Voice Search -- Information Access via Voice Queries

Ye-Yi Wang

Microsoft Research
One Microsoft Way, Redmond, WA 98052, USA

Voice search is the speech understanding technology underlying many
spoken dialog applications that provide users with the information
they request with a spoken query. The information normally exists in a
large database, and the query has to be compared with an unstructured
field in the database to obtain the relevant information. The popular
applications include directory assistance, music/video management,
business and product reviews, stock price quote, and conference
information systems. This talk discusses the challenges of voice
search in acoustic modeling, pronunciation modeling, language
modeling, spoken language understanding and dialog management, and
surveys the technological solutions to these challenges.