APPARATUS AND METHOD FOR ANSWERING QUESTION BASED ON CONCEPTUAL GRAPH

摘要

As a question-and-answer device, a corpus context word is extracted from a corpus, weights are applied to the corpus context words according to a weighting rule, a query context word is extracted from an input query, and a weight is weighted to the query context words according to the weighting rule. Generates a weighting unit to apply, a plurality of corpus sentence embedding vectors for a weighted corpus, and a query sentence embedding vector for a weighted input query, between the plurality of corpus sentence embedding vectors and the query sentence embedding vector A context similarity determining unit determining a target sentence embedding vector among the plurality of corpus sentence embedding vectors based on the vector similarity of, determining a target concept graph corresponding to the target sentence embedding vector, and at least included in the target conceptual graph A candidate response determination unit determining at least one candidate concept corresponding to the input query by calculating a graph matching score for each of the one or more concepts, and a preset number of concepts from the candidate concepts to the input query It includes a response decision unit to determine in response to.