Managing the risks from information --- through semantic information management

• Authors: P. W. Warren;N. J. Davies
• Venue: BT Technology Journal
• Year: 2007

Information is central to the way organisations work -- and with the increasing quantity which organisations must manage, the risks from failure to optimally manage information become ever more significant. The typical organisation uses a wide range of heterogeneous information sources, with differing schemata. This means that information cannot easily be gathered for reporting purposes, and significant connections may be undetected. Moreover, it has been estimated that only 20% of corporate information is in `official' database systems. The remaining 80% is unstructured text, i.e. e-mails, intranet pages, documents, spreadsheets and even transcripts of conversations. This creates risks. Unstructured, informal information is often associated with individuals and when individuals leave, although the information may remain, practical access can be lost. Knowledge management tools are needed to prevent this. More dramatically, when faced with litigation or regulatory requests, companies need effective eDiscovery so as to provide, and be seen to be providing, all the relevant documentation. An important tool in integrating heterogeneous information, and in coping with unstructured information is semantic information management. This paper describes the semantic approach to information management and how it can reduce the risks posed by information. It also describes a research project SEKT, led by BT, which is extending the capability of this semantic approach.