Zenshiro Kawasaki

On the equivalence between scarce-state-transition viterbi decoding and syndrome decoding of convolutional codes

In this paper, we show that Scarce State Transition (SST) Viterbi decoding is equivalent to syndrome decoding by transforming the code trellis module for the main decoder in an SST Viterbi de- coder to a syndrom6 trellis module.

Translator knowledge base for machine translation systems

This paper presents a new method for utilizing translator knowledge bases for machine translation systems. Translator knowledge to be stored and utilized consists of ‘translationally equivalent pattern pairs’: surface-level phrasal, clausal, and sentential correspondences between the source and target languages. This knowledge will be utilized to translate domain-specific idiomatic, nonstandard, or ungrammatical expressions. The proposed method has been implemented in an adaptive English to Japanese machine translation system, HICATS/EJ, as one of its customization facilities.

Symbol reliability estimation using the bidirectional Viterbi algorithm with a code trellis and a syndrome trellis for QLI codes

In this paper, we show that in the case of quick look-in (QLI) codes, the symbol reliability values are obtained by applying the bidirectional Viterbi algorithm (BIVA) either to the code trellis or to the corresponding syndrome trellis.

Symbol reliability estimation using code trellis degeneration for QLI codes

We clarify the relation between trellis degeneration (TDG) and symbol reliability estimation using the bidirectional Viterbi algorithm (BIVA) for the case of quick look-in (QLI) codes.

Small class organization in individualized instruction

Abstract A new method for small class organization is presented. Learners of similar mastery levels are classified into predetermined groups based on the learner-states and group characteristics. The method is applicable to organization of supplementary lectures for individualized instruction or conventional classroom teaching.

A new approach for assigning individually prescribed problems

Abstract This paper describes an exercise-problem selection method which is based on the recently proposed theory of Learning Diagnosis and Treatment (LDT). An optimum problem for each learner is identified by comparing the required readiness for the problem and the learners mastery level. The proposed method will find wide applications in various instruction systems for assigning suitable problems to each learner automatically.

A new expression method for decision logic in learning diagnosis and treatment

Abstract An expression method is presented for decision logic in the Learning Diagnosis and Treatment (LDT) theory reported recently. A conditional AND function is introduced as the basic logic element. The notation of the proposed logic expression requires ten figures (0, 1, 2, 3, …, 9), two special symbols, i.e. (and), and a space. The method is especially suitable for inputting LDT decision logic into computers via punch cards and simple keybord terminals.