| 摘要 |
PURPOSE:To make it possible to retrieve sentence data also from an index word existing on the way of the sentence data by providing the titled system with an index file generation processing part and a hierarchical index work generation processing part. CONSTITUTION:The index file generation processing part P1 divides the sentence data into respective index words, connects the same kind of index words through a path as nodes and adds an identifier of the sentence data to the path as weight to form an index file F4 having a network structure. In the hierarchical index word generation processing part P2, an entry part S4 transfers the path information of an entry node to a hierarchical index word generation processing part IDG and a hierarchical index word generation part S7 forms a hierarchical index word F5 of each entry node in accordance with the forward and backward directions of path tracing. A retrieved index information matching part S10 matches the sentence data identifier added to the hierarchical index word F5 with a key on the basis of information corresponding to the original document data, edits the retrieved information result and outputs an index page PR. |
