The processing of speech recognition is influenced by several factors. Lexical effect is one of these factors. The relationship with lexical characteristics and speech recognition result is explained in the NAM(Neighborhood Activation Model)theory. Le ...

The processing of speech recognition is influenced by several factors. Lexical effect is one of these factors. The relationship with lexical characteristics and speech recognition result is explained in the NAM(Neighborhood Activation Model)theory. Lexical characteristics are defined as word frequency, neighborhood density, neighborhood frequency, etc. Word frequency and neighborhood density are reported as significant variables for speech recognition in Anglophone studies. The purpose of this study is to evaluate lexical effect on the speech recognition process of monosyllable Korean words. Forty adults with normal hearing participated this study. They heard the four lists depending on different lexical characteristics(HFLD, LFLD, HFHD, LFHD, H: high, L: low, F: frequency, D: density) in an audio booth with 20dB SL loudness. The results revealed that a significant difference in speech recognition occurred regarding word frequency. Neighborhood density was not a significant factor in predicting speech recognition. This finding could be explained through language specific characteristics. For example, the number of monosyllabic words in Korean language is limited as compared to the English language. Also only seven phonemes can be placed in the word final position of the CVC structure. Theses characteristics are possible factors to explain the difference of this result with Anglophone study about lexical effect on speech recognition. These results would provide meaningful basic data for making a Korean Monosyllabic Speech Recognition Test.