ICECE2000
International Conference on Engineering and Computer Education

Paper Title:
QSabe II: A Cooperative Service for Knowledge Appropriation and Diffusion Using the Internet.
 

Authors:

Crediné Menezes (UFES); José Pessoa (UFMT); Orivaldo Tavares (UFES)

Affiliation: UFES - Universidade Federal do Espírito Santo
                UFMT - Universidade federal de Mato Grosso

E-mail : credine@inf.ufes.br jmpessoa@gaia.ufes.br tavares@inf.ufes.br
 

Abstract:

The modern learning practices are strongly based on Interactivity. Teachers and learners get involved in a complex and rich process of understanding. A central tool for this process is an environment that supports the formulation of questions and answers.
Today, question and answer services have a certain degree of automation, usually related to such areas as: indexing, information storing and automated search for resources.
Three aspects of these services are still open. The first one refers to the system’s ability of getting a perfect “understanding” of the content of a natural language written text. The second one is related to the system’s capacity of  “choosing” the right person to answer some question. Finally, the last one regards the system’s ability to answer the questions itself. In practice, human agents have to solve all these problems.
The amount of messages (questions, answers, opinions, etc) recorded in cooperative environments reaches intractable dimensions, without automated support. To address this problem, we are developing an environment called QSabe II: A Cooperative Service for Knowledge Appropriation and Diffusion Using the Internet.
This environment is based on the QSabe system, an intelligent system for questions routing. Using the groupware metaphor, the QSabe agent acts as an interpersonal assistant, allowing each member of the group to participate in the process, formulating or answering questions. The agent uses both text comprehension techniques and machine learning theories. In fact, the system is able to interpret the questions about a pre-established subject, and it is capable of dynamically learning the profile of the collaborator user (the expert in certain subject). This way, the system can route the message to the most convenient expert.
The QSabe II enhances the functionalities of the existing system. The new system has the following requirements: extract and compare the questions profile; identify semantics based on onthologies and natural language processing; capture the degree of difficulty of each question and capture the degree of ability of each responder.

--------------------------------------------------------------------------------------------------------------------------
Contact:
UFES - Universidade Federal do Espírito Santo
Department :   Informática
Address:   UFES - Campus de Goiabeira  - Vitória/ES
Country :   Brazil
Phone:   (27) 335-2689
Fax :       (27) 335-2850