vocabulary learning speed prediction parameter generator and vocabulary learning speed prediction apparatus and their method and program

摘要

PROBLEM TO BE SOLVED: To provide a word learning speed prediction parameter generation device capable of predicting word learning speed after vocabulary explosion using small accumulated vocabulary data.;SOLUTION: The word learning speed prediction parameter generation device includes a word learning speed calculation section, a feature amount calculation section and a parameter extraction section. The word learning speed calculation section inputs vocabulary leaning data including a set of vocabulary obtain days and accumulated vocabulary in vocabulary learning process of plural infants to calculate linear function by linear regression between obtaining days in which a period the vocabulary does not increase (plateau) is removed or obtaining days after a vocabulary explosion and accumulated vocabulary to obtain word learning speed v_j of each infant j. The feature amount calculation section calculates a function of a combination of ratios including the ratio of part-of-speech information of the words from the vocabulary leaning data of each infant j as feature amount c_j of each infant j. The parameter extraction section inputs the word learning speed v_j and the feature amount c_j to obtain a regression model between the word learning speed v after the vocabulary explosion and the feature amount c and outputs a parameter for the regression model.;COPYRIGHT: (C)2014,JPO&INPIT