Identifying user expectations in question answering

Patent metadata

Field	Value
Publication number	US-10380150-B2
Application number	US-201715848929-A
Country	US
Kind code	B2
Filing date	Dec 20, 2017
Priority date	Dec 20, 2017
Publication date	Aug 13, 2019
Grant date	Aug 13, 2019

Abstract

Method and system for identifying user expectations in question answering in an on-line social network system are described. The automated support system is configured to address the technical problem of optimization of the processing of user input submitted to a computer in the form of a natural language. The automated support system uses machine learning algorithms to automatically extract, from the user input, information indicative of the user's expectations and obtain data relevant to the input based on said information indicative of the user's expectations.

First claim

The invention claimed is:

1. A computer-implemented method comprising:
generating training data by labeling previously collected data samples comprising submitted questions and associated answers;
training a supervised machine learning model using the training data;
using at least one processor, detecting an input from a user in a form of a natural language;
executing the trained supervised machine learning model using the input from the user as input to produce a topic indicative of the user's expectation;
accessing, in a support database, an answer corresponding to the topic; and
communicating the answer to a device associated with the user.

2. The method of claim 1 , wherein the expectation extractor model is a binary classification model trained using previously collected and labeled training data.

3. The method of claim 1 , comprising using information derived from the input and the answer as training data to train the expectation extractor model.

4. The method of claim 3 , comprising:
identifying the input as associated with a topic;
based on information stored in a member profile that represents the user in an on-line social network system, calculating a weight, the weight indicative of a level of expertise of the user with respect to the topic;
storing the input as associated with the topic and with the weight.

5. The method of claim 4 , wherein the using of the information derived from the input and the answer as training data to train the expectation extractor model comprises using the weight indicative of a level of expertise of the user with respect to the topic.

6. The method of claim 1 , comprising:
executing the expectation extractor model to extract, from the input, a further phrase indicative of a context associated with the user's expectation; and
using the context in accessing; in a support database, of the answer corresponding to the user's expectation.

7. The method of claim 1 ; comprising monitoring interaction of the user with the answer communicated to the device associated with the user, wherein the using of the information derived from the input and the answer as training data to train the expectation extractor model comprises using the monitored interaction.

8. The method of claim 1 , wherein the communicating of the answer to the device associated with the user comprises communication of a request for further information associated with the input.

9. The method of claim 1 ; wherein the input is in a form of text.

10. The method of claim 1 , wherein the input s voice input.

11. A system comprising:
one or more processors; and
a non-transitory computer readable storage medium comprising instructions that when executed by the one or processors cause the one or more processors to perform operations comprising:
generating training data b labeling previously collected data samples comprising submitted questions and associated answers;
training a supervised machine learning model using the training data;
detecting an input from a user in a form of a natural language;
executing the trained supervised machine learning model using the input from the user as input to produce a topic indicative of the user's expectation;
accessing, in a support database, an answer corresponding to the topic; and
communicating the answer to a device associated with the user.

12. The system of claim 11 , wherein the expectation extractor model is a binary classification model trained using previously collected and labeled training data.

13. The system of claim 11 , comprising using information derived from the input and the answer as training data to train the expectation extractor model.

14. The system of claim 13 , comprising:
identifying the input as associated with a topic;
based on information stored in a member profile that represents the user in an on-line social network system, calculating a weight, the weight indicative of a level of expertise of the user with respect to the topic;
storing the input as associated with the topic and with the weight.

15. The system of claim 14 , wherein the using of the information derived from the input and the answer as training data to train the expectation extractor model comprises using the weight indicative of a level of expertise of the user with respect to the topic.

16. The system of claim 11 , comprising:
executing the expectation extractor model to extract, from the input, a further phrase indicative of a context associated with the user's expectation; and
using the context in accessing, in a support database, of the answer corresponding to the user's expectation.

17. The system of claim 11 , comprising monitoring interaction of the user with the answer communicated to the device associated with the user, wherein the using of the information derived from the input and the answer as training data to train the expectation extractor model comprises using the monitored interaction.

18. The system of claim 11 , wherein the communicating of the answer to the device associated with the user comprises communication of a request for further information associated with the input.

19. The system of claim 11 , wherein the input is in a form of text or is voice input.

20. A machine-readable non-transitory storage medium having instruction data executable by a machine to cause the machine to perform operations comprising:
generating training data by labeling previously collected data samples comprising submitted questions and associated answers;
training a supervised machine learning model using the training data;
detecting an input from a user in a form of a natural language;
executing the trained supervised machine learning model using the input from the user as input to produce a topic indicative of the user's expectation;
accessing, in a support database, an answer corresponding to the topic; and
communicating the answer to a device associated with the user.

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Huang Shen
• Zhang Yongzheng
• Kuan Chi-Yi
• Wang Hu
• Zhao Rui
• Jin Zhou

Classifications

• G06Q10/40

  Business processes related to social networking or social networking services · CPC title

• G06F18/24

  Classification techniques · CPC title

• G06F40/30

  Semantic analysis · CPC title

• G06F40/216

  using statistical methods · CPC title

• G06F16/3329 Primary

  Natural language query formulation · CPC title