Method and system for extracting user behavior features to personalize recommendations

• US 10,178,190 B2
• Filed: 09/24/2014
• Issued: 01/08/2019
• Est. Priority Date: 09/25/2013
• Status: Active Grant

• Current Assignee: Alibaba Group Holding Ltd.
• Original Assignee: Alibaba Group Holding Ltd.
• Inventors: Qiao, Ping, Wu, Minggang
• Primary Examiner(s): Obisesan, Augustine K.
• Assistant Examiner(s): Ennaji, Lahcen
• Application Number:
  US14/495,749
• Publication Number:
  US 20150088911A1
• Time in Patent Office: 1,567 Days
• CPC Class Codes: G06F 16/33 Querying
• 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 30/0269 based on user profile or attribute
• CPC Class Codes: H04L 67/535 Tracking the activity of the user network monitoring arrangements H04L43/00; recording of computer activity G06F11/34

Abstract

A method for extracting user features based on user behaviors. The method uses webpage clickstream data of a current user to compute a path correlation between the current user and other users, selects a number of other users whose path correlation with the current user ranks among the highest, and then configures weights in connection to preset tags of websites visited by the selected other users, and computes a user correlation between the current user and the selected other users based on the preset tags and the weights. The method constructs weighted-directed graphs of webpage click paths based on click stream data, and converts computing user correlation to computing a similarity of weighted-directed graphs. The method further combines computing correlation of webpage tags to discover the user'"'"'s clicking habits and personal preferences, and improve the accuracy and efficiency of user clustering.

Claims

1. A method comprising:

obtaining clickstream data of a current user, the clickstream data including a plurality of clickstream data points at a website;

dividing the plurality of clickstream data points into one or more sessions, difference between click times of any two adjacent sessions sequentially sorted by their respective click times being less than or equal to a preset time threshold;

computing a click path correlation between the current user and other users using the clickstream data, the computing the click path correlation including;

generating a clickstream path tree in each session using a current webpage and its source webpage, the clickstream path tree including a node and a path, the node being the current webpage, and the path indicating a connection between the current webpage and its source webpage;

generating a weighted-directed graph by merging one or more clickstream path trees;

assigning a hierarchical weight to each merged node; and

assigning a proportional weight to each merged path;

selecting X other users whose click path correlations with the current user rank among the highest, X being a positive integer;

configuring a comprehensive weight in connection to each of preset tags of webpages visited by the selected X other users; and

computing a user correlation between the current user and the X other users based on the preset tags and comprehensive weights.

2. The method described in claim 1, wherein:

each clickstream data point includes a user identifier, a current webpage, a source webpage, a click time and a length of stay; and

the computing the click path correlation between the current user and the other users further includes;

selecting Y other users who share most common paths with the current user, wherein Y is a positive integer; and

computing the click path correlation between the current user and the Y other users using the weighted-directed graph.

3. The method as recited in claim 1, wherein:

the hierarchical weight to each merged node is determined according to an average length of stay of the node;

the proportional weight to each merged path is determined according to a respective pageview number pv; and

the average length of stay is calculated by dividing a total length of stay of the node by a total number of pageviews of the node, the totals being summed over the plurality of clickstream data points.

4. The method as recited in claim 1, wherein the assigning the hierarchical weight to each merged node comprises:

placing the node into one of the plurality of weight levels divided according to average lengths of stay of the nodes; and

assigning the hierarchical weight to the node according to the weight level to which the node is placed.

5. The method as recited in claim 1, wherein:

each node is a root node or a child node; and

the generating the clickstream path tree using the current webpage and the source webpage comprises;

matching the source webpage of a current clickstream data point with the current webpage of another clickstream data point having a click time prior to that of the current clickstream data point, according to a preset matching criterion; and

if the matching is successful, making the current webpage of the current clickstream data point a child node of a next current webpage of the other matched clickstream data point;

or

if the matching is unsuccessful, generating a new clickstream path tree using the current webpage of the current clickstream data point as the root node thereof.

6. The method as recited in claim 5, wherein the preset matching criterion includes a preset maximum matches, or a preset maximum difference between the click times of the current clickstream data point and the matching clickstream data point.

7. The method as recited in claim 1, wherein the weighted-directed graph is defined by:

U={ . . . ,<

u_i,λ

_i>

, . . . ,<

u_j,λ

_j>

, . . . },

wherein <

u_i, λ

_i>

represents webpage u_i and its user clicking status, u_i is the current webpage visited by the user, and λ

_i is the hierarchical weight of webpage u_i; and

T={ . . . ,<

u_i,u_jδ

_ij>

, . . . ,<

u_i,u_k,δ

_ik>

, . . . }

wherein <

u_i, u_j, δ

_ij>

represents a user click path direction from webpage u_j to webpage u_i, u_i is the current webpage, u_j is the source webpage, and δ

_ij is the proportional weight of the respective user click path direction.

