Filtering information using targeted filtering schemes

• US 9,201,953 B2
• Filed: 03/04/2014
• Issued: 12/01/2015
• Est. Priority Date: 06/26/2008
• Status: Active Grant

• Current Assignee: Alibaba Group Holding Ltd.
• Original Assignee: Alibaba Group Holding Ltd.
• Inventors: Yang, Junjie, Ni, Liang, Zhang, Zhenghua, Zhang, Zhenyu
• Primary Examiner(s): PARK, GRACE A
• Application Number:
  US14/197,118
• Publication Number:
  US 20140188913A1
• Time in Patent Office: 637 Days
• Field of Search: 707/754
• US Class Current: 1/1
• CPC Class Codes: G06F 16/337 Profile generation, learning or modification
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 21/55 Detecting local intrusion or implementing counter-measures
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: H04L 51/04 Real-time or near real-time messaging, e.g. instant messaging [IM]
• CPC Class Codes: H04L 51/212 using filtering or selective blocking
• CPC Class Codes: H04L 63/0245 Filtering by information in the payload

Abstract

A method for filtering user information takes into account not only specific keywords in the user information, but also related user-characteristic data (e.g., user activity data), and allows targeted user characteristics to be determined from multiple aspects of user activities. In one aspect, the disclosed method adopts different filtering schemes for different types of targeted users to improve the recognition accuracy with respect to the target user information. The method determines a suitable filtering scheme using a correspondence relationship between the filtering scheme and keywords and user-characteristic data. The method uses modeling of sample users and multiple candidate filtering schemes to formulate targeted filtering scheme. An apparatus for implementing the method is also disclosed.

Claims

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

generating a plurality of filtering schemes including first multiple filtering schemes and second multiple filtering schemes, wherein;

the first multiple filtering schemes are defined using first user-characteristic parameters to target different types of users and are different from each other in using different variables for defining the first multiple filtering schemes,

the second multiple filtering schemes are defined using second user-characteristic parameters to target a same type of users, and have a same set of variables and different threshold values to provide different emphasis on respective filtering effectiveness of the second filtering schemes;

the first user-characteristic parameters and the second user-characteristic parameters are generated from one or more targeted keywords and targeted user-characteristic data extracted from a data collection of targeted users;

obtaining one or more keywords and user-characteristic data of a user;

selecting a filtering scheme from the plurality of filtering schemes based at least in part on a correspondence relationship among the filtering scheme, the one or more keywords and the user-characteristic data, wherein the correspondence relationship is determined by comparing the one or more keywords and the user-characteristic data of the user with one or more user-characteristic parameters corresponding to the selected filtering scheme; and

assigning a score to the user based at least in part on the selected filtering scheme; and

filtering information of the user according to the selected filtering scheme to prevent the user from circulating undesired information to other users in response to determining that the score of the user score exceeds a preset threshold.

2. The method as recited in claim 1, further comprising configuring the correspondence relationship, the configuring comprising:

generating one or more user-characteristic parameters of a target user of the targeted users based at least in part on the one or more targeted keywords and the targeted user-characteristic data;

filtering one or more anomalies from the one or more user-characteristic parameters of the target user;

formulating a target filtering scheme based on the one or more filtered user-characteristic parameters of the target user; and

establishing a correspondence relationship among the target filtering scheme, the one or more targeted keywords and the targeted user-characteristic data.

3. The method as recited in claim 2, wherein generating the one or more user-characteristic parameters of the target user comprises:

identifying valid data in the one or more targeted keywords and the targeted user-characteristic data;

selecting one or more sample users based on the valid data; and

obtaining the one or more user-characteristic parameters of the target user based on one or more keywords and user-characteristic data of the one or more sample users.

4. The method as recited in claim 3, wherein obtaining the one or more user-characteristic parameters of the target user comprises:

generating an aggregated variable of the target user based on a total frequency according to the one or more keywords and the user-characteristic data of the one or more sample users;

generating a ratio variable of the target user based on a send/receive ratio according to the user-characteristic data of the one or more sample users; and

generating an average variable based on an average frequency according to the one or more keywords and the user-characteristic data of the one or more sample users.

