Providing query recourse with embedded query adjustment options

Patent metadata

Field	Value
Publication number	US-10061866-B2
Application number	US-201514749914-A
Country	US
Kind code	B2
Filing date	Jun 25, 2015
Priority date	Jun 25, 2015
Publication date	Aug 28, 2018
Grant date	Aug 28, 2018

Abstract

A user query may be evaluated to provide a result set. In case the results do not reflect the user's intent, the device may provide recourse options for adjusting the query in a manner that yields more desirable results, e.g., a suggestion at the top of the result set for a different spelling, or recommendations at the end of the results set for additional query techniques that may yield more accurate results. However, such presentation of recourse options may clutter the user interface and/or go unnoticed by the user. Instead, an adjusted query may be identified with an interpreted probability of reflecting the intent of the query. An adjustment option describing the adjusted query may be inserted into the result set, between a higher-probability first result and a lower-probability second result. Selection of the adjustment option may cause the adjusted query to be evaluated on behalf of the user.

First claim

What is claimed is:

1. A server that fulfills a literal query of a user, the server comprising:
a processor; and
a memory storing instructions that, when executed by the processor, cause the server to:
identify, for the literal query, at least two literal query results;
generate result probabilities for the at least two literal query results based on results that were previously selected by other users who previously submitted the literal query, the result probabilities reflecting a probability that a corresponding literal query result matches an intent of the user in submitting the literal query;
identify a sort order according to the result probabilities of the at least two literal query results;
determines, for the literal query, an adjusted query;
evaluate the adjusted query in order to identify, for the adjusted query, one or more adjusted query results;
generate an interpreted probability for the adjusted query based on result probabilities of at least some of the adjusted query results, the interpreted probability reflecting a probability that the user intended the adjusted query;
identify, within the sort order, an adjustment position that is between a first literal query result having a higher result probability than the interpreted probability, and a second literal query result having a lower result probability than the interpreted probability; and
present the at least two literal query results and insert, at the adjustment position, an adjustment option describing the adjusted query.

2. The server of claim 1 , wherein the determining the adjusted query is only performed if
the at least two literal query results have corresponding result probabilities below a result probability accuracy threshold.

3. The server of claim 1 , wherein determining the adjusted query further comprises, after presenting a first result set that is responsive to the literal query:
measuring an engagement of the user with the first result set;
receiving a second query that is responsive to the at least two literal query results; and
after presenting a second result set that is responsive to the second query, responsive to measuring an engagement of the user with the second result set that is higher than the first engagement of the user with the first result set, identify the second query as an adjusted query of the literal query.

4. The server of claim 1 , wherein the memory stores further instructions that, when executed by the processor, cause the server to further:
generate a query adjustment set that correlates literal queries and corresponding adjusted queries, each adjusted query having an interpreted probability higher than a result probability of at least one result of a corresponding literal query; and
determining the adjusted query by reference to a previously generated query adjustment set.

5. The server of claim 1 , wherein the memory stores further instructions that, when executed by the processor, cause the server to further:
evaluate the adjusted query, responsive to a selection of the adjustment option, in order to identify multiple adjusted query results comprising the one or more adjusted query results identified by a prior evaluation of the adjusted query; and
present the identified multiple adjusted query results.

6. A method of fulfilling a literal query, the method comprising:
executing, by one or more processors of one or more servers, instructions that cause at least some of the one or more servers to:
identify, for the literal query, at least two literal query results;
generate result probabilities for the at least two literal query results based on results that were previously selected by other users who previously submitted the literal query, the result probabilities reflecting a probability that a corresponding literal query result matches an intent of the user in submitting the literal query;
identify a sort order according to the result probabilities of the at least two literal query results;
determine, for the literal query, an adjusted query;
evaluate the adjusted query in order to identify, for the adjusted query, one or more adjusted query results;
generate an interpreted probability for the adjusted query based on result probabilities of at least some of the adjusted query results, the interpreted probability reflecting a probability that the user intended the adjusted query;
identify, within the sort order, an adjustment position that is between a first literal query result having a higher result probability than the interpreted probability, and a second literal query result having a lower result probability than the interpreted probability; and
present the at least two literal query results and insert, at the adjustment position, an adjustment option describing the adjusted query.

7. The method of claim 6 , wherein identifying the adjustment position further comprises: identifying, within the sort order, a probability difference between a result probability of the first literal query result and a result probability of the second literal query result, where the probability difference is larger than a probability difference threshold.

8. The method of claim 6 , wherein identifying the adjustment position further comprises: identifying, within the sort order, a group of literal query results having corresponding result probabilities that are all within a result probability range limit.

9. The method of claim 6 , wherein identifying the adjustment position further comprises: weighting selection of the adjustment position proportionally with a result count of the at least two literal query results that were presented in response to the literal query.

10. The method of claim 6 , wherein the instructions further cause the one or more processors of the at least some of the one or more servers to, after presenting the at least two literal query results to the user:
monitor an engagement of the user with the at least two literal query results; and
alter the adjustment position of the adjustment option within the sort order according to the monitored engagement of the user with the at least two literal query results.

11. The method of claim 10 ,
wherein monitoring the engagement of the user with the at least two literal query results further comprises
detecting at least one of:
a consideration of a selected literal query result by the user, and
a skipping of at least some of the at least two literal query results by the user; and

wherein further the altering the adjustment position of the adjustment option further comprises:
responsive to the detecting the consideration of the selected literal query result by the user, reducing the adjustment position of the adjustment option in the sort order; and
responsive to the detecting the skipping of the at least some of the at least two results by the user, raising the adjustment position of the adjustment option in the sort order.

12. The method of claim 6 , wherein the instructions further cause the one or more processors of the at least some of the one or more servers to, after presenting the at least two literal query results from the user and responsive to receiving a request to adjust the literal query:
present a second result set to the user responsive to the request, and
raise the adjustment position of the adjustment option in the sort order of the second result set.

13. The method of claim 6 , wherein the instructions further cause the one or more processors of the at least some of the one or more servers to:
evaluate the adjusted query, responsive to a selection of the adjustment option, in order to identify multiple adjusted query resu

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Vaidya Yogesh Ajit
• Ding Hua
• Wu Nan
• Yuen Aaron Chun Win
• Hasham Karim
• Govindarajen Parthasarathy
• Sacheti Arun
• Sun Yanfeng
• Guo Yandong
• Santhanam Deepak

Classifications

• G06F16/90324 Primary

  using system suggestions · CPC title

• G06F17/3097 Primary

  Physics · mapped topic