Large scale use and commercial application of retrieval technology for spoken content requires a sustained effort focused on merging speech recognition, audio analysis and information retrieval into a concerted discipline with a common vision.
After the success of the 2007 Search in Spontaneous Conversational Speech workshop (SSCS 2007), a second workshop, SSCS 2008, will be held in conjunction with ACM SIGIR 2008 in order to further the cross-pollination between the speech research community and the information retrieval community. The workshop addresses application domains including: conversational broadcast, podcasts, meetings, lectures, discussions, debates, interviews and cultural heritage archives.
We welcome contributions on a range of cross-cutting issues:
• Representation of spoken content for optimal search (e.g., LVCSR, word lattice search, STD on phone lattice)
• Exploitation of evidence beyond word-level (e.g., emotional state, speaker characteristics, topic shifts, audio events)
• Application of text IR techniques to the speech domain
• Speech mining in multimedia data
• Multimodality (integrating features from associated non-speech content)
• Search effectiveness (e.g., evidence combination, expansion)
• Access to large scale collections
• Evaluation resources and benchmarking activities
• Multi-/cross-lingual retrieval
• Cross-media mining (e.g., coupling images or text fragments to speech)
• Interaction design and system development (e.g., query formulation, result presentation, search strategies)
• Spoken audio visualization (e.g. results lists, individual results)
• Spoken query search
Contributions for oral presentations (8 pages) poster presentations (2 pages), demonstration descriptions (2 pages) and position papers for selection of panel members (2 pages) will be accepted.
For submission of papers, please use the submission system.
The templates are available at ACM SIG Proceedings Templates. For Latex chose Option 2: LaTeX2e - Tighter Alternate style
