0118 963 7000

University of Reading

The School of Systems Engineering has an international reputation for its work in the fields of Computer Science, Cybernetics, Electronic Engineering, IT and Robotics.  Of particular note are its excellent Enterprise activities, which include collaboration with Industry, whether it be from industrial placements for students, research and development projects or spin off companies. The school was the first in the University to be involved in Teaching Company Schemes, now called Knowledge Transfer Partnerships (KTPs). These activities started in the late 1980s, led to the formation within the School of a Teaching Company Centre, which is now part of the University's wider Knowledge Transfer Centre, and which works with Schools across all parts of the University. The University of Reading has one of the most successful centres for KTP in the UK.

The collaboration between @UK and the University of Reading began with a single KTP project whose academics were Dr Slawomir Nasuto and Dr Mark Bishop. By the end of that project, when Dr Bishop had moved to Goldsmiths College, it was apparent that there were significant further opportunities for collaboration, and so discussions began involving @UK, Reading and Goldsmiths. These led to the proposal for three linked three-year KTPs, two with Reading and one with Goldsmiths. Joining Drs Nasuto and Bishop, were Dr Richard Mitchell, Dr Victor Becerra and Dr Virginie Ruiz, from Reading and Dr Sebastian Danicic from Goldsmiths. Dr John Howdroyd, the associate on the first project, joined the company and helped supervise the associates on the three projects, Matt Brown, Paul Roberts and Dr Richard Barraclough.

The overall aim of the three projects was to develop a product- and location-aware search engine which would be a key component in the @UK plc eProcurement and eMarketplace platform, adding value by improving the user experience. The three projects focused on spidering, classification and ranking.  Spidering is used to find relevant data from the internet, classification is used to sort the data into appropriate groups (which is necessary for procurement), and ranking is used to sort data in order of relevance in response to queries from a user. The techniques developed were combined and now used to greatly help in the procurement of goods and services. The work has been demonstrated successfully in the NHS market, but can be applied elsewhere.

These projects utilized the expertise and interests of the academics in the areas of machine intelligence, classification and data mining. The projects have provided very useful experience of these methods in the application domain. These were the first three year KTP projects the academics have worked on, and have the added benefit that both Matt Brown and Paul Roberts have been able to register for a PhD as well as doing the work - their thesis should be finished soon.