发明名称 |
DATA PROCESSING DEVICE, DATA PROCESSING METHOD, AND PROGRAM |
摘要 |
<p>A learning-sample data extraction unit (140) extracts sample mails of a plurality of categories. On the basis of a learning sample selection rule (190) indicating the upper limit value of the total number of sample mails to be selected and the selection criteria of the sample mails, a leaning sample selection unit (150) calculates by the unit of the category the number of sample mails to be selected which maximizes within the upper limit value the total number of sample mails to be selected and minimizes the difference between the categories, and selects the sample mails to be used for learning by a learning unit (160) by each category according to the number of sample mails to be selected calculated by the unit of the category from among the sample mails extracted by the learning-sample data extraction unit (140).</p> |
申请公布号 |
WO2011052025(A1) |
申请公布日期 |
2011.05.05 |
申请号 |
WO2009JP68317 |
申请日期 |
2009.10.26 |
申请人 |
MITSUBISHI ELECTRIC CORPORATION;SHIBATA, HIDEYA;KATO, MAMORU;KORI, MITSUNORI |
发明人 |
SHIBATA, HIDEYA;KATO, MAMORU;KORI, MITSUNORI |
分类号 |
G06F17/30 |
主分类号 |
G06F17/30 |
代理机构 |
|
代理人 |
|
主权项 |
|
地址 |
|