Method, apparatus and system of intelligent navigation

• US 9,519,720 B2
• Filed: 04/20/2015
• Issued: 12/13/2016
• Est. Priority Date: 06/12/2010
• Status: Active Grant

• Current Assignee: Alibaba Group Holding Ltd.
• Original Assignee: Alibaba Group Holding Ltd.
• Inventors: Pan, Chunxiang
• Primary Examiner(s): NGUYEN, KIM T
• Application Number:
  US14/691,311
• Publication Number:
  US 20150227635A1
• Time in Patent Office: 603 Days
• Field of Search: 707/706, 707/722
• US Class Current: 1/1
• CPC Class Codes: G06F 16/1734 Details of monitoring file system events, e.g. by the use of hooks, filter drivers, logs
• CPC Class Codes: G06F 16/2455 Query execution
• CPC Class Codes: G06F 16/248 Presentation of query results
• CPC Class Codes: G06F 16/3325 Reformulation based on results of preceding query
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06Q 30/00 Commerce
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

The present disclosure describes a method, an apparatus and a system of intelligent navigation. In one embodiment, a method includes: receiving a user inquiry from a client terminal; searching a navigation dictionary based on the user inquiry to obtain a recommendation result corresponding to the user inquiry, the navigation dictionary including an editor recommendation based on user behavior information; and sending the recommendation result to the client terminal. The present disclosure can enhance the accuracy, relevancy, richness and intelligence of the intelligent navigation, and reduce user search time as well as the search loading on the server.

Claims

1. A method implemented by one or more computing devices, the method comprising:

obtaining user behavior information;

generating reference data based at least in part on statistics obtained from the user behavior information;

compiling the reference data to obtain an editor recommendation corresponding to the user behavior information, the editor recommendation including one or more index keywords, one or more index categories, and corresponding recommendation content based on the reference data;

compiling the editor recommendation into a navigation dictionary;

receiving a user inquiry including a search keyword and search categories from a client terminal;

searching the navigation dictionary based at least in part on the user inquiry to obtain a recommendation result including the recommendation content, index category and a recommendation type, wherein the search keyword of the user inquiry matches the index keyword of the reference data and the search category of the user inquiry matches the index category of the reference data; and

sending the recommendation result including the recommendation content presented based upon the recommendation type to the client terminal.

2. The method as recited in claim 1, wherein searching the navigation dictionary comprises:

searching for the editor recommendation in the navigation dictionary based at least in part on the search keyword;

obtaining the recommendation content corresponding to the search keyword; and

providing the recommendation content as the recommendation result.

3. The method as recited in claim 1, wherein searching the navigation dictionary comprises:

searching for the editor recommendation in the navigation dictionary based at least in part on the search keyword and the search category;

obtaining the recommendation content corresponding to the search keyword and the search category; and

providing the recommendation content as the recommendation result.

4. The method as recited in claim 1, wherein searching the navigation dictionary comprises:

searching for the editor recommendation in the navigation dictionary based at least in part on the search keyword;

obtaining the recommendation type and recommendation content corresponding to the search keyword; and

providing the recommendation type and recommendation content as the recommendation result.

5. The method as recited in claim 1, further comprising:

causing the recommendation content on a user interface to be displayed based at least in part on a format according to the recommendation type.

6. The method as recited in claim 1, further comprising generating a user behavior based at least in part on the user behavior information, wherein the user behavior log comprises a search log, and wherein the reference data comprises a keyword, a search category, and search data.

7. The method as recited in claim 6, wherein compiling the reference data comprises when the search data of the reference data is greater than a predetermined search threshold, generating the editor recommendation which includes an index keyword, an index category, and recommendation content based at least in part on the reference data, wherein the index keyword comprises a keyword of the reference data, the index category comprises a search category of the reference data, and the recommendation content comprises a category, an attribute, or a combination of the category and the attribute corresponding to the search data having a greatest value in the reference data.

8. The method as recited in claim 1, further comprising generating a user behavior log based at least in part on the user behavior information wherein the user behavior log comprises a clicking log, and wherein the reference data comprises a keyword, a search category, and clicking data.

