WG7: Intelligent Tutoring Systems

Objectives

Program debugging has always played an important role in software development and much effort has been spent in order to devise formal techniques being able to diagnose, locate and automatically correct program bugs. As a matter of fact, every programming language is nowadays equipped with debugging facilities which can aid users to develop error-free programs at different levels of automation.

Moreover, these tools permit to avoid laborious proof-reading sessions just for discovering subtle errors in the code. Especially in the context of declarative programming (e.g., functional, logic and functional logic programming) a lot of diagnosis methodologies have been successfully developed in the last thirty years, giving rise to a research area called declarative diagnosis, whose main aim consists in trying to automatize the debugging process of declarative programs as much as possible.

 Description of work

The main activity of this workgroup consist in developing intelligent tutoring systems that can support the learning of a declarative programming language by computer science students. These systems aims at:

1. helping students to recognize syntactic as well as semantic errors in their programs by means of novel abstract diagnosis techniques;

2. facilitating the correction of faulty programs by suggesting possible ways to repair bugs; those kinds of corrections can be computed by using machine learning methodologies;

3. providing error explanations by comparing wrong programs with the corrected ones.

Intelligent tutoring systems seem to be a promising, powerful and natural e-learning mean, which can definitely shorten the time needed to study a given (declarative) language and speed up the process of acquisition of some peculiar programming skills.

Experts involved in this workgroup: 6

Deliverables

The deliverable of the project will be: scientific publications; demonstrations of prototypes.