DOCUMENT SUMMARIZATION DEVICE, METHOD, AND PROGRAM

摘要

PROBLEM TO BE SOLVED: To generate a summary by suppressing increase of fragments.SOLUTION: Each weight according to the importance of each n gram including n words included in an input document is calculated by an n gram weight calculation part 32. A target function represented by using the total value of the weight of the respective n grams included in a summary and the number of pieces constituting a continuous work string in the input document as a penalty, and a constraint formula representing the maximum value of the number of words included in the summary are generated by an integer planning problem generation part 36. By a summary generation part 38, whether to select each word included in the input document where the generated constraint formula is satisfied and the generated target function is maximum is determined by solving the integer planning problem, and the words respectively selected in the solution of the integer planning problem are arrayed again according to the order of appearance in the input document to generate a summary corresponding to the input document.