摘要 |
<P>PROBLEM TO BE SOLVED: To provide an expression image estimating device for much more efficiently entering images expressing the image of a text. <P>SOLUTION: A text emotion recognizing part 14 extracts a word from an input text, and reads an emotion vector value corresponding to the extracted word from a word emotion value storage part 12, and guides the emotion vector value of the input text based on the read emotion vector value. Then, an expression image specifying part 15 specifies the emotion vector value recorded in an expression image emotion value storage part 13 based on the emotion vector value of the input text, and outputs the expression image information corresponding to the specified emotion vector value. <P>COPYRIGHT: (C)2005,JPO&NCIPI |