发明名称 DOCUMENT ANALYSIS DEVICE AND DOCUMENT ANALYSIS PROGRAM
摘要 <p><P>PROBLEM TO BE SOLVED: To shape text data of a mail so that a further accurate sentence unit can be recognized even when a citation layer is disordered in the course of a plurality of exchanges of the mail. <P>SOLUTION: A document layer determination part 30 determines data length and citation layer of each line of the text data of the mail, and a same layer determination part 31 determines a character string of the same layer based on the data length and citation layer of each line determined by the document layer determination part 30. A document shaping part 32 connects lines determined to be in the same layer by the same layer determination part 31 to form a connected character string, and a sentence recognition processing part 33 identifies a sentence from the connected character string, and forms text data of the mail before the citation layer is disordered. <P>COPYRIGHT: (C)2010,JPO&INPIT</p>
申请公布号 JP2010108326(A) 申请公布日期 2010.05.13
申请号 JP20080280765 申请日期 2008.10.31
申请人 TOSHIBA CORP;TOSHIBA SOLUTIONS CORP 发明人 SHIBUYA TAKASHI;YOSHIMURA YUMIKO;SHINDO MASAKI;SAI ENKO
分类号 G06F17/21;G06F13/00 主分类号 G06F17/21
代理机构 代理人
主权项
地址