Nigel Shadbolt seen here during his lively and very well accepted lecture on Web Intelligence.
Nigel started by reminding us of the continuing exponential growth in processing power (Moore’s law has held for four decades), and the facts that the web now has of the order of 108 users and 1010 pages of information.
Nearly all of the pages on the web are designed to be viewed by humans, who, without any particular training, can easily scan a page about a conference, for example, and pick out details such as event timing and speakers.
Such unstructured information is much harder for a computer to process accurately, although because of the vast quantity of information available, we would very much like to be able to sift it automatically.
Nigel reviewed the use of metadata for classifying information. While this is an advance over raw HTML, it requires agreement on the meaning of tags and is hard to extend. Hence the concept of the Semantic Web and the use of an ontology language by means of which the meaning of tags can be defined. One practical example of the use of ontologies is in Gene Ontology.
Nigel described the work of Advanced Knowledge Technologies, one of the IRCs (Interdisciplinary Research Collaborations) funded by the EPSRC and the MRC, and in particular the CS Aktive Space project, winner of the 2003 International Semantic Web Challenge. This continually harvests data from research papers and similar sources and is able to infer who has been collaborating with whom and to create Dynamic Communities of Practice. A practical application of the technology has been in the field of breast cancer screening, where by integrating information from disparate sources such as X-ray, MRI and ultrasound imaging, medical notes, case histories, histopathology etc, a more rapid and accurate diagnosis can be obtained.
He ended by showing us how the AKT visual tools could be used to focus on emerging technologies, again by automatically analysing patterns of publication topics.
Questions flowed after the talk and, when we had to leave the lecture theatre, continued over some liquid refreshment in the Lamb and Flag.