Identifying query fingerprints

• US 9,727,614 B1
• Filed: 03/17/2014
• Issued: 08/08/2017
• Est. Priority Date: 03/17/2014
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Spinelli, Alexander Michael, Sarmento, Luis Antonio Diniz Fernandes de Morais, Hinegardner, Lisa Jane, Hoover, Thomas Jay
• Primary Examiner(s): Mahmoudi, Tony
• Assistant Examiner(s): Dewan, Kamal
• Application Number:
  US14/217,021
• Time in Patent Office: 1,240 Days
• Field of Search: 707737, 707771, 707723, 707728, 707769, 707780, 707999003, 707999004, 707999005, 707999006, 707E17014, 707E17018, 713386, 382124
• CPC Class Codes: G06F 16/2457 with adaptation to user needs
• CPC Class Codes: G06F 16/248 Presentation of query results
• CPC Class Codes: G06F 16/25 Integrating or interfacing systems involving database management systems
• CPC Class Codes: G06F 16/3325 Reformulation based on results of preceding query
• CPC Class Codes: G06F 16/35 Clustering; Classification

Abstract

A method and system for analyzing user behavior as users search for items within an electronic marketplace is provided. A query is submitted by a user of the electronic marketplace, the query is processed to identify a series of actions or behaviors performed by the user in relation to the query and fingerprint information for the query is determined based at least in part on analyzing the actions. A classification for a query is determined based on the fingerprint information. In one embodiment, an electronic marketplace receives a query from a user, accesses a query classification database having fingerprint information, determines a fingerprint for the query based on the fingerprint information and dynamically modifies a user experience for the user based at least in part on the fingerprint information.

Claims

1. A computer-implemented method, comprising:

receiving a query issued by a user of an electronic marketplace;

identifying a first action performed by the user in relation to the electronic marketplace;

accessing a query classification database having a fingerprint information, the fingerprint information comprising a number of correlations between the first action performed with respect to the electronic marketplace and one or more second actions performed with respect to the electronic marketplace by a plurality of users in relation to previous queries submitted by individual users in the plurality of users as well as a time information for each of the one or more second actions, wherein the time information comprises an average amount of time between a performance of the first action and a performance of the one or more second actions for the plurality of users;

comparing the fingerprint information in the query classification database to the first action;

determining a fingerprint for the query based at least in part on comparing the fingerprint information in the query classification database to the first action and based at least in part on the time information, the fingerprint comprising a sequence of actions performed in relation to the previous queries that includes at least the first action followed by a second action of the one or more second actions, the fingerprint being determined based on a similarity measure between the first action and the sequence of actions, and the second action being identified as a next action in the sequence of actions of the fingerprint information likely to be performed subsequent to the first action based on having been performed by a maximum number of the plurality of users in relation to previous queries; and

determining a classification for the query based at least in part on the fingerprint.

2. The computer-implemented method of claim 1, wherein comparing the fingerprint information in the query classification database comprises identifying one or more search terms in the query based at least in part on the first action.

3. The computer-implemented method of claim 2, wherein comparing the fingerprint information in the query classification database further comprises identifying one or more search terms corresponding to a plurality of queries stored in the query classification database.

4. The computer-implemented method of claim 3, further comprising comparing the one or more identified search terms corresponding to the plurality of queries to the one or more search terms in the query issued by the user.

5. The computer-implemented method of claim 4, further comprising obtaining the fingerprint information of at least one of the plurality of queries based on the comparing and determining the fingerprint for the query issued by the user based at least in part on the fingerprint information.

6. The computer-implemented method of claim 5, wherein the fingerprint for the query comprises a set of actions performed by a plurality of users of the electronic marketplace in relation to the query, the set of actions comprising at least one of a re-formulation of the query, a term swap, a term addition, a term deletion, an abandonment of the query, an item detail page click a refinement of the query or a scope change of the query.

