Matsushita Electric Industrial Co., Ltd.

Masako Nomoto

Naohiko Noguchi

Yuji Kanno

Mitsuhiro Sato

Mitsuaki Inaba

Hideki Yasukawa

Yoshio Fukushige

A system for providing keywords to facilitate a search in a text retrieval system. For each of texts constituting a text base, the system creates a word ID of each of words used in the text and a word occurrence count of a corresponding word. The word occurrence count indicates a number of occurrences of a word in each text. For each of words used in any of the texts constituting the text base, the system creates a total word occurrence count and a containing text count indicative of the number of texts containing the word. For each of words contained in the selected texts, a degree of importance is calculated by using the word occurrence count, the total word occurrence count and the containing text count. The words contained in the selected texts are sorted in order of the degree of importance. At least a part of the sorted words are displayed as related keywords.

Keyword extracting system and text retrieval system using the same

A document retrieval system for searching a document coinciding with a retrieval request the user inputs and further ranking the document in accordance with the degree of coincidence between the document and the retrieval request. In the document retrieval system, a word frequency calculating section finds out the number of documents where a word appears, a frequency of occurrence of the word in a document and obtains a weighting parameter for the word, and a frequency score calculating section obtains a frequency score on the basis of the output of the word frequency calculating section. In addition, a word cooccurrence relation checking section checks word cooccurrence relations of the retrieval request and the document, and a cooccurrence score calculating section calculates a cooccurrence score from the degree of coincidence therebetween. A document score calculating section calculates a document score on the basis of the frequency score and the cooccurrence score. The documents are ranked in order of document score and displayed to the user.

Document retrieval system

One or more important sentence structural functional units are extracted from an important portion of each of a plurality of documents, and one or more input sentence structural functional units are extracted from an input sentence. Each of the functional units is a noun phrase, a verb phrase, a pair of nouns connected each other through a preposition or the like. The important sentence structural functional units of each document are collated with the input sentence structural functional units to detect a matching degree, and a similarity level is determined from the matching degree for each document. Thereafter, occurrence frequencies of retrieval words existing in each documents are collated with occurrence frequencies of query words existing in the input sentence to determine an estimation value according to an estimation function for each document, the documents are ranked on condition that one or more documents of a high similarity level are ranked to a higher rank than one or more documents of a low similarity level and a plurality of documents of the same similarity level are ranked in the order of decreasing their estimation values. Accordingly, because the similarity level is determined for each document by considering the functional units, a user can retrieve one or more necessary documents with a high accuracy.

Growing community of inventors

Tokorozawa, Japan

Masako Nomoto