System (SMA). 

Multiagent systems are computer systems composed of several elements positioned in an environment 
and that are autonomous in complying with the objectives assigned to them. These elements are called 
Agents and can be of two types: Reactive - when they simply react with an action to a stimulus received 
or Cognitive - when it has the ability to infer on one or more stimuli and act according to the result of this
 inference. Agents "feel" the environment and act on it, interfering, according to your goals. This present study
 intends to make an analysis of three free tools and open source for simulation of well-known multiagent
 environments currently available for use: JADE, GOAL and JADEX. This analysis will special focus on 
the creation of interaction environments, since the market does not has a well-defined and established
 standard for this. The aim is to adopt the specifications of a standardization proposal called Environment
 Standard Interface (EIS) and analyze its feasibility for the creation of for the three tools analyzed. 
They will be constructed with the use of one of these tools, prototypes of agents - Reactive and Cognitive
 - for the simulation of its operation in a laboratory of Chemistry of Technical High School, of to create 
an interactive and pervasive environment of studies for students.

MEMBROS DA BANCA:
Presidente - Interno ao Programa - 1761107 - RICARDO SUYAMA
Membro Titular - Examinador(a) Interno ao Programa - 2334927 - ANDRE KAZUO TAKAHATA
Membro Titular - Examinador(a) Externo ao Programa - 2271891 - MARCO AURELIO CAZAROTTO GOMES
Membro Suplente - Examinador(a) Interno ao Programa - 1761105 - MURILO BELLEZONI LOIOLA
Membro Suplente - Examinador(a) Externo ao Programa - 2356637 - KENJI NOSE FILHO