Address the complexity of the Web with views
Keywords: Artificial Intelligence, AI, ML, Machine LEarning, RDF, SQL, query, view
Supervision: Ben De Meester Anastasia Dimou
Students: max 1
The web contains a massive amount of knowledge. If a person wants an answer to a question based on this knowledge, they might need to crawl the whole web to obtain it. In fact, this is exactly what certain big companies do: they crawled the whole web, processed its knowledge and earn a lot of money by answering users' search queries. Giving power back to the people would require a small agent to be able to answer their queries without downloading and processing terabytes and terabytes of data. One possible way to achieve this is using views. Views can expose reasonably sized parts of an otherwise huge dataset and a user can, based on view descriptions, select which views are relevant to their information need before downloading them. There is still much to be researched about views on the web: tradeoffs between simplicity and expressivity, query answering strategies, …
In this master thesis you will (i) investigate new ways to define views on the web, (ii) investigate ways to answer queries using this new type of views and (iii) evaluate the performance against existing strategies.