Granted Patent B2

US 12,210,517 · App. 17/483,436 · Granted Jan 28, 2025

Inventors: Alexander Rusnak (Bellevue, WA); Dragomir Dimitrov Yankov (Palo Alto, CA); Shekoofeh Mokhtari (Sunnyvale, CA); Kupa Tsheko Muyambo Mutungu (Mercer Island, WA)

Assignee: Microsoft Technology Licensing, LLC

G06F16/2428G06F16/29G06F40/274

Abstract

Claims (62)

1. A method for suggesting geographic results to a user, comprising:

providing a user interface comprising an input field;

receiving a prefix of a geographic query in the input field of the user interface;

ranking received candidate results for the prefix of the geographic query;

predicting an accuracy of a query expansion using the candidate results, wherein the accuracy comprises a confidence score based on the likelihood of the user selecting the candidate results; and

based on the ranked candidate results, providing the query expansion of a highest ranked candidate result using two binary gating models in response to the predicted accuracy of the query expansion exceeding a threshold,

wherein the two binary gating models include a phrase model and a word model,

wherein if the phrase model has a lower confidence score than the word model, the query expansion is a word expansion, and

if the phrase model has a same or higher confidence score as the word model, the query expansion is a phrase expansion;

wherein the query expansion is provided at a mental pause point in the geographic query and the prefix of the geographic query is automatically updated with a full query expansion in the input field in response to the predicted accuracy of the query expansion exceeding the threshold and a partial expansion in response to the predicted accuracy being below the threshold,

wherein the full query expansion includes additional characters beyond the partial expansion;

providing, based on execution of the query, a ranked results list to the user in the user interface.

2. The method of claim 1 , wherein the prefix of the geographic query includes one or more of a whole word, a whole phrase, or a number.

3. The method of claim 1 , wherein the mental pause point in the geographic query occurs after completing a word, a number, an acronym, or an abbreviation in the prefix of the geographic query.

4. The method of claim 1 , further comprising:

aggregating a list of the ranked candidate results based on the prefix of the geographic query; and

displaying the list of the ranked candidate results.

5. The method of claim 1 , wherein the ranking is based at least on a distance of the ranked candidate results to a current location of the user.

6. The method of claim 1 , further comprising:

displaying a list of the ranked candidate results.

7. The method of claim 1 , wherein the two binary gating models use features related to the prefix or features related to candidates in the ranked candidate results in providing the query expansion.

8. The method of claim 7 , wherein the features related to the prefix include one or more of a length of the prefix or whether the prefix is a numeric prefix, and

the features related to the candidates in the ranked candidate results include one or more of a similarity of the prefix to the candidates or whether the query expansion is supported by the candidates in a list of the ranked candidate results.

9. The method of claim 1 , wherein the full query expansion includes a full expansion of the geographic query.

10. A method for suggesting geographic results to a user, comprising:

providing a user interface comprising an input field;

receiving a prefix of a geographic query in the input field of the user interface;

determining whether the user is at a mental pause point in the geographic query;

if the user is not at the mental pause point, foregoing providing a query expansion until the user is at the mental pause point; and

if the user is at the mental pause point:

predicting an accuracy of the query expansion, wherein the accuracy comprises a confidence score based on the likelihood of the user selecting a set of candidate results provided based on the prefix of the geographic query; and

based on the ranked candidate results, providing the query expansion of a highest ranked candidate result using two binary gating models in response to the predicted accuracy of the query expansion exceeding a threshold,

wherein the two binary gating models include a phrase model and a word model,

wherein if the phrase model has a lower confidence score than the word model, the query expansion is a word expansion, and

if the phrase model has a same or higher confidence score as the word model, the query expansion is a phrase expansion;

wherein the full query expansion includes a full expansion in the input field of the geographic query including additional sufficient characters beyond the partial expansion to be successfully executed by a user to search the geographical location;

providing, based on execution of the query, a ranked results list to the user in the user interface.

11. The method of claim 10 , wherein the mental pause point occurs after the user completes a term in the geographic query or starts to type a new term and the term includes one or more of a word, a number, an acronym, or an abbreviation.

12. The method of claim 10 , wherein the query expansion is a word expansion or a phrase expansion.

13. The method of claim 10 , wherein the query expansion is based on a list of ranked candidate results.

14. The method of claim 10 , further comprising:

determining whether the user is likely to select the query expansion; and

if the user is likely to select the query expansion, providing the query expansion.

15. The method of claim 14 , wherein determining whether the user is likely to select the query expansion is based on one or more of a length of the prefix, whether the prefix is a numeric prefix, or whether the query expansion is supported by candidates in a list of the ranked candidate results.

16. A system, comprising:

one or more processors;

memory in electronic communication with the one or more processors; and

instructions stored in the memory, the instructions executable by the one or more processors to:

provide a user interface comprising an input field;

receive a prefix of a geographic query in the input field of the user interface;

rank received candidate results for the prefix of the geographic query;

predict an accuracy of a query expansion using the candidate results, wherein the accuracy comprises a confidence score based on the likelihood of the user selecting the candidate results; and

based on the ranked candidate results, provide the query expansion of a highest ranked candidate result using two binary gating models in response to the predicted accuracy of the query expansion exceeding a threshold such that the query expansion is executable by a user,

wherein the two binary gating models include a phrase model and a word model,

wherein if the phrase model has a lower confidence score than the word model, the query expansion is a word expansion, and

if the phrase model has a same or higher confidence score as the word model, the query expansion is a phrase expansion;

wherein the query expansion is provided at a mental pause point in the geographic query and the prefix of the geographic query is automatically updated with a full query expansion in the input field in response to the predicted accuracy of the query expansion exceeding a threshold and a partial expansion in response to the predicted accuracy being below the threshold,

wherein the full query expansion includes additional characters beyond the partial expansion, wherein the query expansion is visually distinct from the prefix as the query expansion is visually distinct from characters entered by the user;

provide, based on execution of the query, a ranked results list to the user in the user interface.

17. The method of claim 1 , wherein the query expansion is limited to a partial expansion based on the predicted accuracy.

18. The method of claim 1 , further comprising:

performing the query expansion in response to a predicted likelihood of use of the query expansion exceeding a threshold.

Assignments (2)

Continuity (2)

References Cited (20)

Cited By (1)

US 12,450,268