Method, Device and Equipment for Acquiring Answer Information

• US 20160328469A1
• Filed: 05/03/2016
• Published: 11/10/2016
• Est. Priority Date: 05/04/2015
• Status: Active Grant

• Current Assignee: Shanghai Xiaoi Robot Technology Co., Ltd.
• Original Assignee: Shanghai Xiaoi Robot Technology Co., Ltd.
• Inventors: ZHU, Pinpin
• Granted Patent: US 10,489,435 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

Method, device and equipment for acquiring answer information are provided. The method includes: acquiring a text sequence to be processed from a user equipment; dividing the text sequence into a plurality of word segments and obtaining semantic information of at least one of the plurality of word segments; performing a search in at least one of the plurality of sub-databases for matching the semantic information, successively from high to low in priority of the sub-databases, till an initial answer information matching the semantic information is found; and providing an established answer information corresponding to the initial answer information to the user equipment. Based on better semantic understanding to the text sequence, more accurate answer information may be provided.

Claims

1. A method for acquiring answer information based on a predetermined knowledge database in a computer equipment, wherein the predetermined knowledge database comprises a plurality of sub-databases, and the method comprises:

a) acquiring a text sequence to be processed from a user equipment;

b) dividing the text sequence into a plurality of word segments and obtaining semantic information of at least one of the plurality of word segments;

c) performing a search in at least one of the plurality of sub-databases for matching the semantic information, successively from high to low in priority of the plurality of sub-databases, till an initial answer information matching the semantic information is found; and

d) providing an established answer information corresponding to the initial answer information to the user equipment.

2. The method according to claim 1, prior to the step c), further comprising:

determining the priority order of the plurality of sub-databases based on the semantic information.

3. The method according to claim 1, wherein the step c) comprises:

if answer information, which has a higher matching rate with the semantic information of the word segment than a first threshold, is found in the search, determining the answer information to be the initial answer information;

or

if one or more answer information, which has a higher matching rate with the semantic information of the word segment than a second threshold and lower than the first threshold, is found when the search is completed, providing description information of the one or more answer information to the user equipment; and

determining answer information which is fed back from the user equipment and corresponds to a portion of the description information which is selected by the user equipment to be the initial answer information.

4. The method according to claim 1, wherein the step c) comprises:

c1) performing the search to multiple groups of semantic matching information in at least one of the plurality of sub-databases for matching the semantic information, successively from high to low in priority of the plurality of sub bases, till a group of semantic matching information matching the semantic information is found; and

determining answer information corresponding to the group of semantic matching information to be the initial answer information, wherein each of the multiple groups of semantic matching information comprises one or more semantic matching structure.

5. The method according to claim 4, wherein the semantic matching structure comprises at least one of the following items:

a plurality of words and semantic information of at least one of the plurality of words; and

logic relations between the plurality of words.

6. The method according to claim 4, wherein the step c1) comprises:

if the group of semantic matching information matching the semantic information corresponds to a plurality of answer information, selecting one answer information from the plurality of answer information to be the initial answer information based on at least one of equipment information of the user equipment and application information of an application which sends the text sequence to be processed.

7. The method according to claim 4, wherein the sub-databases comprises at least one ontology instance, each of the at least one ontology instance has at least one property, each of the at least one property comprises one group of semantic matching information and an initial answer information corresponding to the group of semantic matching information, and the method further comprises:

x) obtaining an instance item of an ontology, the ontology having at least one property, and each of the at least one property of the ontology comprising an ontology answer information and one group of semantic matching information with a default item; and

y) processing the default item in the at least one property of the ontology to obtain the ontology instance which inherits all the at least one property of the ontology based on the instance item of the ontology.

8. The method according to claim 1, wherein the step d) comprises:

if it is determined that the initial answer information cannot be directly provided to the user equipment, obtaining parameter information from the semantic information of the word segment;

processing the initial answer information based on the parameter information to obtain the established answer information; and

providing the established answer information to the user equipment.

9. The method according to claim 1, further comprising:

if the text sequence to be processed meets a predetermined trigger condition, obtaining historical search information of the user equipment,

wherein the step c) comprises;

by considering the historical search information of the user equipment, performing the search in at least one of the plurality of sub-databases for matching the semantic information, successively from high to low in priority of the sub-databases, till the initial answer information matching the semantic information is found.

10. The method according to claim 1, wherein the semantic information of the word segment comprises at least one of the followings:

a synonym of the word segment, and/or a combination of synonyms of the word segment, a word that belongs to a same class with the word segment, and/or a combination of words that belong to the same class with the word segment, and an entity having a same or similar structure with the word segment.

