Web Graph Validity (Video Lecture)

Validity Questions in Web Graph Analysis. August 2007 presentation given at the annual meeting of the American Political Science Association (APSA), Pre-Conference, MACHINE POLITICS/POLITICS OF THE MACHINE: New Technology in Political Communication Research and Teaching, Chicago, IL. [video link]

References added to "webgraph as content analysis" paper

I have appended a "references" section to the end of my paper, Web Graph Analysis in Perspective: Description and Evaluation in terms of Krippendorff's Conceptual Framework for Content Analysis (version 1.0).

I will post a more complete version of the original paper, with additional analyses of specific networks, in the next few weeks. If you're currently doing research in this area and would like to see it sooner, please drop me an email.

Web Graphs Rising

After a brief hiatus, I'm again doing active work on web graph analysis. The definitive reference to my methodological approach is still my (2005) Web Graph as Content Analysis paper, but a major version update is in the works.

Validity Issues (informal survey)

I'm doing an informal survey about key validity questions communication researchers have about web graph analysis via Govcom.org's Issue Crawler software. If you've got a comment or question please post it to the blog and/or shoot me an email. For an idea of some of the questions see the validity section my online paper at:

http://farrall.org/papers/webgraph_as_content.html#valid

Web Graph as Content Analysis

I find that a great way to approach web graph analysis methodologically is in terms of a more traditional and established method in communications research, content analysis. Although it may seem counterintuitive, web graph analysis fits perfectly well within the conceptual framework of content analysis. My paper "Web Graph Analysis in Perspective: Description and Evaluation in terms of Krippendorff's Conceptual Framework for Content Analysis" presents this argument in detail. If you have questions about the validty of web graph analysis as a research method, this paper will provide some answers.

AoIR, Chicago

If you happen to be in Chicago at the Association for Internet Reseacher's Conference, Internet Generations 6.0, join us in the Lincolnshire room of the Marriot Hotel at 3:30, Thursday (October 6th) for the panel "Emerging Methods for Analyzing A New Generation of Civic Engagement on the Web." I'll be presenting the paper " Issue Politics, Social Networks and the Web Graph" co-authored with Annenberg Dean Michael X. Delli Carpini.

Why Map the Web?

The Webgrapher blog will discuss issues relating on what some might consider to be a narrow domain of networks within cyberspace: static, public HTML documents accessible via the World Wide Web. Other domains (application spaces) of cyberspace, including email, instant messaging, and peer to peer networks, have been and are being mapped by academics and industry researchers, but they will not be the focus of this blog.

Why the relatively narrow approach? This is one question we will explore in this blog, but the simplest answer is that it is the only domain of which I'm currently aware for which solid tools with transparent, open mapping algorithms are publicly available. Specifically, I am referring to the Issue Crawler software developed by Richard Rogers and the Govcom.org foundation. A more well-known tool in the blog circuits, TouchGraph, is, in my opinion, not suitable for academic research, specifically because it makes use of Google's proprietary "similar pages" algorithm. (If you aware of other publicly available tools that can map local portions of the web graph please let me know.)

Another reason to map networks within the Web instead of networks mediated by email or peer to peer applications is that the web tends to be more public and spans a wider range of media objects (text, video, sound, interactive databases) than many other Internet applications. If you are interested in studying broad social structures such as global political policy networks, the Web is an ideal place to look. Of course, this statement is subject to challenge and begs of the question of why we would want to map any kind of network within the Internet in the first place. These issues will come up in the course of this blog.

In the not too distant future, expect a blog entry placing web graph networks within the context of cyberspace networks in general. This will include a distinction between the term cyberspace and Internet, where Internet is a real-time instantiation of the more abstract cultural object cyberspace.

As this is a blog, expect the content here to be fairly informal and somewhat unstructured. Post your comments and questions and I'll do my best to follow up with them. More to come.

Welcome

Welcome to the webgrapher blog, an open space for exploration and discussion of techniques for mapping and analyzing both link and semantic structure in the World Wide Web. Initial postings are by Kenneth Farrall, 3rd Year PhD student at the Annenberg School for Communication, University of Pennsylvania. Postings are open to comment by interested readers.