Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
SOCIS: A Prototype Scene of Crime Information System Khurshid Ahmad, Department of Computing, University of Surrey, Guildford, Surrey. GU2 7XH The Scene of Crime Information System project, sponsored by the UK Engineering and Physical Sciences Research Council, is investigating this link through the construction of a computer system - the prototype Scene of Crime Information System (SOCIS). This is a three-year collaborative project that began in 1999, involving the Universities of Sheffield and Surrey, working closely with four UK Police Forces - Hampshire, Kent, South Yorkshire and Surrey – and together with the Metropolitan Police Scientific Support College. The prototype can be made available to Police Forces and other enterprises involved in the administration of justice. The system has been developed in the Java programming language and is based on a 3-tier architecture including a client, a server and a database and can thus be accessed via local Intranets as well. The SOCIS system can be regarded as a variant of image retrieval systems in that it uses not only the vision-specific features of an image, particularly colours in the image and some data about possible shapes of objects, but also uses the terms found in a collateral description of the image. The SOCIS system has three other distinguishing features: First, the system can extract specialist terms and organise the terms in a conceptual hierarchy; second, the system has a component that can autonomously learn vision-specific features of an image and learn to associate the keywords used in the description of the image with the vision-specific features; and third, identify meaning-bearing relations amongst the objects depicted in the image. The SOCIS project can be described as a user-oriented project that is at the same time investigating an open-ended question - the relationship between an image and its description. Biographical Note: Khurshid Ahmad is Professor of Artificial Intelligence and the Head of Department of Computing, University of Surrey. He works in the areas of information extraction and neural computing. His work has concentrated on extracting terms automatically from texts, summarizing large texts and on building neural computing systems comprising a number of neural networks. The SOCIS project has two principal investigators: Khurshid Ahmad at Surrey and Yorick Wilks at Sheffield. The research officers working on the project are: Bogdan Vrusias (Surrey), Horatio Saggion and Katerina Pastra (Sheffield). Mariam Tariq is working on her doctoral dissertation for conceptually organising a set of candidate terms extracted from texts. Chris Handy is looking at the variability in the description of images. Both Mariam and Chris are at the University Surrey. John Armstrong, Head of Scientific Support, Surrey Police, chairs the SOCIS Round Table. The Round Table includes Andy Hawley (South Yorkshire Police), Tim Curtis (Hampshire Police), Jim Fraser (Kent Police), David Ince (Principal Scientific Support College, Metropolitan Police). Bob Milne (Metropolitan Police Forensic Intelligence Unit). Enterprises that have attended some of the Round Table meetings include Solcara PLC, CME Software Systems, Advantage Systems, CMG Admiral and the Forensic Science Services.