Method and system for providing query suggestions based on user feedback

1. A method, implemented on at least one computing device each of which has at least one processor, storage, and a communication platform connected to a network for providing query suggestions, the method comprising:

receiving, in a search session, an input including a first prefix of a query from a user;

fetching a first plurality of query suggestions based on the first prefix of the query;

providing, in the search session, the first plurality of query suggestions to the user as a response to the input;

prior to receiving a second prefix of the query that is subsequent to the first prefix, monitoring reaction of the user with respect to each of the first plurality of query suggestions based on the first prefix of the query in the search session, wherein the reaction includes a negative user feedback or a neutral user feedback with respect to at least one of the first plurality of query suggestions which is not selected by the user after being provided to the user, and wherein:

if the at least one of the first plurality of query suggestions is ranked top in the first plurality of query suggestions, the negative user feedback is represented by a first time spent by the user on the at least one of the first plurality of query suggestions during the input of the first prefix, wherein the first time spent indicates that the user has examined the at least one of the first plurality of query suggestions before deciding not to select the at least one of the first plurality of query suggestions; and

if the at least one of the first plurality of query suggestions is not ranked top in the first plurality of query suggestions, the neutral user feedback is represented by a second time, shorter than the first time, spent by the user on the at least one of the first plurality of query suggestions during the input of the first prefix, wherein the second time spent indicates that the user did not examine the at least one of the first plurality of query suggestions;

receiving, in the search session, an expanded input of the query including the second prefix from the user;

obtaining a second plurality of query suggestions based on the expanded input of the query, the second plurality of query suggestions having an initial ranking;

modifying the initial ranking of the second plurality of query suggestions based on the monitored reaction of the user indicating the negative user feedback or the neutral user feedback with respect to the at least one of the first plurality of query suggestions based on the first prefix of the query;

providing, in the search session, the second plurality of query suggestions to the user based on the modified rankings as a response to the expanded input; and

responsive to the user submitting a query, obtaining query results.

2. The method of claim 1 , wherein the search session extends from the receiving of a first character of the input until an execution of a search based on the query submitted by the user.

3. The method of claim 1 , wherein determining rankings of the plurality of query suggestions comprises:

obtaining initial rankings of the first plurality of query suggestions based on bias information associated with each of the first plurality of query suggestions; and

adjusting the initial rankings of the first plurality of query suggestions based on the user's previous reaction in the search session with respect to the at least one of the first plurality of query suggestions.

4. The method of claim 3 , wherein

the initial rankings of the first plurality of query suggestions are determined offline; and

the initial rankings of the first plurality of query suggestions are dynamically adjusted at runtime.

5. The method of claim 1 , wherein the user's previous reaction in the search session with respect to the at least one of the first plurality of query suggestions is measured based on at least one of:

number of occurrence of the at least one of the first plurality of query suggestions previously provided to the user in the search session; and

ranking of the at least one of the first plurality of query suggestions when it was previously provided to the user in the search session.

6. The method of claim 1 , wherein the user's previous reaction in the search session with respect to at least one of the first plurality of query suggestions is associated with a weighting parameter in determining rankings of the second plurality of query suggestions.

7. The method of claim 6 , wherein the weighting parameter is estimated offline based on a machine learning model.

8. The method of claim 6 , wherein the weighting parameter is personalized for the user.

9. A system for providing query suggestions, the system comprising:

a communication platform;

a storage having executable instructions stored thereon; and

at least one processor, in response to the executable instructions, operative to:

receive, in a search session, an input including a first prefix of a query from a user;

fetch a first plurality of query suggestions based on the first prefix of the query;

provide, in the search session, the first plurality of query suggestions to the user as a response to the input;

prior to receiving a second prefix of the query that is subsequent to the first prefix, monitor reaction of the user with respect to each of the first plurality of query suggestions based on the first prefix of the query in the search session, wherein the reaction includes a negative user feedback or a neutral user feedback with respect to at least one of the first plurality of query suggestions which is not selected by the user after being provided to the user, and wherein:

if the at least one of the first plurality of query suggestions is ranked top in the first plurality of query suggestions, the negative user feedback is represented by a first time spent by the user on the at least one of the first plurality of query suggestions during the input of the first prefix, wherein the first time spent indicates that the user has examined the at least one of the first plurality of query suggestions before deciding not to select the at least one of the first plurality of query suggestions; and

if the at least one of the first plurality of query suggestions is not ranked top in the first plurality of query suggestions, the neutral user feedback is represented by a second time, shorter than the first time, spent by the user on the at least one of the first plurality of query suggestions during the input of the first prefix, wherein the second time spent indicates that the user did not examine the at least one of the first plurality of query suggestions;

receive, in the search session, an expanded input of the query including the second prefix from the user;

obtain a second plurality of query suggestions based on the expanded input of the query, the second plurality of query suggestions having an initial ranking;

modify the initial ranking of the second plurality of query suggestions based on the monitored reaction of the user indicating the negative user feedback or the neutral user feedback with respect to the at least one of the first plurality of query suggestions based on the first prefix of the query;

provide, in the search session, the second plurality of query suggestions to the user based on the modified rankings as a response to the expanded input; and

responsive to the user submitting a query, obtain query results.

