China
Africa
Explore chapters and articles related to this topic
Methodological Frameworks for Opportunity Discovery in Innovation and Technology Management
Published in Tugrul Daim, Marina Dabić, Yu-Shan Su, The Routledge Companion to Technology Management, 2023
Citation networks exploit the property of papers and patents of having a list of references pointing back to previous research or inventions from where new knowledge is built. A citation network is built by connecting a set of documents to their cited references, or citing documents, and then connecting those to their subsequent references forming a network of documents. Citation networks can be created by following different approaches, namely direct citations (de Solla Price, 1965), bibliographic coupling (Kessler, 1963), and co-citations (Small, 1973). Properties of citation networks are well understood in the ITM community. They are used particularly in the context of science mapping and emerging technology detection – for instance, to analyze a large field like sustainability research (Kajikawa et al., 2017), or even snapshots of the whole scientific literature (Boyack et al., 2005). Topics represented by citation networks can be decomposed into granular subtopics by applying algorithms rooted in network theory (Lancichinetti and Fortunato, 2009).
Intellectual structure of cybersecurity research in enterprise information systems
Published in Enterprise Information Systems, 2023
Nitin Singh, Venkataraghavan Krishnaswamy, Justin Zuopeng Zhang
A paper’s centrality in the citation network is measured by closeness, betweenness, and degree centrality. In a citation network, each node represents a publication. Betweenness centrality is related to how it acts as an intermediary in terms of connections or links in the network (White and Borgatti 1994). Its score represents the number of times any given node requires it to reach any other node through the shortest path (geodesic distance). The higher the score of any given node, the higher would be its role as a broker. Betweenness scores also indicate if nodes are viewed as leaders. The degree centrality of a node is measured as the number of direct connections of that node. The higher the number of connections, the more central and active the node is. In-degree centrality reflects how other publications recognise a specific publication and, therefore, we consider in-degree centrality to represent the importance of publication. Closeness centrality is a measure of the average shortest distance from one node to another. Thus, it is an indication of how central a node is to others. We compute the centrality measures, including betweenness centrality, closeness centrality, and degree centrality, using Pajek. Table 2 lists the top 20 documents by citation count. The centrality measures for the top five authors in each of the centrality measures are shown in Table 3.
Construction Management and Economics 40th anniversary: investigating knowledge structure and evolution of research trends
Published in Construction Management and Economics, 2023
Islam H. El-adaway, Gasser G. Ali, Radwa Eissa, Mohamad Abdul Nabi, Muaz O. Ahmed, Tamima Elbashbishy, Ramy Khalef
The objective of the ML model is to classify the probability of papers becoming “high-impact” or not after 5 years from publications. To achieve that, a large citation network was created using the list of references for each paper. Then, the citation count for each paper was calculated by tracing the citations from the network. As such, the authors were able to generate a variable reflecting the citation counts of each article after 5 years of publication. Since the authors adopted a classification approach to develop the ML model, a target classification variable is established for each paper to mark if a paper is within the top 95 percentile of papers by citation counts after 5 years of publication. In other words, the top 95 percentile of paper by citation count are labeled as “high-impact” articles. The selection of cutoff years for citation counts as well as the percentile threshold for “high-impact” citations were based on the prior study of Weis and Jacobson (2021). Unlike other previous regression-based studies, the work of Weis and Jacobson (2021) aligns with the scope of the presented study since they provided a classification-based ML algorithm for the prediction of high-impact works (in the domain of biotechnology) instead of particular citation counts. As such, their cutoff years and high-impact thresholds were adopted as reasonable values to serve the purpose of the current model. Nevertheless, it should be noted that both the cutoff year and high-impact thresholds are user-defined values, which can be altered by future researchers without impacting the robustness of the model.
Cognition and emotion in the information systems field: a review of twenty-four years of literature
Published in Enterprise Information Systems, 2022
Wen-Lung Shiau, Xiaoqun Wang, Fei Zheng, Yung Po Tsang
We apply citation analyses combined with SNA to identify influential papers. The importance of a paper is determined by its position in a citation network. There are two ways to judge a given paper: one uses degree centrality, and the other uses betweenness centrality (Verma 2018). Degree centrality is based on the degree of a node, namely, the number of edges directly connected to it (Zhang et al. 2011). The more arcs a node has, the more important it is. Because a citation network is a directed network, distinguishing between in-degree centrality and out-degree centrality is necessary; the former measures the number of papers that cite a particular paper, while the latter measures the number of papers that a particular paper cites. In-degree centrality is more suitable for illustrating a paper’s importance than out-degree centrality, as it reflects the recognition of peer scholars. According to the work of Wasserman and Faust (1994), let g be the number of nodes (papers) in a citation network, be the number of edges that point to node , be the total number of possible edges incident upon node , and be the in-degree centrality of ; then: