THematic Indexing of Spoken Language
Summary of Activities in 1999
BBC News Retrieval Demonstrator
- A BBC news retrieval demonstrator has been produced containing over 1000 hours of BBC radio and TV news material covering a two year period.
- The demonstrator has been exhibited extensively within the BBC, and also at several international conferences and to potential customers from the media monitoring industry.
- The demonstrator is currently being evaluated by the BBC Information & Archives department. Initial feedback has been encouraging and is being incorporated in an enhanced system.
- Versions of the demonstrator tailored towards BBC Monitoring and the BBC Natural History Unit are also being developed.
- The THISL project participated in the Spoken Document Retrieval (SDR) track at the TREC-8 text retrieval evaluation organized by NIST in the USA.
- The THISL system performed extremely competitively.
Spoken Query Interface
French Language System
- Work has begun on adapting the THISL system to cover French language news.
- A database of RTBF news broadcasts has been recognized and indexed.
Other Research Developments
- Although the THISL project concludes on January 31 2000, the BBC Demonstrator and related technologies will continue to be developed at the BBC, Sheffield University and SoftSound.
- Industrial partners SoftSound and Thomson-CSF are exploring ways of exploiting the technology commercially.