9. The method as recited in claim 8, wherein compiling the reference data information comprises when the clicking data of the reference data is greater than a predetermined clicking threshold, generating the editor recommendation which includes an index keyword, an index category, and recommendation content based at least in part on the reference data, wherein the index keyword comprises a keyword of the reference data, the index category comprises a search category of the reference data, and the recommendation content comprises a category, an attribute, or a combination of the category and the attribute corresponding to the clicking data having a greatest value in the reference data.

10. The method as recited in claim 1, further comprising:

generating a user behavior log based at least in part on the user behavior information which comprises a purchase log, and wherein the reference data comprises a keyword, a search category, and purchase data.

11. The method as recited in claim 10, further comprising:

compiling the reference data wherein when the purchase data of the reference data is greater than a predetermined purchase threshold generating the editor recommendation which includes an index keyword, an index category, and recommendation content based at least in part on the reference data, wherein the index keyword comprises a keyword of the reference data, the index category comprises a search category of the reference data, and the recommendation content comprises a category, an attribute, or the category and attribute corresponding to the purchase data having a greatest value in the reference data.

12. The method as recited in claim 1, further comprising:

obtaining user feedback information corresponding to the editor recommendation;

determining whether the editor recommendation is valid based at least in part on the user feedback information;

keeping the editor recommendation if the editor recommendation is determined to be valid; and

modifying the editor recommendation if the editor recommendation is determined to be invalid.

13. The method as recited in claim 12, further comprising:

receiving user feedback information corresponding to the editor recommendation, the user feedback information comprising a number of clicks corresponding to the editor recommendation and a number of clicks of a given category, and wherein when the number of clicks corresponding to the editor recommendation is smaller than the number of clicks of the given category;

determining that the editor recommendation is valid when the number of clicks corresponding to the editor recommendation is smaller than the number of clicks of the given category.

14. A system comprising:

one or more processors;

a memory;

an acquisition module stored in the memory and executable by the one or more processors to receive a user inquiry from a client terminal, wherein the user inquiry comprises a search keyword and search categories;

an inquiry module stored in the memory and executable by the one or more processors to search a navigation dictionary based at least in part on the user inquiry to obtain a recommendation result corresponding to the user inquiry, types of content included in the user inquiry, and the navigation dictionary including an editor recommendation which is obtained based at least in part on user behavior information, wherein the editor recommendation is obtained by;

generating reference data based at least in part on statistics obtained from the user behavior information including the recommendation content, one or more index categories and a recommendation type, wherein the search keyword of the user inquiry matches one or more index keywords of the reference data and the search category of the user inquiry matches the index category of the reference data;

compiling the reference data to obtain the editor recommendation corresponding to the user behavior information, the editor recommendation including the index keywords, the index categories, and corresponding recommendation content based on the reference data; and

compiling the editor recommendation into the navigation dictionary; and

a transmission module stored in the memory and executable by one or more processors to send the recommendation result including the recommendation content presented based upon the recommendation type to the client terminal.

15. A method implemented by one or more computing devices, the method comprising:

obtaining the user behavior information;

providing a user behavior log based at least in part on statistics obtained from the user behavior information;

generating reference data based at least in part on statistics obtained from the user behavior information;

compiling the reference data to obtain an editor recommendation corresponding to the user behavior information wherein the editor recommendation comprises at least a search keyword, a first category and a first recommendation content;

compiling the editor recommendation into the navigation dictionary;

determining whether a number of clicks associated with a second category that correspond to the search keyword and not included in the editor recommendation is fraudulent data in an event that the number of clicks associated with the second category is more than a number of clicks associated with the first category included in the editor recommendation; and

performing a corresponding action in response to determining that the number of clicks associated with the second category is not fraudulent data wherein the corresponding action comprises one of deleting the editor recommendation which includes the first category from the navigation dictionary or replacing the first category by the second category and further the first recommendation content by a second recommendation content which corresponds to the second category in the editor recommendation.

16. The method of claim 15, further comprising:

receiving a user inquiry from a client terminal;

searching the navigation dictionary wherein the search is based at least in part on the user inquiry to obtain a recommendation result corresponding to the user inquiry; and

sending the recommendation result to the client terminal.

17. The method of claim 16, further comprising:

obtaining types of content from the recommendation results based at least in part on types of content included in the user inquiry.