8. The method as recited in claim 2, wherein the click path correlation between the current user and the Y other users is computed using the equation:

9. The method as recited in claim 1, wherein each webpage has a page weight, each preset tag on each webpage has an original weight, and the comprehensive weight of a preset tag is computed using an equation:

10. The method as recited in claim 9, wherein the user correlation between the current user and the X other users is computed using an question:

11. A method comprising:

obtaining user information including a user identifier;

selecting Z other users who have a user similarity ranked among the highest to the user, where Z is a positive integer; and

making a recommendation to the user based on information of the selected Z other users,

wherein the user similarity is generated by a process comprising;

obtaining clickstream data of the user, the clickstream data including a plurality of clickstream data points;

dividing the plurality of clickstream data points into one or more sessions, a difference between click times of any two adjacent sessions sequentially sorted by their respective click times being less than or equal to a preset time threshold;

computing a click path correlation between the user and other users using the clickstream data, the computing the click path correlation including;

generating a clickstream path tree in each session using a current webpage and its source webpage, the clickstream path tree including a node and a path, the node being the current webpage, and the path indicating a connection between the current webpage and its source webpage;

generating a weighted-directed graph by merging one or more clickstream path trees;

assigning a hierarchical weight to each merged node; and

assigning a proportional weight to each merged path;

selecting X other users whose click path correlations with the current user rank among the highest, where X is a positive integer;

configuring a comprehensive weight in connection to each of preset tags of webpages visited by the selected X other users; and

computing a user correlation between the current user and the X other users based on the preset tags and comprehensive weights.

12. A computer-based apparatus comprising:

one or more processors; and

one or more memories stored thereon computer-executable instructions that when executed by the one or more processors, cause the one or more processors to perform acts comprising;

obtaining clickstream data of a current user, the clickstream data including a plurality of clickstream data points;

dividing the plurality of clickstream data points into one or more sessions, difference between click times of any two adjacent sessions sequentially sorted by their respective click times being less than or equal to a preset time threshold;

computing a click path correlation between the current user and other users using the clickstream data, the computing the click path correlation including;

generating a clickstream path tree in each session using a current webpage and its source webpage, the clickstream path tree including a node and a path, the node being the current webpage, and the path indicating a connection between the current webpage and its source webpage;

generating a weighted-directed graph by merging one or more clickstream path trees;

assigning a hierarchical weight to each merged node; and

assigning a proportional weight to each merged path;

selecting X other users whose click path correlation with the current user ranks among the highest, where X is a positive integer;

configuring a comprehensive weight in connection to each of preset tags of webpages visited by the selected X other users; and

computing a user correlation between the current user and the X other users based on the preset tags and comprehensive weights.

13. The computer-based apparatus described in claim 12, wherein:

each clickstream data point includes a user identifier, a current webpage, a source webpage, a click time and a length of stay; and

the computing the click path correlation between the current user and the other users further comprises;

selecting Y other users who share most common paths with the current user, wherein Y is a positive integer; and

computing the click path correlation between the current user and the Y other users, using the weighted-directed graph.

14. The computer-based apparatus as recited in claim 12, wherein:

the hierarchical weight to each merged node is determined according to an average length of stay of the node;

the proportional weight to each merged path is determined according to a respective pageview number pv; and

the average length of stay is calculated by dividing a total length of stay of the node by a total number of pageviews of the node, the totals being summed over the plurality of clickstream data points.

15. The computer-based apparatus as recited in claim 12, wherein the assigning the hierarchical weight to each merged node comprises:

placing the node into one of a plurality of weight levels divided according to the average lengths of stay of the nodes; and

assigning the hierarchical weight to the node according to the weight level to which the node is placed.

16. The computer-based apparatus as recited in claim 12, wherein:

each node is a root node or a child node; and

the generating the clickstream path tree using the current webpage and the source webpage comprises;

matching the source webpage of a current clickstream data point of the plurality of clickstream data points with the current webpage of another clickstream data point of the plurality of clickstream data points having a click time prior to that of the current clickstream data point, according to a preset matching criterion; and

if the matching is successful, making the current webpage of the current clickstream data point a child node of the current webpage of the other matched clickstream data point;

or

if the matching is unsuccessful, generating a new clickstream path tree using the current webpage of the current clickstream data point as the root node thereof.

17. The computer-based apparatus as recited in claim 12, wherein the weighted-directed graph is defined by:

U={ . . . <

u_i,λ

_i>

, . . . ,<

u_j,λ

_j>

, . . . },

wherein <

u_i, λ

_i>

represents webpage u_i and its user clicking status, u_i being the current webpage visited by the user, and λ

_i the hierarchical weight of webpage u_i; and

T={ . . . ,<

u_i,u_j,δ

_ij>

, . . . ,<

u_t,u_k,δ

_ik>

, . . . },

wherein <

u_i,u_j, δ

_ij>

represents a user click path direction from webpage u_j to webpage u_i, u_i is the current webpage, u_j is the source webpage, and δ

_ij is the proportional weight of the respective user click path direction.

18. The computer-based apparatus as recited in claim 13, wherein the click path correlation between the current user and the Y other users is computed using an equation:

19. The computer-based apparatus as recited in claim 12, wherein each webpage has a page weight, each preset tag on each webpage has an original weight, and the comprehensive weight of the preset tag is computed using an equation:

20. The computer-based apparatus as recited in claim 19, wherein the user correlation between the current user and the X other users is computed using an question: