Can computers talk a universal language?

To make computers speak the same language we cannot just use technology. First an analysis of the main concepts involved is needed: at ISTC this is the job of the Laboratory for Applied Ontology (LOA). Here, theory comes before practice. This simple claim contains a deep overturning of the traditional approach, which used to start straightly from computational models. This means involving both people and computers, which in any case have always people at their ends. 


Databases are made up of codes expressed in binary systems and stored together. But behind data there are concepts: how can they properly be translated so that they maintain the same meaning for everyone? For instance, how can different social services (like health services) build up a common glossary? To answer these questions, an interdisciplinary approach is needed. Both theoretical and technological aspects are essential.

At ISTC the Laboratory for Applied Ontology (LOA) tries to put together Computer Science and Linguistics using the instruments of Philosophy and Logic: its aim is to build up a new unified paradigm of computer language. The key is considering Computer science not just as a technical subject: computers have always people at their end, so it is crucial to build a global service framework able to account for complex processes involving both people and computers.

If we want machines to understand each other and people to agree with the meaning of machines language we need to make words explicit. To do that, we have to face ontology before technology. An ontology is a description of the concepts and of the relations between concepts: its aim is finding unambiguous meanings of words. It is useless to perfectly learn Java or HTML languages if we don't start from day-to-day language. This is the reason why a real interaction between people and technology is essential: it is the only way to build a new unified paradigm of computer language. 

Contact: Nicola Guarino

ISTC Group: Laboratory for Applied Ontology