[Paper Review] Mapping the UK Webspace: Fifteen Years of British Universities on the Web

Scott A. Hale, Taha Yasseri+4 more|arXiv (Cornell University)|May 12, 2014

Web visibility and informetrics17 references16 citations

TL;DR

This paper maps the UK web domain (.uk) and its academic subdomain (.ac.uk) from 1996 to 2010 using web archive data, applying network analysis to hyperlink structures. It reveals that geographical proximity strongly influences hyperlink strength between universities, with a gravity-law relationship (exponent a = 0.28 ± 0.02), while institutional ranking and affiliation show only weak or inconsistent effects on link centrality.

ABSTRACT

This paper maps the national UK web presence on the basis of an analysis of the .uk domain from 1996 to 2010. It reviews previous attempts to use web archives to understand national web domains and describes the dataset. Next, it presents an analysis of the .uk domain, including the overall number of links in the archive and changes in the link density of different second-level domains over time. We then explore changes over time within a particular second-level domain, the academic subdomain .ac.uk, and compare linking practices with variables, including institutional affiliation, league table ranking, and geographic location. We do not detect institutional affiliation affecting linking practices and find only partial evidence of league table ranking affecting network centrality, but find a clear inverse relationship between the density of links and the geographical distance between universities. This echoes prior findings regarding offline academic activity, which allows us to argue that real-world factors like geography continue to shape academic relationships even in the Internet age. We conclude with directions for future uses of web archive resources in this emerging area of research.

Motivation & Objective

• To analyze longitudinal changes in the UK web domain (.uk) and its academic subdomain (.ac.uk) using archived web data.
• To understand how institutional characteristics such as league table ranking, geographic location, and affiliation influence hyperlink patterns.
• To assess the feasibility and utility of large-scale web archive data for studying societal and institutional dynamics over time.
• To identify structural patterns in hyperlink networks that reflect real-world social and geographic constraints.

Proposed method

• Constructed a dataset of hyperlinks from archived web pages in the .uk domain using the Internet Archive's Wayback Machine snapshots from 1996 to 2010.
• Applied network analysis techniques to measure hyperlink strength between universities in the .ac.uk subdomain, using both raw and normalized link strength metrics.
• Normalized hyperlink strength using the formula σ_ij = S_ij / (S_i^out * S_j^in) to control for node-level variation in out- and in-degree.
• Applied a gravity-law model to test the relationship between hyperlink strength and geographical distance, fitting the power-law exponent a in σ_ij ∝ d_ij^−a.
• Used moving average smoothing (500-point window) to reduce noise and improve visualization of temporal trends.
• Mapped university networks geographically, coloring links by normalized strength to visualize spatial clustering and distance decay.

Experimental results

Research questions

• RQ1 How has the structure of the UK web domain evolved from 1996 to 2010 in terms of hyperlink density and distribution across subdomains?
• RQ2 To what extent do institutional characteristics such as league table ranking and geographic location influence hyperlink centrality in the academic web?
• RQ3 Does institutional affiliation or status correlate with higher inlink counts or stronger network positions in the .ac.uk domain?
• RQ4 How does geographical distance between universities affect the strength of their hyperlink connections?
• RQ5 To what extent do real-world factors like proximity and institutional prestige shape digital academic relationships on the web?

Key findings

• The academic subdomain .ac.uk receives proportionally more inlinks per domain than commercial or government subdomains, indicating higher network prominence.
• There is no significant clustering of hyperlinks based on institutional affiliation, suggesting that affiliation alone does not drive linking behavior.
• League table ranking shows only partial evidence of influencing network centrality, with no strong or consistent correlation across all measures.
• A clear inverse relationship exists between hyperlink strength and geographical distance, with the normalized strength following a gravity-law model (exponent a = 0.28 ± 0.02).
• The spatial pattern of links is strongly influenced by geography, with dense clusters in regions like London, Oxford, and Cambridge, and longer-distance links showing significantly weaker strength.
• The gravity-law relationship holds across all years studied, indicating persistent geographic constraints on digital academic networking despite the global reach of the web.