Abstract:

"Advanced Artificial Intelligence" is an optional subject in the Computer Engineering degree at University Jaume I, Castellón, Spain. Its methodology is totally practical and involves introducing students to research by carrying out a small project in Artificial Intelligence.

The initial methodology of this subject was totally open: students should investigate what Artificial Intelligence deals with, select a topic from it, made a state of the art about it and propose a small contribution to it orally in front of their classmates and teachers. The evaluation of the course was based on the content of the work and the presentation of the contribution.

The shortcomings that we found to this initial methodology were: (1) as the scope of work was all the AI, it was too broad and students took too much time to find a bounded problem that could be solved in the time required for the subject, (2) as there was only one evaluation at the end of the course, students could not practice and get feedback to improve their work during the year, and (3) as there was no interaction between students during the academic year, they only learn about their individual topic of research and some of them ended up feeling 'alone' with the problem and unmotivated.

In order improve the success of this subject; an educational innovative project was carried out, which consisted of:
(1) continuing with the methodology of learning by projects because students develop important skills, such as independent learning and problem solving. But limiting the research and the contributions to make to a list of problems to be addressed more specifically, proposed by teachers and negotiable by students, so that results can be obtained faster and students did not discourage;
(2) introducing a form of regulatory assessment of the students’ work through an initial evaluation at the beginning of course, a mid-course evaluation and submission of biweekly reports. This was intended to ease teachers to keep track of students’ work and to force students to work in the subject throughout the course, do not leave it to the end and obtain results of quality;
(3) introducing a formative and qualifying assessment by applying a co-evaluation or a method of assessment between classmates. This is intended to avoid students only being interested on their own works, but to get involved in the work of their classmates as they need to evaluate it. In addition, the co-evaluation is also intended for students to develop general skills of critical and constructive thinking, very important in professional practice.

The general objectives of this project were that students:
1. develop independent learning skills, problem solving and critical thinking, and work skills such as communicating effectively orally and in writing;
2. learn to receive critical feedback on their work and use them to improve;
3. learn to make constructive and critical comments on the work of their peers;
4. learn and apply Artificial Intelligence techniques to solve problems.

This innovative methodology has been applied for three academic courses (2008/2009, 2009/2010 and 2010/2011) with successful results: the students have been involved and motivated in the subject and they also have obtained results of quality, such as, five projects published as conference papers, two projects continued as final degree projects and another one continued in a training stay in an enterprise.

Keywords:

Learning by projects, co-evaluation, and artificial intelligence.