System, method, and recording medium for regular rule learning

• US 10,120,863 B2
• Filed: 10/24/2017
• Issued: 11/06/2018
• Est. Priority Date: 03/31/2016
• Status: Active Grant

• Current Assignee: AirBNB Incorporated
• Original Assignee: International Business Machines Corporation
• Inventors: Popescu, Octavian, Sheinin, Vadim
• Primary Examiner(s): Baker, Charlotte M
• Application Number:
  US15/791,791
• Publication Number:
  US 20180046615A1
• Time in Patent Office: 378 Days
• Field of Search: 704 9, 704277, 704 5, 704E11001, 704E21001, 704E1504, 706 16
• CPC Class Codes: G06F 40/211 Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

A regular rule learning system, including an analyzing circuit configured to analyze a corpus of sentences to find semantic relationships between sentence constituents that are responsible for specific senses of words in that sentence by describing the semantic relationships and grammatical relations that are actuated in the sentence.

Claims

1. A regular rule learning system, comprising:

a processer; and

a memory, the memory storing instructions to cause the processor to execute;

an analyzing circuit configured to analyze a corpus of sentences to find semantic relationships between sentence constituents that are responsible for specific senses of words in that sentence by describing the semantic relationships and grammatical relations that are actuated in the sentence to create a set of regular rules, a regular rule does not depend on a history or applying any other rules in the set of regular rules.

2. The system of claim 1, wherein the analyzing further analyzes the corpus of sentences stored in a database to discover lexical features of the sentences and conjunctively create the set of regular rules based on the discovered lexical features and syntactical features of the sentences.

3. The system of claim 2, wherein the syntactical features of the corpus of sentences are given.

4. The system of claim 2, wherein the lexical features are unknown prior to the analyzing circuit discovering the lexical features.

5. The system of claim 2, wherein the lexical features are calculated from the corpus of sentences stored in the database prior to the syntactical features being given to the corpus of sentences.

14. The system of claim 1, wherein the set of regular rules predicts a meaning of a user input in the corpus of sentences without referring to a database and independently of any other rule.

6. A regular rule learning method, comprising:

analyzing a corpus of sentences to find semantic relationships between sentence constituents that are responsible for specific senses of words in that sentence by describing the semantic relationships and grammatical relations that are actuated in the sentence, wherein the syntactical features of the corpus of sentences are given to create a set of regular rules, a regular rule does not depend on a history or applying any other rules in the set of regular rules.

7. The method of claim 6, wherein the analyzing further analyzes the corpus of sentences stored in a database to discover lexical features of the sentences and conjunctively create the set of regular rules based on the discovered lexical features and syntactical features of the sentences.

8. The method of claim 7, wherein the lexical features are unknown prior to the analyzing discovering the lexical features.

9. The method of claim 7, wherein the lexical features are calculated from the corpus of sentences stored in the database prior to the syntactical features being given to the corpus of sentences.

15. The method of claim 6, wherein the set of regular rules predicts a meaning of a user input in the corpus of sentences without referring to a database and independently of any other rule.

10. A non-transitory computer-readable recording medium recording a regular rule learning program, the program causing a computer to perform:

analyzing a corpus of sentences to find semantic relationships between sentence constituents that are responsible for specific senses of words in that sentence by describing the semantic relationships and grammatical relations that are actuated in the sentence, wherein the syntactical features of the corpus of sentences are given to create a set of regular rules, a regular rule does not depend on a history or applying any other rules in the set of regular rules.

11. The non-transitory computer-readable recording medium of claim 10, wherein the analyzing further analyzes the corpus of sentences stored in a database to discover lexical features of the sentences and conjunctively create the set of regular rules based on the discovered lexical features and syntactical features of the sentences.

12. The non-transitory computer-readable recording medium of claim 10, wherein the lexical features are unknown prior to the analyzing discovering the lexical features.

13. The non-transitory computer-readable recording medium of claim 10, wherein the lexical features are calculated from the corpus of sentences stored in the database prior to the syntactical features being given to the corpus of sentences.

16. The non-transitory computer-readable recording medium of claim 10, wherein the set of regular rules predicts a meaning of a user input in the corpus of sentences without referring to a database and independently of any other rule.