Using voice information to influence importance of search result categories

Patent metadata

Field	Value
Publication number	US-9875740-B1
Application number	US-201615187251-A
Country	US
Kind code	B1
Filing date	Jun 20, 2016
Priority date	Jun 20, 2016
Publication date	Jan 23, 2018
Grant date	Jan 23, 2018

Abstract

Approaches provide for using voice information to influence the importance of search result categories for a search query. For example, various embodiments may provide search results for a search query based on a most relevant search result category to the search query. Voice information associated with a subsequent user interaction may be analyzed to identify whether the search result category is correct or if search results from a different category should be provided. Additionally, the voice information may be used to update the relevance score of the search result category to the search query to improve the category matching of future queries.

First claim

What is claimed is:

1. A computing system, comprising:
a device processor;
a memory device including instructions that, when executed by the device processor, cause the computing system to:
receive first audio input data from a voice-enabled communications device, the first audio input data corresponding to a first utterance of a user received by a microphone of the voice-enabled communications device;
identify a first volume level of the first audio input data;
identify a first search query from the first audio input data by performing automated speech recognition on the first audio input data;
identify a set of categories associated with the first search query, the set of categories including two or more categories, each of the two or more categories being ranked according to respective relevance scores to the first search query;
identify a first set of results associated with a first category of the set of categories, the first category having a largest relevance score to the first search query;
cause the first set of results to be transmitted to the voice-enabled communications device, the first set of results being presented to the user of the voice-enabled communications device;
receive second audio input data from the voice-enabled communications device, the second audio input data corresponding to a second utterance of the user received by the microphone of the voice-enabled communications device;
identify a second volume level of the second audio input data;
identify a second search query from the second audio input data by performing automated speech recognition on the second audio input data;
determine that the first category is incorrect by:
calculating a voice volume difference by comparing the first volume level and the second volume level; and
determining whether the voice volume difference is above a volume difference threshold, the volume difference threshold indicating whether the volume level of the second utterance is at least a predetermined amount louder than the volume level of the first utterance;

decrease the relevance score of the first category for the first search query when the voice volume difference is determined to be above the volume difference threshold;
identify a second category of the set of categories associated with the first search query; and
increase the relevance score of the second category for the first search query when the voice volume difference is determined to be below the volume difference threshold.

2. The computing system of claim 1 , wherein the instructions, when executed further enable the computing system to:
determine that the second category has a second largest relevance score to the first search query;
identify a second set of results associated with the second category; and
cause the second set of results to be transmitted to the voice-enabled communications device, the second set of results being presented to the user of the voice-enabled communications device.

3. The computing system of claim 1 , wherein determining that the first category is incorrect further comprises:
determining the voice volume difference is above the volume difference threshold; or
calculating a search query similarity score by comparing the first search query and the second search query; and
determining the search query similarity score is above a similarity score threshold indicating that the second search query is a repeat search query of the first search query.

4. The computing system of claim 2 , wherein the instructions, when executed further enable the computing system to:
receive third audio input data from the voice-enabled communications device, the third audio input data corresponding to a third utterance of the user received by the microphone of the voice-enabled communications device;
identify a third volume level of the third audio input data;
identify a third search query from the third audio input data by performing automated speech recognition on the third audio input data;
determine that the second category is correct by:
calculating a second voice volume difference by comparing the first volume level and the third volume level; and
determining the second voice volume difference is below the volume difference threshold indicating that the volume level of the third utterance is not the predetermined amount louder than the volume level of the first utterance; and
calculating a second search query similarity score by comparing the first search query and the third search query; and
determining the second search query similarity score is below the similarity score threshold indicating that the third search query is different than the first search query; and

increase the relevance score of the second category for the first search query.

5. A computing system, comprising:
a computing device processor;
a memory device including instructions that, when executed by the computing device processor, enable the computing system to:
receive a first search request, the first search request associated with first audio input data from a device;
identify a first search query from the first search request;
identify a set of categories associated with the first search query, the set of categories including two or more categories, each of the two or more categories being ranked according to respective relevance scores to the first search query;
identify a first set of results associated with a first category, the first category having a highest relevance score to the first search query;
cause the first set of results to be transmitted to the device, the first set of results being presented to the user of the device;
receive a second search request, the second search request associated with second audio input data from the device;
analyze the second search request for indicators that an incorrect category was presented;
determine the second search request includes one or more of the indicators that the incorrect category was presented;
decrease the relevance score of the first category for the first search query when a voice volume difference is determined to be above a volume difference threshold;
identify a second category of the set of categories associated with the first search query; and
increase the relevance score of the second category for the first search query when the voice volume difference is determined to be below the volume difference threshold.

6. The computing system of claim 5 , wherein the instructions, when executed further enable the computing system to:
identify a second category from the set of categories, the second category having a second highest relevance score to the first search query;
identify a second set of results associated with the second category; and
cause the second set of results to be transmitted to the device, the second set of results being presented to the user of the device.

7. The computing system of claim 6 , wherein the instructions, when executed further enable the computing system to:
receive a third search request, the third search request associated with third audio input data from the device;
analyze the third search request for the indicators that the incorrect category was presented;
determine the third search request does not include one or more indicators that the incorrect category was presented; and
increase the relevance score of the second category for the first search query.

8. The computing system of claim 5 , wherein analyzing the second search request for indicators that an incorrect category was presented further comprises:
identifying a first volume level of the first audio input data;
identifying a second volume level of the second audio input data;
calculating a voice volume d

Assignees

• A9 Com Inc

Inventors

• Kumar Mukul Raj
• Pankaj Balpreet Singh

Classifications

• G06F16/2425

  Iterative querying; Query formulation based on the results of a preceding query · CPC title

• G10L2015/225

  Feedback of the input speech · CPC title

• G10L15/30 Primary

  Distributed recognition, e.g. in client-server systems, for mobile phones or network applications · CPC title

• G10L25/51

  for comparison or discrimination · CPC title

• G06F16/63

  Querying · CPC title