Programming Research Group

Knowledge Fusion Platform for Humans and Machines: PinWeb

PinWeb is a hybrid declarative application being developed under a research program at the Programming Research Group of University of Tübingen, Germany. The goal of PinWeb is to create a knowledge fusion platform which can be utilized by humans and machines to manage, share and merge their knowledge. By means of various existing tools and applications, both humans and machines can collect knowledge in diverse domains, which must not necessarily be similar in its representation. The heterogeneity of knowledge representation is actually not a big issue as it is assumed, since Gruber says that the problem of knowledge representation in different languages can be addressed by following the approach of translation. According to Berners-Lee's vision, web content needed to be composed in a syntactically and semantically correct manner, so that it is understandable by machines.

By contrast, autonomous interpretation (without human intervention) of a web page proves yet to be a challenge for machines. As a matter of fact, machines need human knowledge to work in the virtual and real world which is still human oriented. If the machines had the ability to translate and extract ontology from some information source using preexisting ontologies developed by humans, just as humans do while acting in new domains or reading books, both could share their collected knowledge among them. Following this idea of collective intelligence of humans and machines, this project introduces an extensible plug-in Java-based application which abstracts over different internal knowledge representations, facilitates knowledge interoperability and supports diverse reasoning methods and language independent knowledge queries with the help of existing technologies.

Contact: Mandeep Singh

Address

Universität Tübingen

Wilhelm-Schickard-Institut für Informatik

Jasminweg 21

72076 Tübingen

GERMANY