11. A question-answering device for acquiring answer information based on a predetermined knowledge database in a computer equipment, wherein the predetermined knowledge database comprises a plurality of sub-databases, and the question-answering device comprises:

a first acquiring device, configured to acquire a text sequence to be processed from a user equipment;

a semantic device, configured to;

divide the text sequence into a plurality of word segments; and

obtain semantic information of at least one of the plurality of word segments;

a retrieval device, configured to;

perform a search in at least one of the plurality of sub-databases for matching the semantic information, successively from high to low in priority of the sub-databases, till an initial answer information matching the semantic information is found; and

a first providing device, configured to provide an established answer information corresponding to the initial answer information to the user equipment.

12. The question-answering device according to claim 11, further comprising:

a first determining device, configured to determine the priority order of the plurality of sub-databases based on the semantic information.

13. The question-answering device according to claim 11, wherein the retrieval device comprises:

a second determining device, configured to;

if answer information, which has a higher matching rate with the semantic information of the word segment than a first threshold, is found in the search, determine the answer information to be the initial answer information;

or

a second providing device, configured to;

if one or more answer information, which has a higher matching rate with the semantic information of the word segment than a second threshold and lower than the first threshold, is found when the search is completed, provide description information of the one or more answer information to the user equipment; and

a third determining device, configured to;

determine answer information which is fed back from the user equipment and corresponds to a portion of the description information which is selected by the user equipment to be the initial answer information.

14. The question-answering device according to claim 11, wherein the retrieval device comprises:

a first sub-retrieval device, configured to;

perform the search to multiple groups of semantic matching information in at least one of the plurality of sub-databases for matching the semantic information, successively from high to low in priority of the sub-databases, till a group of semantic matching information matching the semantic information is found; and

determine answer information corresponding to the group of semantic matching information to be the initial answer information, wherein each of the multiple group of semantic matching information comprises one or more semantic matching structure.

15. The question-answering device according to claim 14, wherein the semantic matching structure comprises at least one of the following items:

a plurality of words and semantic information of at least one of the plurality of words; and

logic relations between the plurality of words.

16. The question-answering device according to claim 14, wherein the first sub-retrieval device comprises:

a second sub-retrieval device, configured to;

perform the search to multiple groups of semantic matching information in at least one of the plurality of sub-databases for matching the semantic information, successively from high to low in priority of the sub-databases, till a group of semantic matching information matching the semantic information is found; and

a selecting device, configured to;

if the group of semantic matching information matching the semantic information corresponds to a plurality of answer information, select one answer information from the plurality of answer information to be the initial answer information based on at least one of equipment information of the user equipment and application information of an application which sends the text sequence to be processed.

17. The question-answering device according to claim 14, wherein the sub-databases comprises at least one ontology instance, each of the at least one ontology instance has at least one property, each of the at least one property comprises one group of semantic matching information and an initial answer information corresponding to the group of semantic matching information, and the question-answering device further comprises:

a second acquiring device, configured to obtain an instance item of an ontology, the ontology having at least one property, and each of the at least one property of the ontology comprising an ontology answer information and one group of semantic matching information with a default item; and

an instance generating device, configured to process the default item in the at least one property of the ontology to obtain the ontology instance which inherits all the at least one property of the ontology based on the instance item of the ontology.

18. The question-answering device according to claim 11, wherein the first providing device comprises:

a third acquiring device, configured to;

if it is determined that the initial answer information cannot be directly provided to the user equipment, obtain parameter information from the semantic information of the word segment;

a fourth acquiring device, configured to process the initial answer information based on the parameter information to obtain the established answer information; and

a third providing device, configured to provide the established answer information to the user equipment.

19. The question-answering device according to claim 11, further comprising:

a fifth acquiring device, configured to;

if the text sequence to be processed meets a predetermined trigger condition, obtain historical search information of the user equipment,

wherein the retrieval device comprises a third sub-retrieval device, configured to;

by considering the historical search information of the user equipment, perform the search in at least one of the plurality of sub-databases for matching the semantic information, successively from high to low in priority of the sub-databases, till an initial answer information matching the semantic information is found.

20. The question-answering device according to claim 11, wherein the semantic information of the word segment comprises at least one of the followings:

a synonym of the word segment, and/or a combination of synonyms of the word segment, a word that belongs to a same class with the word segment, and/or a combination of words that belong to the same class with the word segment, and an entity having a same or similar structure with the word segment.

21. A computer equipment, comprising the question-answering device according to claim 11.