automatic recognition of company names in linguistic expressions

摘要

The method involves a speech recognition device producing at least one word sequence hypothesis from a spoken statement consisting of one or more words, comparing the hypothesis with company names stored in a database and depending on the result of the comparison selecting a company name as a recognition result. The speech recognition unit produces a probability value per hypothesis that is taken into account during a comparison. Independent claims are also included for the following: a dialog system, especially an information system.