Browse Prior Art Database

A Method of finding reliable approach in social networks

IP.com Disclosure Number: IPCOM000236727D
Publication Date: 2014-May-13
Document File: 6 page(s) / 111K

Publishing Venue

The IP.com Prior Art Database

Abstract

Social network analysis is becoming to a key technique in modern computing. It has also gained a significant following in a lot of areas which commonly available as a consumer tool. With the rapid development of the social network analysis some mining methodology within a relationship network becoming realistically. What will be introduced here is a key contact allocation method using relationship mining in social network, which will obviously help out finding relationships in specific areas, significantly reducing corporate effort and cost of in communication and public relationship even sales, and also getting advise or reference from a real hot experiences. And the method will increase the reliability of the recommended approach with social network interaction analysis.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 100% of the total text.

Page 01 of 6

A Method of finding reliable approach in social networks

With keywords and other parameters input by the user, the friends of the user are searched layer by layer in the social network.

For each friend in the searching scope, the score of the path between this friend and the user is computed.

The score metrics include: 1) the relevance between the keywords and the friend's profile or history blogs; 2) the length of the path; 3) the closeness between the friend and the user.

With the score metrics, the score fomular is generated using machine learning methods.


(1) A user interface is provided for users to input the relationship search query parameters, including the keywords, social network parameters (number of layers of friends, original blogs or not, etc.)


(2) Search the friends of the query in the social network, and the friends of friends, in layer by layer.

1



Page 02 of 6


(3) For each friend access, calculate the relevance and closeness to the query.

Score(R) = fomular( rel(D), len(R), closeness(R)), where rel(D) = rel(keyword(query), ProfileD, BlogsD);

len(R) = 3

closeness(R) = w(A, B) + w(B, C) + w(C, D)

2



Page 03 of 6


(4) Generate the score formula based on training data and analytics model.

Training data:

3



Page 04 of 6

analytics model

4



Page 05 of 6


(5)Use the training result for score computing

5



Page 06 of 6

6