Rating scale data for evaluating coded speech quality by the mean opinion score (MOS) are factor‐analyzed to show the possibility of extracting rich information from them. Two factors, softness‐shrillness and clearness‐noisiness, are extracted. Factor scores expressing a kind of multidimensional scale of speech quality are related to several physical parameters of speech samples, e.g., differential spectral distortion, phase distortion, and dissonance. Individual subject differences in evaluating speech quality are extracted as the factor loadings. It is found also that speech specialists and nonspecialists place different weights on different factor axes.

This paper presents a language model and its application to sentence structure manipulations for various natural language applications including human-computer communications. Building a working natural language dialog systems requires the integration of solutions to many of the important subproblems of natural language processing. In order to materialize any of these subproblems, handling of natural language expressions plays a central role; natural language manipulation facilities axe indispensable for any natural language dialog systems. Concept Compound Manipulation Language (CCML) proposed in this paper is intended to provide a practical means to manipulate sentences by means of formal uniform operations.