Social data filtering system, method and non-transitory computer readable storage medium of the same

• US 9,607,031 B2
• Filed: 05/20/2014
• Issued: 03/28/2017
• Est. Priority Date: 06/18/2013
• Status: Active Grant

• Current Assignee: Institute For Information Industry
• Original Assignee: Institute For Information Industry
• Inventors: Li, Yi-Hsung, Lu, Chun-Hung, Hsieh, Wen-Tai, Chen, Tai-Hung, Chang, Yu-Chi
• Primary Examiner(s): Fleurantin, Jean B
• Application Number:
  US14/281,912
• Publication Number:
  US 20140372459A1
• Time in Patent Office: 1,043 Days
• Field of Search: 707/754
• US Class Current: 1/1
• CPC Class Codes: G06F 16/23 Updating

Abstract

A social data filtering system is provided. The social data filter system comprises a database, a retrieving module, a filtering module and a determining module. The database stores personal data and corresponding identification information. The retrieving module retrieves the personal data and the identification information corresponding to a designated person and generates search information accordingly to retrieve user information and corresponding social interaction information from a plurality of social data sources accordingly to the search information The filtering module performs filtering on the user information and the social interaction information according to the personal data to retrieve filtered information. The determining module determines at least one key word corresponding to the designated person according to the filtered information.

Claims

1. A social data filtering system disposed in a computer comprising:

a database in the computer to store a plurality pieces of personal data and a plurality pieces of corresponding identification information; and

a processor coupled to the database comprises;

a retrieving module to retrieve one piece of the personal data and the identification information corresponding to a target person in the database, and to generate a plurality pieces of search information according to the identification information corresponding to the target person for retrieving a plurality pieces of user information and at least one piece of corresponding social interaction information from a plurality of social data sources according to the plurality pieces of search information;

a filtering module coupled to the retrieving module to perform a filtering process on the plurality pieces of user information and the at least one piece of corresponding social interaction information according to the personal data of the target person for retrieving at least one piece of filtered information, wherein each piece of the filtered information is determined to be as the user information and the social interaction information corresponding to the target person; and

a determining module coupled to the filtering module to determine at least one keyword corresponding to the target person according to the filtered information.

2. The social data filtering system of claim 1, further comprising a social database coupled to the retrieving module and the filtering module to store the retrieved user information and the at least one piece of corresponding social interaction information by the retrieving module, and to provide the stored information to the filtering module for performing the filtering process.

3. The social data filtering system of claim 1, wherein each piece of the retrieved at least one piece of corresponding social interaction information further comprises a corresponding piece of time information;

and the determining module further determines the at least one keyword corresponding to the target person and a piece of designated time information according to the filtering information and the designated time information.

4. The social data filtering system of claim 1, wherein the filtering module further comprises a semantic analysis and matching component to perform a semantic analysis and matching process on the at least one piece of social interaction information corresponding to the user information during the filtering process performed by the filtering module for generating a behavioral relation value among the social interaction information of each of the piece of the user information, wherein the behavioral relation value is generated according to a time relation value, a location relation value and a wording relation value;

and wherein the filtering module searches for one piece of the user information corresponding to the social interaction information that has the behavioral relation value larger than a threshold of the behavioral relation value, which is used to determine if corresponding to a candidate person, and compares the personal data of the target person with all the user information of the candidate person to determine whether the user information and the social interaction information corresponding to the candidate person is the at least one piece of filtered information.

5. The social data filtering system of claim 4, wherein the filtering module further comprises:

a user name retrieving unit to retrieve a user name of the plurality pieces of the user information for performing a hybrid comparison with the identification information of the target person;

a personal data retrieving unit to retrieve an individual data of the plurality pieces of the user information; and

a data clustering and disambiguating unit to perform the data clustering and disambiguating process according to the individual data, the behavioral relation value and a result of the hybrid comparison.

6. The social data filtering system of claim 4, wherein the semantic analysis and matching component further comprises:

an information type preference analysis unit to perform an information type analysis on the plurality pieces of the social interaction information;

a dynamic analysis unit to perform a time analysis and a location analysis on the plurality pieces of the social interaction information; and

a word segmentation unit to perform a relation analysis and a keyword detection on the plurality pieces of the social interaction information.

7. The social data filtering system of claim 1, wherein the filtering module further comprises a social relation retrieving unit to retrieve a piece of social relation information from each piece of the user information and the corresponding social interaction information for each piece of the user information;

and

the filtering module further searches for at least two group of the user information in which a similarity value calculated according to the at least two group is larger than a similarity threshold according to the user information and the corresponding social relation information, use the user information of the at least two group to determine if corresponding to a candidate person, and compares the personal data of the target person with all the user information of the candidate person to determine whether the user information and the social interaction information corresponding to the candidate person is the at least one piece of filtered information.