10. The system of claim 9 , wherein the search session extends from the receiving of a first character of the input until an execution of a search based on the query submitted by the user.

11. The system of claim 9 , wherein the at least one processor is further operative to:

obtain initial rankings of the first plurality of query suggestions based on bias information associated with each of the first plurality of query suggestions; and

adjust the initial rankings of the first plurality of query suggestions based on the user's previous reaction in the search session with respect to the at least one of the first plurality of query suggestions.

12. The system of claim 11 , wherein

the initial rankings of the first plurality of query suggestions are determined offline; and

the initial rankings of the first plurality of query suggestions are dynamically adjusted at runtime.

13. The system of claim 9 , wherein the user's previous reaction in the search session with respect to the at least one of the first plurality of query suggestions is measured based on at least one of:

number of occurrence of the at least one of the first plurality of query suggestions previously provided to the user in the search session; and

ranking of the at least one of the first plurality of query suggestions when it was previously provided to the user in the search session.

14. The system of claim 9 , wherein the user's previous reaction in the search session with respect to at least one of the first plurality of query suggestions is associated with a weighting parameter in determining rankings of the second plurality of query suggestions.

15. The system of claim 14 , wherein the weighting parameter is estimated offline based on a machine learning model.

16. The system of claim 14 , wherein the weighting parameter is personalized for the user.

17. A non-transitory machine-readable medium having information recorded thereon for providing query suggestions, wherein the information, when read by the machine, causes the machine to perform the following:

receiving, in a search session, an input including a first prefix of a query from a user;

fetching a first plurality of query suggestions based on the first prefix of the query;

providing, in the search session, the first plurality of query suggestions to the user as a response to the input;

prior to receiving a second prefix of the query that is subsequent to the first prefix, monitoring reaction of the user with respect to each of the first plurality of query suggestions based on the first prefix of the query in the search session, wherein the reaction includes a negative user feedback or a neutral user feedback with respect to at least one of the first plurality of query suggestions which is not selected by the user after being provided to the user, and wherein:

if the at least one of the first plurality of query suggestions is ranked top in the first plurality of query suggestions, the negative user feedback is represented by a first time spent by the user on the at least one of the first plurality of query suggestions during the input of the first prefix, wherein the first time spent indicates that the user has examined the at least one of the first plurality of query suggestions before deciding not to select the at least one of the first plurality of query suggestions; and

if the at least one of the first plurality of query suggestions is not ranked top in the first plurality of query suggestions, the neutral user feedback is represented by a second time, shorter than the first time, spent by the user on the at least one of the first plurality of query suggestions during the input of the first prefix, wherein the second time spent indicates that the user did not examine the at least one of the first plurality of query suggestions;

receiving, in the search session, an expanded input of the query including the second prefix from the user;

obtaining a second plurality of query suggestions based on the expanded input of the query, the second plurality of query suggestions having an initial ranking;

modifying the initial ranking of the second plurality of query suggestions based on the monitored reaction of the user indicating the negative user feedback or the neutral user feedback with respect to the at least one of the first plurality of query suggestions based on the first prefix of the query;

providing, in the search session, the second plurality of query suggestions to the user based on the modified rankings as a response to the expanded input; and

responsive to the user submitting a query, obtaining query results.

18. The medium of claim 17 , wherein determining rankings of the plurality of query suggestions comprises:

obtaining initial rankings of the first plurality of query suggestions based on bias information associated with each of the first plurality of query suggestions; and

adjusting the initial rankings of the first plurality of query suggestions based on the user's previous reaction in the search session with respect to the at least one of the first plurality of query suggestions.

19. The medium of claim 17 , wherein the user's previous reaction in the search session with respect to the at least one of the first plurality of query suggestions is measured based on at least one of:

number of occurrence of the at least one of the first plurality of query suggestions previously provided to the user in the search session; and

ranking of the at least one of the first plurality of query suggestions when it was previously provided to the user in the search session.

20. The medium of claim 17 , wherein the user's previous reaction in the search session with respect to at least one of the first plurality of query suggestions is associated with a weighting parameter in determining rankings of the second plurality of query suggestions.

21. The medium of claim 20 , wherein the weighting parameter is estimated offline based on a machine learning model and personalized for the user.

22. The method of claim 6 , wherein the weighting parameter is determined based on a typing speed of user.