首页>
外国专利>
Method for indexing set of documents distributed in semantic space in relational database, involves obtaining prototype vector, and obtaining vector running from list according to vector associated with documents forming current list parts
Method for indexing set of documents distributed in semantic space in relational database, involves obtaining prototype vector, and obtaining vector running from list according to vector associated with documents forming current list parts
The method involves providing a creation phase in which a weighted matrix or word/document is created (301). A regrouping phase is provided for regrouping of documents (302) according to occurrence of terms of the matrix within a regrouping list (LG), where the list belongs to a list of documents including one document of the list of documents. A prototype vector is obtained (303), where the vector is associated with the regrouping list. The vector running from the list is obtained according to a vector associated with the documents forming current list parts. Independent claims are also included for the following: (1) a method for performing research in an assembly of documents (2) a computer program comprising instructions for performing a method for indexing a set of documents (3) a computer program comprising instructions for performing a method for research in the assembly of documents.
展开▼
