Determining influence in social networks using Social Capital

Loading...
Thumbnail Image

Persistent link to this item

Statistics
View Statistics

Journal Title

Journal ISSN

Volume Title

Title

Determining influence in social networks using Social Capital

Published Date

2014-05

Publisher

Type

Thesis or Dissertation

Abstract

The proliferation of online social networks enables the influence of a person or an event to propagate to every corner of the globe in a very short duration of time. The problem of identifying such key sources of influence is important for a wide variety of applications from sales and marketing to public health and policies. Most of the existing methods for identifying influencers use the process of information diffusion to discover the nodes (people) with the maximum expected information spread. In this work we have developed a novel method for identifying key influencers in a given network. This method works on the premise that people generate more value for their work by collaborating with peers having high influence. The social value generated through such collaborations denotes the notion of individual social capital. At the core of this method we use the popular valuation-allocation approach for finding the individual social capital value. In this approach first we determine the value of the entire network using a valuation function and then we do a fair allocation of this entire network's value amongst the participating nodes (people). We show that our Valuation and Allocation functions satisfy several axioms of fairness and fall under the Myerson's allocation rule class.Also, we implement our allocation rule using an efficient algorithm and show that our algorithm outperforms the baselines in several real life datasets. Especially, for the DBLP collaboration network our algorithm outperforms PageRank, PMIA and Weighted Degree baselines by up to 8% in terms of precision recall and F1-measure.Furthermore, we use Hypergraphs as a tool to model group collaborations more effectively and empirically show the superiority of hypergraph edge weights as compared to dyadic edge weights for identifying influencers.To conclude with we discuss a couple of popular distributed programming paradigms, namely MapReduce and BSP (Bulk Synchronous Parallel) and the implementation of the algorithm on these.

Description

University of Minnesota M.S. Thesis. May 2014. Major: Computer science. Advisor:Jaideep Srivastava.1 computer file (PDF); xiii, 126 pages.

Related to

Replaces

License

Series/Report Number

Funding information

Isbn identifier

Doi identifier

Previously Published Citation

Other identifiers

Suggested citation

Sharma, Dhruv. (2014). Determining influence in social networks using Social Capital. Retrieved from the University Digital Conservancy, https://hdl.handle.net/11299/165583.

Content distributed via the University Digital Conservancy may be subject to additional license and use restrictions applied by the depositor. By using these files, users agree to the Terms of Use. Materials in the UDC may contain content that is disturbing and/or harmful. For more information, please see our statement on harmful content in digital repositories.