RESPONDING TO AN INDIRECT UTTERANCE BY A CONVERSATIONAL SYSTEM

• US 20200073941A1
• Filed: 11/07/2019
• Published: 03/05/2020
• Est. Priority Date: 02/01/2018
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Canim, Mustafa, Farrell, Robert G., Fokoue-Nkoutche, Achille B., Gunnels, John A., Musa, Ryan A., Saraswat, Vijay A.
• Granted Patent: US 10,832,006 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

A method, apparatus and computer program product for responding to an indirect utterance in a dialogue between a user and a conversational system is described. An indirect utterance is received. A parse structure of the indirect utterance is generated. The indirect utterance is an utterance which does not match a user goal expressed as elements of a knowledge graph. The parse structure is connected through the knowledge graph to a user goal to issue a user request which is not stated in the indirect utterance. The parse structure is connected using a matching process which matches the parse structure with the connected user goal in the knowledge graph according to a similarity of the parse structure and a portion of the knowledge graph including the connected user goal. A system response is performed based on the connected user goal.

Claims

1. A method for responding to an indirect utterance in a dialogue between a user and a conversational system comprising:

responsive to receiving an indirect utterance, generating a parse structure of the indirect utterance, wherein the indirect utterance is an utterance which does not match a user goal expressed as elements of a knowledge graph, wherein the knowledge graph comprises user goals expressed as elements;

connecting the parse structure through the knowledge graph to a user goal to issue a user request which is not stated in the indirect utterance, wherein the connecting includes a matching process which matches the parse structure with the connected user goal in the knowledge graph according to a similarity of the parse structure and a portion of the knowledge graph including the connected user goal;

performing a system response, wherein the system response is a system response based on the connected user goal.

2. The method as recited in claim 1, wherein the matching process is selected from the group of a “

term to predicate”

similarity matching process, a linguistic matching process, a semantic matching process, a syntactic matching process, a probabilistic matching process and a structural similarity matching process.

3. The method as recited in claim 1, wherein the parse structure is one of a logical form, a constituency parse, phrase structure parse, dependency parse, universal dependency parse and an enhanced dependency parse.

4. The method as recited in claim 1, wherein the indirect utterance is one of a short text, an ungrammatical text, a chat text, a text message, a sentence in a natural language and transcribed speech in a dialogue exchange.

5. The method as recited in claim 2, wherein a term to predicate similarity measure is used in the matching process.

6. The method as recited in claim 2, wherein the matching process is selected from a group consisting of a linguistic matching process, a semantic matching process, a structural matching process, a syntactic matching process and a probabilistic matching process.

7. The method as recited in claim 1, wherein the parsing structure is a logical form produced by a semantic parser which depends upon weights and at least one weight is initialized differently due to making a logical connection to a dialogue goal.

8. The method as recited in claim 7, wherein the logical form includes at least one term or predicate mapped to a web resource.

9. The method as recited in claim 1, wherein the system response is based on a connected goal and the system response includes the goal.

10. The method as recited in claim 9, wherein the connected goal is not communicated as a natural language response immediately but is instead communicated to the user in a later dialogue exchange.

18. The method as recited in claim 1, wherein the user goal is selected based on a ranking process.

19. The method as recited in claim 1, wherein the ranking process uses feedback from users on previous system responses that included a connected goal.

11. Apparatus, comprising:

a processor;

computer memory holding computer program instructions executed by the processor for responding to an indirect utterance in a dialogue between a user and a conversational system, the computer program instructions comprising;

program code responsive to receiving an indirect utterance, operative to generate a parse structure of the indirect utterance, wherein the indirect utterance is an utterance which does not match a user goal expressed as element of a knowledge graph;

program code, operative to connect connecting the parse structure through the knowledge graph to a user goal to issue a user request which is not stated in the indirect utterance, wherein the connecting includes a matching process which matches the parse structure with the connected user goal in the knowledge graph according to a similarity of the parse structure and a portion of the knowledge graph including the connected user goal;

program code, operative to perform a system response, wherein the system response is a system response based on the connected user goal.

12. The apparatus as recited in claim 11, wherein the matching process is selected from the group of a “

term to predicate”

similarity matching process, a linguistic matching process, a semantic matching process, a syntactic matching process, a probabilistic matching process and a structural similarity matching process.

13. The apparatus as recited in claim 11, wherein the parse structure is one of a logical form, a constituency parse, phrase structure parse, dependency parse, universal dependency parse and an enhanced dependency parse

14. The apparatus as recited in claim 11, wherein at least one part of the parse is connected to at least one part of the user goal through subsumption, specialization, or abstraction relationships in the knowledge graph.

15. The apparatus as recited in claim 11, wherein the system response performs a speech act, wherein the speech act can include confirming, explaining, clarifying, or suggesting the goal.

17. The computer program product as recited in claim 15, wherein the matching process is selected from the group of a “

term to predicate”

similarity matching process, a linguistic matching process, a semantic matching process, a syntactic matching process, a probabilistic matching process and a structural similarity matching process.

16. A computer program product in a non-transitory computer readable medium for use in a data processing system, the computer program product holding computer program instructions executed by the data processing system for responding to an indirect utterance in a dialogue between a user and a conversational system, the computer program instructions comprising:

program code responsive to receiving an indirect utterance, operative to generate a parse structure of the indirect utterance, wherein the indirect utterance is an utterance which does not match a user goal expressed as elements of a knowledge graph;

program code, operative to connect the parse structure through the knowledge graph to a user goal to issue a user request which is not stated in the indirect utterance, wherein the connecting includes a matching process which matches the parse structure with the connected user goal in the knowledge graph according to a similarity of the parse structure and a portion of the knowledge graph including the connected user goal;

program code, operative to perform a system response, wherein the system response is a system response based on the connected user goal.