5. The method as recited in claim 2, wherein filtering the one or more anomalies from the one or more user-characteristic parameters of the target user comprises one or more of:

replacing a missing value in the one or more user-characteristic parameters of the target user by a replacement value;

or

replacing an irregular data in the one or more user-characteristic parameters of the target user by a regular value.

6. The method as recited in claim 1, further comprising modeling at least one filtering scheme to filter a particular type of users, the modeling comprising:

generating one or more user-characteristic parameters of a target user of the targeted users based at least in part on the one or more targeted keywords and the targeted user-characteristic data;

selecting one or more rule generation parameters from the one or more user-characteristic parameters of the target user;

generating multiple candidate filtering schemes based on the one or more rule generation parameters; and

selecting the filtering scheme having a highest accuracy among the multiple candidate filtering schemes to be the at least one filtering scheme for the particular type of users.

7. The method as recited in claim 1, wherein the user-characteristic data of the user comprises user activity data, user information data and network user-characteristic data.

8. The method as recited in claim 7, wherein:

the user activity data includes one or more of frequencies of characteristic word groups in information sent by the user within a time frame, a number of times the user sends/receives information within the time frame, or an amount of information the user sends/receives within the time frame;

the user information data includes one or more of an initial login time of the user, user activities after login, and a number of contacts of the user; and

the network user-characteristic data includes one or more of a number of users being associated with a same IP as the user, and number of users being associated with a same machine ID as the user.

9. An apparatus comprising:

one or more processors;

memory;

a configuration module stored in the memory and executable by the one or more processors to generate a plurality of filtering schemes comprise first multiple filtering schemes and second multiple filtering schemes,

the plurality of filtering schemes including first multiple filtering schemes and second multiple filtering schemes,

the first multiple filtering schemes being defined using first-user-characteristic parameters to target different types of users and are different from each other in using different variables for defining the first multiple filtering schemes,

the second multiple filtering schemes being defined using second user-characteristic parameters to target a same type of users, and have a same set of variables and different threshold values to provide different emphasis on respective filtering effectiveness of the second filtering schemes, and

the first user-characteristic parameters and the second user-characteristic parameters are generated from one or more targeted keywords and targeted user-characteristic data extracted from a data collection of targeted users;

an acquisition module stored in the memory and executable by the one or more processors to obtain one or more keywords and user-characteristic data of a user; and

a filtering module stored in the memory and executable by the one or more processors to;

select a filtering scheme from the plurality of filtering schemes according to a correspondence relationship based at least in part on the one or more keywords and the user-characteristic data of the user, the correspondence relationship being determined by comparing the one or more keywords and the user-characteristic data of the user with one or more user-characteristic parameters corresponding to the selected filtering scheme,

assign a score to the user based at least in part on the selected filtering scheme, and

filter information of the user according to the selected filtering scheme to prevent the user from circulating undesired information to other users in response to determining that the score of the user score exceeds a preset threshold.

10. The apparatus as recited in claim 9, wherein the configuration module comprises:

a parameter generation sub-module used for extracting the one or more targeted keywords and the targeted user-characteristic data from the data collection of targeted users, and for generating one or more user-characteristic parameters of a target user based at least in part on the one or more targeted keywords and the targeted user-characteristic data;

a filtering sub-module used for filtering one or more anomalies from the one or more user-characteristic parameters of the target user generated by the parameter generation sub-module; and

a rule generation sub-module used for generating a targeted filtering scheme for the target user based at least in part on the one or more filtered user-characteristic parameters.

11. The apparatus as recited in claim 10, wherein the parameter generation sub-module comprises:

a recognition unit used for recognizing valid data in the targeted user-characteristic data;

a selection unit used for selecting one or more sample users based on the valid data; and

a computation unit used for computing the one or more user-characteristic parameters of the target user based on user-characteristic data of the one or more sample users selected by the selection unit.

12. The apparatus as recited in claim 10, wherein the filtering sub-module comprises:

a first filtering unit used for replacing a missing value in the one or more user-characteristic parameters of the target user by a replacement value; and

a second filtering unit used for replacing an irregular value in the one or more user-characteristic parameters of the target user by a regular value.

