METHOD AND APPARATUS FOR VOICE SEARCHING FOR STORED CONTENT USING UNITERM DISCOVERY

摘要

A method, system, and communication device (100) for enabling voice-to-voice searching and ordered content retrieval via audio tags (830) assigned to individual content (835), which tags (830) generate uniterms (319) that are matched against components of a voice query (201). The method includes storing content (835) and tagging at least one of the content (835) with an audio tag (830). The method further includes receiving a voice query (201) to retrieve content (835) stored on the device (100). When the voice query (201) is received, the method completes a voice-to-voice search utilizing uniterms (319) of the audio tag (830) scored against the phoneme latent lattice model (215) generated by the voice query (201) to identify matching terms within the audio tags (830) and corresponding stored content (835). The retrieved contents (835) associated with the identified audio tags (830) having uniterms (319) that score within the phoneme lattice model (215) are outputted in an order corresponding to an order in which the uniterms (319) are structured within the voice query (201).