Simon Price’s Research Blog

I’m very much looking forward to getting my copy of Kernels for Structured Data penned by University of Bristol machine learning group alumnus Thomas Gartner.

Synopsis:
This book provides a unique treatment of an important area of machine learning and answers the question of how kernel methods can be applied to structured data. Kernel methods are a class of state-of-the-art learning algorithms that exhibit excellent learning results in several application domains. Originally, kernel methods were developed with data in mind that can easily be embedded in a Euclidean vector space. Much real-world data does not have this property but is inherently structured. An example of such data, often consulted in the book, is the (2D) graph structure of molecules formed by their atoms and bonds. This book guides the reader from the basics of kernel methods to advanced algorithms and kernel design for structured data. It is thus useful for readers who seek an entry point into the field as well as experienced researchers.

I attended UK-KDD 2008, a locally organised, one day symposium held in Bristol on 16 April 2008. As advertised, it provided a very enjoyable and interesting forum for discussion, dissemination and exchange between practitioners and researchers working within the broad field of Knowledge Discovery and Data Mining (KDD). The flavour of the event was very much a conversation between the two worlds of academia and industry/government, with both sides learning in the process. Of relevance to my own work was a talk titled, “Ways to Use Inconsistent Knowledge: Argumentation and Merging” by Dr Anthony Hunter of the Intelligent Systems Group at UCL. This overlapped with Stephen Muggleton’s work on stochastic ILP but also contrasted with it in interesting ways. The words of the late, great Douglas Adams sprung to mind, “We demand rigidly defined areas of doubt and uncertainty!”

Lee Feigenbaum et al have written a very readable and (relatively) jargon free article entitled, “The Semantic Web in Action”, which was originally published in Scientific American and is now reproduced online by permission.
Read the rest of this entry »

The newly announced Google Social Graph API indexes the public Web for XHTML Friends Network (XFN), Friend of a Friend (FOAF) markup and other publicly declared connections, returning web addresses of public pages and publicly declared connections between them.

I attended the Epistemic Networks and GRID + Web 2.0 for Arts and Humanities workshop organised by Dolores Iorizzo at Imperial College’s Internet Centre, London (30 - 31 January 2008). The workshop brought together GRID computing specialists with researchers from Classics, Literature and History who have been involved in the creation and use of electronic resources. The core technologies discussed were: infrastructure; named entity, identity and co-reference services; morphological services and parallel texts; and epistemic networks and virtual research environments.
Read the rest of this entry »