Learning apparatus and learning method

• US 10,643,152 B2
• Filed: 03/22/2018
• Issued: 05/05/2020
• Est. Priority Date: 03/30/2017
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Mizobuchi, Yuji
• Primary Examiner(s): Gauthier, Gerald
• Application Number:
  US15/928,431
• Publication Number:
  US 20180285781A1
• Time in Patent Office: 775 Days
• Field of Search: 382159, 382173, 704 2, 704 3, 704 9, 704245, 704 8, 704 1, 704236, 704244, 706 12, 706 20, 707737, 707748, 707749, 707769, 707738, 434362, 715753
• CPC Class Codes: G06F 16/353 into predefined classes
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

A learning apparatus includes a memory and a processor configured to acquire a plurality of documents, perform clustering of the plurality of documents for each of a plurality of words included in the plurality of document, when a plurality of clusters are generated for a first word among the plurality of words by the clustering, perform assignment of different labels corresponding to the plurality of clusters to the first word included in the plurality of documents, and perform re-clustering of the plurality of documents including the first word with the assigned different labels, for other words among the plurality of words.

Claims

1. A learning apparatus comprising:

a memory; and

a processor coupled to the memory and the processor configured to;

acquire a plurality of words from a plurality of documents;

generate a plurality of contexts represented in a vector for each word of the plurality of words;

perform clustering of the plurality of contexts for each word of the plurality of words;

when a plurality of clusters are generated for a first word among the plurality of words by the clustering, perform assignment, to the first word, different labels corresponding to each cluster of the plurality of clusters;

generate first contexts for each first word distinguished by the assigned different labels; and

perform re-clustering of the plurality of contexts including the first contexts for each first word with the assigned different labels.

2. The learning apparatus according to claim 1, wherein the first word with the assigned different labels is treated as different words.

3. The learning apparatus according to claim 1, the processor further configured to:

extract the plurality of words from the plurality of documents to generate a word dictionary,

wherein each context of the plurality of contexts is to indicate whether each document of the plurality of documents includes each word of the plurality of words included in the word dictionary.

4. The learning apparatus according to claim 3, wherein the assignment includes changing the first word included in the word dictionary to a plurality of words with the assigned different labels.

5. The learning apparatus according to claim 1, the processor further configured to:

determine whether a plurality of second clusters for a second word for other words among the plurality of words are generated as a result of the re-clustering;

when the plurality of second clusters are generated for the second word, assign, to the second word, different labels corresponding to each cluster of the plurality of second clusters; and

perform other re-clustering of the plurality of contexts including the contexts for each second word with the assigned different labels.

6. The learning apparatus according to claim 5, the processor further configured to:

when the plurality of second clusters are not generated for the second word, output information indicating a result of the re-clustering.

7. A learning method executed by a computer, the method comprising:

acquiring a plurality of words from a plurality of documents;

generating a plurality of contexts represented in a vector for each word of the plurality of words;

performing clustering of the plurality of contexts for each word of the plurality of words;

when a plurality of clusters are generated for a first word among the plurality of words by the clustering, performing assignment, to the first word, different labels corresponding to each cluster of the plurality of clusters;

generating first contexts for each first word distinguished by the assigned different labels; and

performing re-clustering of the plurality of contexts including the first contexts for each first word with the assigned different labels.

8. The learning method according to claim 7, wherein the first word with the assigned different labels is treated as different words.

9. The learning method according to claim 7, the method further comprising:

extracting the plurality of words from the plurality of documents to generate a word dictionary,

wherein each context of the plurality of contexts is to indicate whether each of the plurality of documents includes each of the plurality of words included in the word dictionary.

10. The learning method according to claim 9, wherein the assignment includes changing the first word included in the word dictionary to a plurality of words with the assigned different labels.

11. The learning method according to claim 7 further comprising:

determining whether a plurality of second clusters for a second word for other words among the plurality of words are generated as a result of the re-clustering;

when the plurality of second clusters are generated for the second word, assigning, to the second word, different labels corresponding to each cluster of the plurality of second clusters; and

performing other re-clustering of the plurality of contexts including the contexts for each second word with the assigned different labels.

12. The learning method according to claim 11 further comprising:

when the plurality of second clusters are not generated for the second word, outputting information indicating a result of the re-clustering.

13. A non-transitory computer-readable medium storing a learning program that causes a computer to execute a process comprising:

acquiring a plurality of words from a plurality of documents;

generate a plurality of contexts represented in a vector for each word of the plurality of words;

performing clustering of the plurality of contexts for each word of the plurality of words;

when a plurality of clusters are generated for a first word among the plurality of words by the clustering, performing assignment, to the first word, different labels corresponding to each cluster of the plurality of clusters;

generate first contexts for each first word distinguished by the assigned different labels; and

performing re-clustering of the plurality of contexts including the first contexts for each first word with the assigned different labels.