DOCUMENT-PROCESSING DEVICE, DOCUMENT-PROCESSING METHOD, AND PROGRAM

PROBLEM TO BE SOLVED: To extract the feature quantity of a document and determine similarity at a very high speed. SOLUTION: A document-processing device is equipped with: a data-inputting part 101 for inputting document data to be processed; a morpheme analyzing part 102 which divides a document into morphemes; a statistic data analyzing part 103 which acquires the frequencies of the respective morphemes in the document; a waveform converting part 104 which associates the respective morphemes with frequencies and also associates the frequencies of morphemes with intensities to generate frequency spectrum data which correspond to the document data; a DA conversion part 105 which converts the frequency spectrum data to waveform data; a pass filter part 106 which filters the waveform data by means of specific processing; and a resonance filter part 107 which extracts similar waveform data from among a plurality of waveform data after filtering. COPYRIGHT: (C)2011,JPO&INPIT