APPARATUS OF SENTENCE SIMILARITY CALCULATION USING MORPHEME TRANSFORM TYPE AND METHOD THEREOF

摘要

The present invention relates to a device and a method for calculating sentence similarity by changing a sentence order and a sentence structure. The method for calculating the sentence similarity by using the device for calculating the sentence similarity comprises: a step of extracting a plurality of morphemes from a comparative object sentence; a step of generating a hash value of the comparative object sentence by using a character code value and a position code value of each of morphemes; a step of determining a morpheme transformation type of the comparative object sentence by comparing the hash value of the comparative object sentence with a hash value of a stored original sentence; and a step of calculating similarity of the comparative object sentence and the original sentence by using the morpheme transformation type. The morpheme transformation type includes a first type of changing contents of a morpheme in the comparative object sentence; a second type of changing arrangement of the morpheme in the comparative object sentence; and a third type of omitting the morpheme in the comparative object sentence. The present invention accurately confirms the similarity between the sentences by calculating the similarity of the sentence according to the morpheme transformation type. The present invention is provided to calculate the similarity by considering a distance between transformed morphemes or a distance between the center of a sentence and a transformed morpheme, thereby accurately calculating the similarity in comparison to similarity calculation through simple pattern comparison.