8. The social data filtering system of claim 7, wherein the filtering module further performs a data clustering and disambiguating process before comparing the personal data of the target person and the user information corresponding to the candidate person to determine whether the user information and the social interaction information corresponding to the candidate person is the at least one piece of filtered information, such that the user information and the social interaction information is determined to be the filtered information after the data clustering and disambiguating process.

9. A social data filtering method used in a computer comprising:

retrieving, by a processor in the computer, one of a plurality pieces of personal data and a plurality pieces of identification information corresponding to a target person from the plurality of personal data and the plurality pieces of corresponding identification information stored in a database in the computer;

generating, by the processor, a plurality pieces of search information according to the plurality pieces of identification information corresponding to the target person;

retrieving, by the processor, a plurality pieces of user information and at least one piece of corresponding social interaction information from a plurality of social data sources according to the plurality pieces of search information;

performing, by the processor, a filtering process on the plurality pieces of user information and the at least one piece of corresponding social interaction information according to the personal data of the target person for retrieving at least one piece of filtered information, wherein each piece of the filtered information is determined to be as the user information and the social interaction information corresponding to the target person; and

determining, by the processor, at least one keyword corresponding to the target person according to the filtered information.

10. The social data filtering method of claim 9, further comprising storing the retrieved user information and the at least one piece of corresponding social interaction information to a social database, and providing the stored information to the filtering module for performing the filtering process.

11. The social data filtering method of claim 9, wherein each piece of the retrieved at least one piece of corresponding social interaction information further comprises a corresponding piece of time information;

and the step of determining the at least one keyword further comprises determining the at least one keyword corresponding to the target person and a piece of designated time information according to the filtering information and the designated time information.

12. The social data filtering method of claim 9, wherein the step of retrieving the filtered information further comprises:

performing a semantic analysis and matching process on the at least one piece of social interaction information corresponding to the user information during the filtering process for generating a behavioral relation value among the social interaction information of each of the piece of the user information, wherein the behavioral relation value is generated according to a time relation value, a location relation value and a wording relation value;

searching for one piece of the user information corresponding to the social interaction information that has the behavioral relation value larger than a threshold of the behavioral relation value, which is used to determined if corresponding to a candidate person;

comparing the personal data of the target person with all the user information of the candidate person; and

determining whether the user information and the social interaction information corresponding to the candidate person is the at least one piece of filtered information.

13. The social data filtering method of claim 12, wherein the step of retrieving the filtered information further comprises:

retrieving a user name of the plurality pieces of the user information for performing a hybrid comparison with the identification information of the target person;

retrieving an individual data of the plurality pieces of the user information; and

performing the data clustering and disambiguating process according to the individual data, the behavioral relation value and a result of the hybrid comparison.

14. The social data filtering method of claim 12, wherein the semantic analysis and matching process comprises an relation analysis, a keyword detection, an information type analysis, a time analysis, a location analysis or a combination of the above.

15. The social data filtering method of claim 9, wherein the step of retrieving the filtered information further comprises:

retrieving a piece of social relation information from each piece of the user information and the corresponding social interaction information to correspond for each piece of the user information;

searching for at least two group of the user information in which a similarity value calculated according to the at least two group is larger than a similarity threshold according to the user information and the corresponding social relation information;

using the user information of the at least two group to determine if corresponding to a candidate person;

comparing the personal data of the target person with all the user information of the candidate person; and

determining whether the user information and the social interaction information corresponding to the candidate person is the at least one piece of filtered information.

16. The social data filtering method of claim 15, wherein the step of retrieving the filtered information further comprises:

performing a data clustering and disambiguating process before comparing the personal data of the target person and the user information corresponding to the candidate person to determine whether the user information and the social interaction information corresponding to the candidate person is the at least one piece of filtered information, such that the user information and the social interaction information is determined to be the filtered information after the data clustering and disambiguating process.

17. A non-transitory computer readable storage medium to store a computer program to execute a social data filtering method by a computer, wherein the object-sharing method comprises:

retrieving one of a plurality pieces of personal data and a plurality pieces of identification information corresponding to a target person from the plurality of personal data and the plurality pieces of corresponding identification information stored in a database in the computer;

generating, by a processor in the computer, a plurality pieces of search information according to the plurality pieces of identification information;

retrieving, by the processor, a plurality pieces of user information and at least one piece of corresponding social interaction information from a plurality of social data sources according to the search information;

performing by the processor, a filtering process on the plurality pieces of user information and the at least one social interaction information according to the personal data of the target person to retrieve at least one piece of filtered information, wherein each piece of the filtered information is determined to be as the user information and the social interaction information corresponding to the target person; and

determining by the processor, at least one keyword corresponding to the target person according to the filtered information.