| 摘要 |
PROBLEM TO BE SOLVED: To provide an information processor, its control method and a program for extracting meta data for precisely retrieving non-text information in a document. SOLUTION: The characteristic information of non-text information such as an image, figure and chart in a document is extracted based on caption information or the like relating to the non-text information. Also, relational information between respective non-text information shown by a symbol such as an arrow symbol connecting a plurality of non-text information in the document is analyzed. Then, meta data for retrieving the non-text information in the document are extracted from the document based on the relational information shown by the characteristic information and symbol of the non-text information. That is, the meta data are extracted under the consideration of the characteristics of the non-text information such as the image, figure and chart or their relations. COPYRIGHT: (C)2008,JPO&INPIT
|
