Electronic Social Network Analysis

From E-Learning Faculty Modules


Contents

Module Summary

In the age of Web 2.0, there has been a wide adoption of social media technologies and platforms. There are content sharing sites like YouTube and Flickr. There are microblogging sites like Twitter. There are social networking sites like LinkedIn and Facebook. To extract meanings from these social platforms, there are a range of technologies that help map the network structures of social relationships.

Takeaways

Learners will...

  • Describe the statistical reasoning behind the creation of social network graphs
  • Summarize the basic types of social media platforms in Web 2.0
  • Explain some basic differences between the layout algorithms for social graphs in NodeXL
  • Explain the various applications of technologies (like NodeXL, Maltego Radium / Tungsten, and NCapture of NVivo) for social media network graphing
  • Highlight some of the enablements and affordances of electronic social network analysis

Module Pretest

1. What is a social network graph?

2. How are nodes (entities) and links (relationships) defined for various social media platforms?

3. What are some basic differences between how relational data is portrayed with various layout algorithms?

4. What are some technologies which may be used for electronic social network analysis, and what are some of their unique contributions?

5. What are some of the enablements and affordances of electronic social network analysis?

Main Contents

A Social Network Graph

A social network graph is a two- or three-dimensional node-link diagram that represents social entities and relationships. The social entities may be individuals or groups; the relationships may represent a range of linkages. A social network graph is one tool in structure-mining, which looks at the structures of interrelationships to understand organizations, power, and any number of other features.

An undirected network is expressed using links without directionality (no arrows). A directed network is expressed using links with directionality (with arrows). The latter can show whether a relationship is reciprocal or not vs. whether it just exists or not.

A dual mode graph has at least two types of entities. One common form of this includes entities that are individuals and events, for example.


Definition of Nodes (Entities) and Links (Relationships) for Social Media Platforms

In general, “nodes” on social media platforms are merely human-created accounts. Some of them represent individuals; others represent groups or companies or organizations. Some accounts on microblogging sites were set up by people but may represent robots that microblog in an automated way; likewise, there are accounts that are cyborgs (mixes of human and robot microblogging).

Links are somewhat harder to define. In content sites, they may relate to commenting on another’s video / photo / videostream / photostream as a relationship. There may be following / follower relationships. There may be replies to others’ works or mentions of others’ accounts. In microblogging sites, relationships are follower / following ones, mentions of others @ certain accounts, shared #hashtag conversations, or other links. In social network sites, relationships are direct follower / following types of relationships.


Image:BasicElements.jpg


Relational Data Portrayed with Layout Algorithms

Different layout algorithms depict the interrelationships between entities in different ways. Variances between these visualizations involve the following:

  • The visual representation of the entities,
  • The visual representation of the links, and
  • The uses of the two-dimensional or three-dimensional space.

To sample some of these differences, please visit the NodeXL Graph Gallery and also view the slideshow below (the slideshow shows how one data set manifests differently depending on the layout algorithm).


Technologies for E-SNA

NodeXL: The Network Overview, Discovery and Exploration for Excel tool is a freeware tool created by the Social Media Research Foundation. It enables data extractions from a range of social media platforms (YouTube, Flickr; Twitter; Facebook, and others)…and offers a wide range of layout algorithms.
Maltego Tungsten / Chlorine: Maltego Tungsten is a penetration testing tool that enables fast and easy crawls of the World Wide Web and various http links. Further, it enables a broad range of “transforms” to disambiguate various types of data. This tool also enables the extraction of information from various Twitter accounts. This tool has a broad range of capabilities which are not mentioned here because of unrelatedness to this context.
NVivo 11 Plus: The NCapture tool in NVivo enables the extraction of (some of) the Twitter messages in a microblogging account and their analysis using a variety of text frequency counts and text search capabilities. (This tool has many more capabilities than in e-social network analysis.)


Some Enablements and Affordances of E-Social Network Analysis

Currently, electronic social network analysis can reveal some of the structured relationships between various electronic accounts online. It can capture some of the global features of social networks and the types of relationships that exist in that context (through motif analysis).

This section offers a broad introduction. There are much better and more solid works that introduce this research technique, its main thinkers, its main practices, and insights from this that have enabled advancements in law enforcement, social computing, online ethnographic research, marketing, and other fields.

Examples

How To

The various technologies have highly variant ways to extract the social media network data and then to portray the findings in data visualizations.

Possible Pitfalls

The main pitfall to electronic social media analysis is to assume that the electronic relationships represent actual real-world ties. A lot of what is happening in the world may not show up in electronic social media platforms. Given automation, given the low costs of “reciprocation,” such ties may actually be very weak ones. Also, the social grooming of highly public online relationships will limit what is real vs. what is superficial and vetted.

The nature of structure mining is also natively limiting. A structure suggests how ideas (and other transmittables) might percolate through the social networks, but they do not actually show what is moving through the networks. Additional research will have to be done to actual probe the inner dynamics of an electronic social network and those individuals and others who are running certain accounts.

Module Post-Test

1. What is a social network graph?

2. How are nodes (entities) and links (relationships) defined for various social media platforms?

3. What are some basic differences between how relational data is portrayed with various layout algorithms?

4. What are some technologies which may be used for electronic social network analysis, and what are some of their unique contributions?

5. What are some of the enablements and affordances of electronic social network analysis?

References

NodeXL Graph Gallery


Robert A. Hanneman and Mark Riddle’s “Introduction to Social Network Methods”


The following is a slideshow that introduces the use of NodeXL for social network analysis, in particular data extractions from social media platforms like Twitter, Flickr, YouTube, and to a lesser extent, Facebook. This slideshow is in the format of Shockwave Flash (.swf).



A Brief Overview of Social Network Analysis and NodeXL

Resources

Hansen, D.L., Schneiderman, B., & Smith, M.A. (2011). Analyzing Social Media Networks with NodeXL: Insights from a Connected World. Elsevier. 284.


Review of "Analyzing Social Media Networks..." (in three parts below):


Crawling social media and depicting social networks with NodeXL (Part 1 of 3)

Crawling social media and depicting social networks with NodeXL (Part 2 of 3)

Crawling social media and depicting social networks with NodeXL (Part 3 of 3)