| 摘要 |
PURPOSE: To reduce the labor and time for image generation by automatically acquiring information required for the generation of an image from natural language by which the content of the image is described. CONSTITUTION: This device is equipped with a natural language processing part 102 which analyzes the natural language using the information stored in a language dictionary 101, an information managing part 104 which manages the information analyzed by the natural language processing part 102 and also, performs inference using knowledge in a knowledge data base 103, an image data base 105 in which the image is accumulated, an image data base managing part 106 which performs the registration of the image on the image data base 105 and retrieval from and arrangement on the image data base 105 according to the information managed by the information managing part 104, and an image display part 107 which displays image data, and two pieces of information on a noun that becomes an image component in an input natural language text and spatial relation between the image components of the input natural language text are sampled. |
