MODEL GENERATION DEVICE, MODEL ADJUSTMENT DEVICE, MODEL GENERATION METHOD, MODEL ADJUSTMENT METHOD, AND RECORDING MEDIUM

摘要

This model generation device generates, by using learning data in a plurality of source domains, model parameters corresponding to a model to be used, and intermediate parameter relevance information that shows the relevance between the model parameters of the plurality of source domains and intermediate parameters. This model adjustment device generates target model parameters, which are model parameters corresponding to a target domain and including the intermediate parameters, on the basis of learnt model parameters for each of the plurality of source domains and the intermediate parameter relevance information. In addition, the model adjustment device determines the intermediate parameters included in the target model parameters by using data for the evaluation of the target domain.