Scholarlink

<p>This article examines the Webometrics ranking (WR) system, particularly using Wikidata, which ranks universities globally twice yearly. The problem of not all universities being highly ranked and of having low positions in international rankings has persisted over time. Our analytic approaches for frequency tables, similarity and hierarchical clustering (agglomerative) identified crucial indicators, such as potential identities and statements, that universities can examine to increase their exposure to web ranking. The world’s top 100 universities share common characteristics in their <i>Identifiers</i> and <i>Statements</i>. The clustering results of five popular algorithms named K-means (KM), hierarchical clustering (HC), fuzzy C-means (FCM), Gaussian mixture model (GMM) and density-based spatial clustering of applications with noise (DBSCAN) were computed on the characteristics of university ranking data. KM and FCM show identical clustering findings, showing a perfect correlation according to their methods. However, HC, DBSCAN and GMM produce different results. DBSCAN represents the ranking process by placing most universities in a single group. These clustering findings suggest that top universities worldwide exhibit common patterns in their ranking strategies. The results highlight that specific <i>Identifiers</i> and <i>Statements</i> play a crucial role in shaping web visibility, key elements that institutions may prioritise to enhance their future online presence and reputation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimising university web visibility: strategies using Wikidata identifiers and statements in Webometrics rankings

  • 期刊论文

  • 发表日期：

    2026年4月17日

查看全文
  • Sanju Backory,
  • Arshad Ahmud Iqbal Peer,
  • Mohammad Sameer Sunhaloo

摘要

This article examines the Webometrics ranking (WR) system, particularly using Wikidata, which ranks universities globally twice yearly. The problem of not all universities being highly ranked and of having low positions in international rankings has persisted over time. Our analytic approaches for frequency tables, similarity and hierarchical clustering (agglomerative) identified crucial indicators, such as potential identities and statements, that universities can examine to increase their exposure to web ranking. The world’s top 100 universities share common characteristics in their Identifiers and Statements. The clustering results of five popular algorithms named K-means (KM), hierarchical clustering (HC), fuzzy C-means (FCM), Gaussian mixture model (GMM) and density-based spatial clustering of applications with noise (DBSCAN) were computed on the characteristics of university ranking data. KM and FCM show identical clustering findings, showing a perfect correlation according to their methods. However, HC, DBSCAN and GMM produce different results. DBSCAN represents the ranking process by placing most universities in a single group. These clustering findings suggest that top universities worldwide exhibit common patterns in their ranking strategies. The results highlight that specific Identifiers and Statements play a crucial role in shaping web visibility, key elements that institutions may prioritise to enhance their future online presence and reputation.