This work is partly supported by the National Grand Fundamental Research 973 Program of China (No.2004CB318106), the National Natural Science Foundation of China (No.10574140, No.60535030) and Beijing Municipal Science & Technology Commission (No.Z0005189040391).
This paper presents a fast fuzzy search algorithm to extract keyword candidates from Syllable confusion networks (SCNs) in Mandarin spontaneous speech. Since the recognition accuracy of spontaneous speech is quite p...