Abstract: Document classification systems are valuable tools for searching and retrieving classified documents but can be prohibitively complex and cumbersome for users. A system for the indexing and retrieval of classified documents inserts keywords, titles or definitions of previously applied classifications into the document record and provides the resulting record to a search engine. Searchers are able to retrieve documents by searching on keywords from the classification system without looking up class coding.

Abstract: Document classification systems are valuable tools for searching and retrieving classified documents but can be prohibitively complex and cumbersome for users. A system for the indexing and retrieval of classified documents inserts keywords, titles or definitions of previously applied classifications into the document record and provides the resulting record to a search engine. Searchers are able to retrieve documents by searching on keywords from the classification system without looking up class coding.

Abstract: Textual elements and unambiguous locations paths corresponding to textual elements and/or their ancestors are extracted from a tree-structured document such as an XML document and stored in relational database structures. Textual elements are stored in a table comprising a column of textual elements and an identity column. The unambiguous location paths are stored in a second table in rows comprising the location path, the identity form the first table corresponding to the first textual element that is a descendant of the location path, the identity from the first table corresponding to the last textual element that is a descendant of the location path, and the name of the element located by the location path.

Abstract: A multilingual electronic transfer dictionary provides for automatic topic disambiguation by including one or more topic codes in definitions contained the dictionary. Automatic topic disambiguation is accomplished by determining the frequencies of topic codes within a block of text. Dictionary entries having more frequently occurring topic codes are preferentially selected over those having less frequently occurring topic codes. When the topic codes are members of a hierarchical topical coding system, such as the International Patent Classification system, an iterative method can be used with starts with a coarser level of the coding system and is repeated at finer levels until an ambiguity is resolved. The dictionary is advantageously used for machine translation, e.g. between Japanese and English.

Abstract: A multilingual electronic transfer dictionary provides for automatic topic disambiguation by including one or more topic codes in definitions contained the dictionary. Automatic topic disambiguation is accomplished by determining the frequencies of topic codes within a block of text. Dictionary entries having more frequently occurring topic codes are preferentially selected over those having less frequently occurring topic codes. When the topic codes are members of a hierarchical topical coding system, such as the International Patent Classification system, an iterative method can be used with starts with a coarser level of the coding system and is repeated at finer levels until an ambiguity is resolved. The dictionary is advantageously used for machine translation, e.g. between Japanese and English.