Document search system, document search method, program, and non-transitory computer readable storage medium

Patent metadata

Field	Value
Publication number	US-12488011-B2
Application number	US-202418635181-A
Country	US
Kind code	B2
Filing date	Apr 15, 2024
Priority date	Mar 23, 2018
Publication date	Dec 2, 2025
Grant date	Dec 2, 2025

Abstract

A highly accurate document search, particularly a search for a document relating to intellectual property, is achieved with an easy input method. A document search system includes a processing portion. The processing portion has a function of extracting a keyword included in text data, a function of extracting a related term of the keyword from words included in a plurality of pieces of first reference text analysis data, a function of giving a weight to each of the keyword and the related term, a function of giving a score to each of a plurality of pieces of second reference text analysis data on the basis of the weight, a function of ranking the plurality of pieces of second reference text analysis data on the basis of the score to generate ranking data, and a function of outputting the ranking data.

First claim

The invention claimed is:

1 . A document search device, for searching a document related or similar to an input document, comprising a processing portion that is configured to:
obtain a first distributed representation vector and a first weight of the first distributed representation vector from the input document; extract related terms of a keyword extracted from words included in the input document based on a similarity degree with between the first distributed representation vector and a second distributed representation vector of words included in a plurality of pieces of reference text analysis data; obtain a second weight of the related terms, the second weight comprising a product of the first weight by the similarity degree; output the first weight and the second weight; receive a change to value of one or more of the first weight and the second weight; and execute the search of the document by using the first weight and the second weight that were compiled.

2 . The document search device according to claim 1 , wherein the first distributed representation vector and the second distributed representation vector are obtained through a machine learning of distributed representation of a word included in the plurality of pieces of reference text analysis data.

3 . The document search device according to claim 2 , wherein the plurality of pieces of the reference text analysis data is generated by performing a morphological analysis on reference text data.

4 . The document search device according to claim 1 , wherein the first distributed representation vector is obtained from a keyword extracted from the input document.

5 . The document search device according to claim 2 , wherein the machine learning uses a neural network.

6 . A document search device comprising:
a memory portion having instructions stored thereon that, when executed by a processing portion, cause the processing portion to perform operations for searching a document in view of an input document, the operations comprising:
obtaining a first distributed representation vector and a first weight of the first distributed representation vector from the input document,
extracting related terms of a keyword extracted from words included in the input document based on a similarity degree with between the first distributed representation vector and a second distributed representation vector of the words included in a plurality of pieces of reference text analysis data,
obtaining a second weight of the related terms, the second weight comprising a product of the first weight and the similarity degree,
outputting the first weight and the second weight, and
executing a search of the document using a changed value of one or more of the first weight and the second weight, the changed value being provided through compiling of keyword data and related term data.

7 . The document search device according to claim 6 , wherein the first distributed representation vector and the second distributed representation vector are obtained through a machine learning of distributed representation of a word included in the plurality of pieces of reference text analysis data.

8 . The document search device according to claim 7 , wherein the plurality of pieces of the reference text analysis data is generated by performing a morphological analysis on reference text data.

9 . The document search device according to claim 7 , wherein the machine learning uses a neural network.

10 . The document search device according to claim 6 , wherein the first distributed representation vector is obtained from a keyword extracted from the input document.

Assignees

• Semiconductor Energy Lab

Inventors

• Higashi Kazuki
• Momo Junpei

Classifications

• G06Q50/184

  Intellectual property management · CPC title

• G06Q10/10

  Office automation; Time management · CPC title

• G06N3/08

  Learning methods · CPC title

• G06F2216/11

  Patent retrieval · CPC title

• G06F40/268

  Morphological analysis · CPC title