SEMANTIC DISAMBIGUATION USING A STATISTICAL ANALYSIS

• US 20150178268A1
• Filed: 10/08/2014
• Published: 06/25/2015
• Est. Priority Date: 12/19/2013
• Status: Active Grant

• Current Assignee: ABBYY Development LLC
• Original Assignee: ABBYY InfoPoisk LLC
• Inventors: Bogdanova, Daria Nikolaevna, Zuev, Konstantin Alekseevich
• Granted Patent: US 9,740,682 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/2457 with adaptation to user needs
• CPC Class Codes: G06F 16/24573 using data annotations, e.g. user-defined metadata
• CPC Class Codes: G06F 16/284 Relational databases
• CPC Class Codes: G06F 16/288 Entity relationship models
• CPC Class Codes: G06F 16/36 Creation of semantic tools, e.g. ontology or thesauri
• CPC Class Codes: G06F 16/367 Ontology
• CPC Class Codes: G06F 16/374 Thesaurus
• CPC Class Codes: G06F 40/247 Thesauruses; Synonyms
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/45 Example-based machine translation; Alignment
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning

Abstract

A text containing a word is received by a computing device. The word is compared to inventory words in a sense inventory. The sense inventory comprises at least one inventory word and at least one concept corresponding to the at least one inventory word. Upon matching the word to an inventory word in the sense inventory, a concept for the word is identified by comparing each concept related to the inventory word to the word. The concept is assigned the word.

Claims

1. A method comprising:

receiving text by a computing device, the text including a word;

comparing, by a processor of the computing device, the word in the text to inventory words in a sense inventory, wherein the sense inventory comprises at least one inventory word and at least one concept corresponding to the at least one inventory word;

responsive to matching the word to an inventory word in the sense inventory, identifying a concept for the word by comparing each concept related to the inventory word to the word;

responsive to identifying the concept that is correct for the word, assigning the concept to the word; and

responsive to not identifying the concept that is correct for the word, adding a new concept to the sense inventory for the inventory word.

2. The method of claim 1, further comprising performing a statistical analysis to identify the concept for the word.

3. The method of claim 1, further comprising, responsive to not matching the word to an inventory word in the sense inventory, adding a new inventory word and a corresponding new concept to the sense inventory based on the word.

4. The method of claim 1, wherein the sense inventory further comprises a semantic hierarchy of parent-child relationships of semantic classes, and wherein each semantic class comprises a deep model that determines a semantic relationship between a parent of the semantic class and a child of the semantic class.

5. The method of claim 4, wherein a child semantic class inherits and adjusts a deep model of a direct parent semantic class.

6. The method of claim 1, wherein adding the new concept to the sense inventory for the inventory word comprises statistically analyzing each level of a semantic hierarchy associated with the sense inventory starting from a root of the hierarchy to determine a probability of the concept for the word, and wherein the new concept comprises a most probable concept.

7. The method of claim 6, wherein the probability of the concept for word is based on a corpora corresponding to the text.

8. A system comprising:

one or more data processors; and

one or more storage devices storing instructions that, when executed by the one or more data processors, cause the one or more data processors to perform operations comprising;

receiving text by a computing device, the text including a word;

comparing, by a processor of the computing device, the word in the text to inventory words in a sense inventory, wherein the sense inventory comprises at least one inventory word and at least one concept corresponding to the at least one inventory word;

responsive to matching the word to an inventory word in the sense inventory, identifying a concept for the word by comparing each concept related to the inventory word to the word;

responsive to identifying the concept that is correct for the word, assigning the concept to the word; and

responsive to not identifying the concept that is correct for the word, adding a new concept to the sense inventory for the inventory word.

9. The system of claim 8, wherein the operations further comprising performing a statistical analysis to identify the concept for the word.

10. The system of claim 8, wherein the operations further comprising, responsive to not matching the word to an inventory word in the sense inventory, adding a new inventory word and a corresponding new concept to the sense system based on the word.

11. The system of claim 8, wherein the sense inventory further comprises a semantic hierarchy of parent-child relationships of semantic classes, and wherein each semantic class comprises a deep model that determines a semantic relationship between a parent of the semantic class and a child of the semantic class.

12. The system of claim 11, wherein a child semantic class inherits and adjusts a deep model of a direct parent semantic class.

13. The system of claim 8, wherein adding the new concept to the sense inventory for the inventory word comprises statistically analyzing each level of a semantic hierarchy associated with the sense inventory starting from a root of the hierarchy to determine a probability of the concept for the word, and wherein the new concept comprises a most probable concept.

14. The system of claim 13, wherein the probability of the concept for word is based on a corpora corresponding to the text.

15. A computer-readable storage medium having machine instructions stored therein, the instructions being executable by a processor to cause the processor to perform operations comprising:

receiving text by a computing device, the text including a word;

comparing, by a processor of the computing device, the word in the text to inventory words in a sense inventory, wherein the sense inventory comprises at least one inventory word and at least one concept corresponding to the at least one inventory word;

responsive to matching the word to an inventory word in the sense inventory, identifying a concept for the word by comparing each concept related to the inventory word to the word;

responsive to identifying the concept that is correct for the word, assigning the concept to the word; and

responsive to not identifying the concept that is correct for the word, adding a new concept to the sense inventory for the inventory word.

16. The computer-readable storage medium of claim 15, further comprising performing a statistical analysis to identify the concept for the word.

17. The computer-readable storage medium of claim 15, further comprising, responsive to not matching the word to an inventory word in the sense inventory, adding a new inventory word and a corresponding new concept to the sense inventory based on the word.

18. The computer-readable storage medium of claim 15, wherein the sense inventory further comprises a semantic hierarchy of parent-child relationships of semantic classes, and wherein each semantic class comprises a deep model that determines a semantic relationship between a parent of the semantic class and a child of the semantic class.

19. The computer-readable storage medium of claim 18, wherein a child semantic class inherits and adjusts a deep model of a direct parent semantic class.

20. The computer-readable storage medium of claim 15, wherein adding the new concept to the sense inventory for the inventory word comprises statistically analyzing each level of a semantic hierarchy associated with the sense inventory starting from a root of the hierarchy to determine a probability of the concept for the word, and wherein the new concept comprises a most probable concept.

21. The computer-readable storage medium of claim 20, wherein the probability of the concept for word is based on a corpora corresponding to the text.