Research On Calculating Method Of Chinese Speech Audiometry

In recent years, Chinese speech audiometric tests are widely used in clinical evaluation of the patients with hearing loss. The speech assessment methods of equivalence about audiometric materials are always depending on large-scale data collection. Along with the needs of smaller scale word list for clinical application is more and more intense, a standardized method for measuring the difficulty degrees of the word list is pressing demands in speech audiometry field.To assess the equivalence of clinical speech audiometric materials, we need to calculate the perceptual articulation of usual language units(monosyllable, disyllabic word, short sentences, etc). This paper investigates the independent contribution of initial and vowel based on ZHANG’s conditional probability model[22], emphasizing the effect to the perceptual articulation of monosyllable, disyllabic word, which is resulting from the factors of tone, junctures and word frequency. We organized a monosyllable and a disyllabic word audiometric tests to get data base just as the phonematic perceptual confusion matrix and the measured value of articulation in some transfer conditions (all-pass, LP1600, HP1600, etc). These tests help to determine the frequency coefficient of the perceptual definition model. Compared with the existing independent probability model and conditional probability model, the new perceptual model has a lesser deviation degree and a better correlation with the measured value.The main work contents are accomplished as follows:1. We put forward a procedural perception model for monosyllable articulation. We redescribed the process of monosyllabic perception by analyzing the existing independent probability model and conditional probability model. The overall-perception model as an improved one is put forward by detailing these three factors--innersyllabic juncture, tone articulation and syllablic appearance frequency.2. We put forward a procedural perception model for monosyllable articulation.The monosyllabic articulation test organized in this research proved the improvement of overall-perception model. On the other hand, this model is not perfect on reflecting the influence produced by other similar phoneme. As the analysis of the front, a new model called process perception model is raised. The test data proved this model is much better than the models mentioned earlier.3. We put forward a procedural perception model for disyllabic word articulation.In order to examine the Influence degree produced by these three factors-the disyllabic tone juncture, the disyllabic juncture and the word frequency, a disyllabic word articulation test is organized. A word perception model is set which is proved effective for calculating disyllabic intelligibility by adjusting the model parameters based on the test data.