Research topics

We are active in the areas related to the management and analysis of large volumes of data and knowledge. Our overall goal is to process simple and (often) uncorrelated facts about individual entities, and organize them into a meaningful context that can significantly enhance the human capacity to take effective actions in varied and uncertain situations.

More specifically, we are interested in designing and developing management systems for different types of data (relational, XML, social, textual, streams, business process data, and others). We explore techniques for efficiently storing, querying, and analyzing these data. Our research includes, but is not limited to, the development of novel processing techniques for static and continuous data, and the identification of structure, correlations, or other patterns in large collections of data. This line of research addresses the needs of business, scientific, and operational environments for organizing, processing and mining huge amounts of data, both in an offline and an online (i.e., real-time) fashion. 

Large part of our research is on Social data, i.e., any form of data that has been generated by users, independently of whether these are documents, short texts, or actions like likes, purchases, or connections. We are trying to develop models of query answering techniques on such systems that go beyond the traditional similarity search and try to identify the goals that users have in mind and predict the elements that they are actually interested in. 

Special focus in all these have the Big Data, i.e., data that are so large, so heterogeneous and have so large arriving rate that existing technologies fall short in processing them. 

We are studying the problems of data management in peer to-peer systems. In particular, we are exploring the applications of game theory in this environment, modelling the cooperation of nodes to work on answering queries, an extensive game, and studying conditions under which subgame perfect equilibria exist, and when they are optimal. Another direction of our research is working on problems related to information integration and metadata management. We propose a methodology for the efficient management of metadata, which makes use of existing infrastructures. We are building tools, based on rigorous foundations, for schema and ontology mapping, interoperability, data translation, information integration, data exchange, view updates, view maintenance, and meta-data management. These tools are designed to support the evolution of data, while requiring minimum human support.

We are working on knowledge representation and knowledge management, with a focus on how to achieve interoperability across multiple local representations. This approach enables the effective management of diversity in knowledge. By treating diversity as a feature (and not as a problem), we can devise algorithms and systems that support the user in the creation, acquisition, adaptation, evolution, and sharing of knowledge. At the foundational level, we work on the logics of context in knowledge representation and reasoning, and on the notion of identity in web-based knowledge representation. The results of this research are applied mainly in the area of the Semantic Web, where we are working on the development of a global Entity Name System and entity-centric applications in the areas of knowledge management, authoring, search.

Faculty members

Paolo Bouquet
Fausto Giunchiglia
Gabriel Mark Kuper
Yannis Velegrakis (coordinator)