Research
As a research fellow at the Computer Science Department of Udine University I am currently involved in the following research areas:
- Software engineering, which is my primary activity;
- Natural language processing, information filtering and retrieval, automatic text classification, and neural networks, that are fields of artificial intelligence in which I worked in the past;
- Digital libraries, where my research focus on service-oriented architectures, knowledge representation formalisms, and Web 2.0 technologies. All these are fields where both software engineering concepts and artificial intelligence techniques can marriage.
Software Engineering
What does Computer-Aided Software Design (CASD) really means, in particular concerning software quality and Model-Driven Engineering? Current state-of-the-art modeling tools do a pretty good job supporting the most important features of the UML standard notation, but they completely lack the "assistant" role they promise, especially with respect to assuring quality in the software design process. Designing good software is definitely more than building pretty UML diagrams. We believe that good software comes from a solid understanding of software engineering principles, a strong knowledge of the problem domain, and a good experience. In this context, tools should aid the designer to catch critical aspects of a modeled system as early as possible, and in a possibly unobtrusive way. My research activity is prevalently focalized on building such tools. At the moment I am working on two broad, but interconnected areas: assuring quality in the design process and check the consistency/correctness of the model as a preliminary step to build correct software.
Natural Language Processing
I am involved on NLP techniques since 1999 when I started my Master degree thesis on computer science, addressing the automatic categorization of Web pages and building systems based on software agents, semantic networks, and Kohonen self-organizing neural networks. More recently, my research shifted a little bit, moving from automatic unsupervised clustering to automatic supervised text categorization based on user profiles. In 2008 I focused more specifically to Semantic Web and Web 2.0 as a new empowered (collaborative) environment where to apply NLP algorithms.
Semantic Digital Libraries
In 2008 I joined several research groups in Udine University working on a FIRB project concerning digital libraries. My research is targeted to integrate the ideas, strategies, and techniques proper of the NLP discipline to the concept of digital library. The main goal of this research is to transorm traditional digital libraires into semantic digital libraries.