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

Patent metadata

Field	Value
Publication number	US-12019636-B2
Application number	US-202017064871-A
Country	US
Kind code	B2
Filing date	Oct 7, 2020
Priority date	Mar 23, 2018
Publication date	Jun 25, 2024
Grant date	Jun 25, 2024

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,
wherein the processing portion is configured to generate a keyword data comprising a keyword and a weight of the keyword,
wherein the processing portion is configured to generate a related term data of the keyword,
wherein the processing portion is configured to output the related term data that was generated,
wherein the processing portion is configured to obtain a distributed representation vector through a machine learning of distributed representation of a word included in a plurality of pieces of first reference text analysis data,
wherein the processing portion is configured to extract a related word to the keyword among words included in the plurality of pieces of the first reference text analysis data by using the distributed representation vector,
wherein the processing portion is configured to accept compiling the weight of the keyword,
wherein the related term data comprises a weight of a related term,
wherein the processing portion is configured to accept compiling the weight of the related term data that was output, and
wherein the processing portion is configured to execute the search of the document by using the keyword and the related term data that was compiled.

2. A document search method using the document search device according to claim 1 .

3. A document search method for searching data related or similar to a document data, comprising the steps of:
extracting keywords included in a document data;
obtaining a distributed representation vector through a machine learning of distributed representation of a word included in a plurality of pieces of first reference text analysis data;
extracting related words to each of the keywords among words included in the plurality of pieces of the first reference text analysis data by using the distributed representation vector;
accepting a deletion of a word in the keywords and the related words; and
searching the data related or similar to the document data by using the keywords and the related words after accepting the deletion.

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

5. The document search method according to claim 3 , wherein the distributed representation vector of the word is generated by using a neural network.

6. A document search system comprising a server executing the document search method according to claim 3 , and an electronic device having a function of supplying text data to the server and being supplied with a ranking data from the server, through one or both of wire communication and wireless communication.

7. A program making a computer execute the document search method according to claim 3 .

8. A document search method for searching data related or similar to a document data, comprising the steps of:
extracting a keyword data comprising a keyword and a weight of the keyword, the keyword included in a document data;
obtaining a distributed representation vector through a machine learning of distributed representation of a word included in a plurality of pieces of first reference text analysis data;
extracting a related word to the keyword among words included in the plurality of pieces of the first reference text analysis data by using the distributed representation vector;
accepting a change of the weight of the keyword and the related word; and
searching the data related or similar to the document data by using the weight of the keyword and the related word after accepting the change.

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

10. The document search method according to claim 8 , wherein the distributed representation vector of the word is generated by using a neural network.

11. A document search system comprising a server executing the document search method according to claim 8 , and an electronic device having a function of supplying text data to the server and being supplied with a ranking data from the server, through one or both of wire communication and wireless communication.

12. A program making a computer execute the document search method according to claim 8 .

13. A document search method for searching data related or similar to a document data, comprising the steps of:
extracting a keyword included in a document data;
obtaining distributed representation vectors through a machine learning of distributed representation of a word included in a plurality of pieces of first reference text analysis data;
extracting a related word data comprising a related word and a weight of the related word to the keyword among words included in the plurality of pieces of the first reference text analysis data by using the distributed representation vectors of the keyword and the word;
accepting a change of the keyword and the weight of the related word; and
searching the data related or similar to the document data by using the keyword and the related word after accepting the change.

14. The document search method according to claim 13 , wherein the plurality of pieces of the first reference text analysis data is generated by performing a morphological analysis on first reference text data.

15. The document search method according to claim 13 , wherein the distributed representation vector of the word is generated by using a neural network.

16. A document search system comprising a server executing the document search method according to claim 13 , and an electronic device having a function of supplying text data to the server and being supplied with a ranking data from the server, through one or both of wire communication and wireless communication.

17. A program making a computer execute the document search method according to claim 13 .

Assignees

• Semiconductor Energy Lab

Inventors

• Higashi Kazuki
• Momo Junpei

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/08

  Learning methods · CPC title

• G06N3/063

  using electronic means · CPC title

• G06F40/279 Primary

  Recognition of textual entities · CPC title