METHOD AND APPARATUS FOR INCREASING LEARNING DATA BY USING SEMANTIC TENSOR SPACE MODEL

摘要

A method of propagating training data using a Semantic Tensor Space Model includes receiving training data including a plurality of documents, and words and concepts for each of the plurality of documents included in the input training data. extracting information, generating a semantic tensor space model by embedding word and concept information of each extracted document into a matrix, and at least one of a matrix corresponding to a first document constituting the generated semantic tensor space model and exchanging a row or column of R with at least one row or column of a matrix corresponding to a second document constituting the generated semantic tensor space model to multiply the training data.