13. The apparatus as recited in claim 10, wherein the rule generation sub-module comprises:

a parameter selection unit used for selecting one or more rule generation parameters;

a rule computing unit used for generating multiple candidate filtering schemes based at least in part on the one or more rule generation parameters selected by the parameter selection unit; and

a rule selection unit used for selecting a filtering scheme having a highest accuracy among the multiple candidate filtering schemes to be the targeted filtering scheme of the target user.

14. One or more non-transitory storage media storing executable instructions that, when executed by one or more processors, cause the one or more processors to perform acts comprising:

generating a plurality of filtering schemes including first multiple filtering schemes and second multiple filtering schemes, wherein;

the first multiple filtering schemes are defined using first user-characteristic parameters to target different types of users and are different from each other in using different variables for defining the first multiple filtering schemes,

the second multiple filtering schemes are defined using second user-characteristic parameters to target a same type of users, and have a same set of variables and different threshold values to provide different emphasis on respective filtering effectiveness of the second filtering schemes; and

the first user-characteristic parameters and the second user-characteristic parameters are generated from one or more targeted keywords and targeted user-characteristic data extracted from a data collection of targeted users;

obtaining one or more keywords and user-characteristic data of a user;

selecting a filtering scheme from the plurality of filtering schemes based at least in part on a correspondence relationship among the filtering scheme, the one or more keywords and the user-characteristic data, wherein the correspondence relationship is determined by comparing the one or more keywords and the user-characteristic data of the user with one or more user-characteristic parameters corresponding to the selected filtering scheme; and

assigning a score to the user based at least in part on the selected filtering scheme; and

filtering information of the user according to the selected filtering scheme to prevent the user from circulating undesired information to other users in response to determining that the score of the user score exceeds a preset threshold.

15. The one or more storage media as recited in claim 14, the acts further comprising configuring the correspondence relationship, the configuring comprising:

generating one or more user-characteristic parameters of a target user of the targeted users based at least in part on the one or more targeted keywords and the targeted user-characteristic data;

filtering one or more anomalies from the one or more user-characteristic parameters of the target user;

formulating a target filtering scheme based on the one or more filtered user-characteristic parameters of the target user; and

establishing a correspondence relationship among the target filtering scheme, the one or more targeted keywords and the targeted user-characteristic data.

16. The one or more storage media as recited in claim 15, wherein generating the one or more user-characteristic parameters of the target user comprises:

identifying valid data in the one or more targeted keywords and the targeted user-characteristic data;

selecting one or more sample users based on the valid data; and

obtaining the one or more user-characteristic parameters of the target user based on one or more keywords and user-characteristic data of the one or more sample users.

17. The one or more storage media as recited in claim 16, wherein obtaining the one or more user-characteristic parameters of the target user comprises:

generating an aggregated variable of the target user based on a total frequency according to the one or more keywords and the user-characteristic data of the one or more sample users;

generating a ratio variable of the target user based on a send/receive ratio according to the user-characteristic data of the one or more sample users; and

generating an average variable based on an average frequency according to the one or more keywords and the user-characteristic data of the one or more sample users.

18. The one or more storage media as recited in claim 15, wherein filtering the one or more anomalies from the one or more user-characteristic parameters of the target user comprises one or more of:

replacing a missing value in the one or more user-characteristic parameters of the target user by a replacement value;

or

replacing an irregular data in the one or more user-characteristic parameters of the target user by a regular value.

19. The one or more storage media as recited in claim 14, the acts further comprising modeling at least one filtering scheme to filter a particular type of users, the modeling comprising:

generating one or more user-characteristic parameters of a target user of the targeted users based at least in part on the one or more targeted keywords and the targeted user-characteristic data;

selecting one or more rule generation parameters from the one or more user-characteristic parameters of the target user;

generating multiple candidate filtering schemes based on the one or more rule generation parameters; and

selecting the filtering scheme having a highest accuracy among the multiple candidate filtering schemes to be the at least one filtering scheme for the particular type of users.

20. The one or more storage media as recited in claim 14, wherein the user-characteristic data of the user comprises user activity data, user information data and network user-characteristic data.