7. A computer-implemented method, comprising:

receiving a query issued by a user of an electronic marketplace;

identifying a first action performed by the user in relation to the electronic marketplace;

accessing a query classification database having information on associations between one or more second actions performed by a plurality of users in relation to previous queries submitted by individual users in the plurality of users, as well as a time information for each of the previous queries, wherein the time information comprises an average amount of time between a performance of the first action and a performance of the one or more second actions for the previous queries;

determining a fingerprint for the query by comparing the information in the query classification database to an action performed by the user in relation to the query, the fingerprint for the query comprising a sequence of actions performed in relation to a similar previous query of the previous queries that includes at least the first action followed by one or more second actions, the one or more second actions being likely to be performed based on having been performed by a maximum number of the plurality of users in relation to the previous queries; and

determining a classification for the query based at least in part on the fingerprint and based at least in part on the time information.

8. The computer-implemented method of claim 7, further comprising identifying one or more search terms in the query issued by the user, based at least in part on the identified first action.

9. The computer-implemented method of claim 8, wherein determining the fingerprint for the query further comprises comparing the fingerprint information of one or more queries in the query classification database to the one or more search terms identified in the query.

10. The computer-implemented method of claim 9, wherein determining the fingerprint for the query comprises combining the fingerprint information of the one or more queries with fingerprint information related to the query.

11. The computer-implemented method of claim 10, wherein the one or more second actions comprise at least one of a re-formulation of the query, a term swap, a term addition, a term deletion, an abandonment of the query, an item detail page click, a refinement of the query or a scope change of the query.

12. The computer-implemented method of claim 7, further comprising storing the fingerprint for the query and the classification for the query in a query classification database.

13. A computerized system, comprising:

memory that stores computer-executable instructions; and

at least one processor configured to access the memory, the at least one processor configured to execute the computer-executable instructions to collectively at least;

identify a query submitted by a user of an electronic marketplace and at least one first action taken with respect to the query;

obtain a fingerprint information stored in a query classification database, the fingerprint information including information on relationships between actions performed in relation to a plurality of previous search sessions submitted by individual users in a plurality of users, as well as time information for each of the actions performed, wherein the time information comprises an average amount of time between a performance of each of the actions performed;

determine a fingerprint for the query based at least in part on the fingerprint information and the at least one first action taken with respect to the query and the time information, the fingerprint for the query comprising a sequence of actions performed in relation to at least one previous search session, the sequence of actions including the at least one first action followed by a second action, the second action having been performed by a maximum number of the one or more users with respect to the query with respect to the time information; and

determine a classification for the query based at least in part on the fingerprint.

14. The system of claim 13, wherein the at least one processor is configured to execute the computer-executable instructions to identify one or more search terms in the query, based at least in part on the identified at least one first action.

15. The system of claim 14, wherein the at least one processor is configured to execute the computer-executable instructions to compare the fingerprint information of a plurality of queries in the query classification database to the one or more search terms identified in the query.

16. The system of claim 15, wherein the at least one processor is configured to execute the computer-executable instructions to determine the fingerprint for the query based at least in part on comparing the fingerprint information of a plurality of queries in the query classification database to the one or more search terms identified in the query.

17. The system of claim 13, wherein the second action comprises at least one of a re-formulation of the query, a term swap, a term addition, a term deletion, an abandonment of the query, an item detail page click a refinement of the query or a scope change of the query.

18. One or more non-transitory computer-readable media collectively storing computer-executable instructions that, when executed by one or more computer systems, configure the one or more computer systems to collectively perform operations comprising:

identifying a first action performed and at least one second action by a user of an electronic marketplace in relation to a query as well as time information that comprises an amount of time between a performance of the first action and a performance of the at least one second action;

accessing a query classification database having a fingerprint information;

comparing the fingerprint information in the query classification database to the action;

determining a fingerprint for the query based at least in part on the fingerprint information and based at least in part on the time information, the fingerprint for the query comprising a sequence of actions performed in relation to a similar previous query performed by at least one second user, the sequence of actions comprising at least the first action followed by the at least one second action, the at least one second action one that is likely to be taken by the user based on the at least one second action having been taken by a maximum number of a plurality of users that have performed the query in the past; and

storing the fingerprint for the query in the query classification database.

19. The computer-implemented method of claim 18, further comprising determining a classification for the query based at least in part on the fingerprint.

20. The computer-implemented method of claim 18, wherein determining the classification for the query comprises classifying the query as a broad query, a well-defined query or a brand query based at least in part on the fingerprint.

21. The computer-implemented method of claim 18, further comprising modifying a search result related to the query based at least in part on the fingerprint.