Using machine learning techniques for rising star prediction in co-author network

Daud, Ali; Ahmad, Munir; Malik, M. S. I.; Che, Dunren

doi:10.1007/s11192-014-1455-8

Cited by 77 publications

(37 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several methods have already been proposed to search for promising researchers using a co-authorship network. These include evaluation indexes that combine parameters such as the co-authorship of papers and the names of journals [12,13], along with classification methods using machine learning [14,15]. However, these studies do not focus on young researchers who have only just started to publish their research results, and they are not designed to find and actively evaluate promising young researchers and students.…”

Section: Related Researchmentioning

confidence: 99%

Analyzing Promising Researchers Using Network Centralities of Co-authorship Networks from Academic Literature

2020

View full text Add to dashboard Cite

We propose a new way of using the betweenness centrality measure with co-author networks from an academic literature database to evaluate young researchers. It is difficult to discover and evaluate promising young researchers with indexes based on the number of cited papers, such as the h-index to which published papers introduce a lag and whose impact only becomes apparent after they have been cited by other papers. We validated the effectiveness of the measure as an index for evaluating young researchers. Our investigation of 1.92 million publications in the biological sciences shows that Research Fellows with the Japan Society for the Promotion of Science (JSPS) have higher rankings and progress more quickly than other researchers. In addition, differences between JSPS Research Fellows and other researchers were observed at earlier stages using the proposed method than with the h-index and with centralities from literature published in the past 4 years. We expect that the proposed use of the betweenness centrality measure can be applied effectively to extract promising young researchers.

show abstract

Section: Related Researchmentioning

confidence: 99%

Analyzing Promising Researchers Using Network Centralities of Co-authorship Networks from Academic Literature

2020

View full text Add to dashboard Cite

show abstract

“…Social network analysis has been employed to investigate relevant outcomes in academia, e.g. by investigating how coauthor networks predict h-index (McCarty, Jawitz, Hopkins, & Goldman, 2013) and future success (Daud, Ahmad, Malik, & Che, 2015). Further, focal individuals within local networks who have the opportunity to lower the cost or increase the benefits of public goods contributions can facilitate cooperation (McAuliffe, Wrangham, Glowacki, & Russell Andrew, 2015).…”

Section: Creds and The Spread Of Open Sciencementioning

confidence: 99%

Practice what you preach: Credibility-enhancing displays and the growth of open science

Kraft‐Todd¹,

Rand²

2020

Preprint

View full text Add to dashboard Cite

How can individual scientists most effectively spread the adoption of open science practices? Engaging in open science practices presents a social dilemma because they are individually costly (given the current incentive schemes in academia) but collectively beneficial (due to production of higher quality and more accessible science). Mechanisms for promoting cooperation in social dilemmas typically rely on normativity—but open science practices are still comparatively rare. Further, individuals may be tempted to dishonestly “virtue signal” due to growing support for open science. We formulate a solution based on the theory of credibility-enhancing displays: advocates who are known to themselves practice the behavior they are advocating for (particularly if they are prestigious) are more effective at convincing others—specifically because their actions provide an honest signal of their belief in the behavior’s value. Thus, advocates for open science practices should find ways to engage in those practices visibly and often.

show abstract

“…A deep-learning model is applied by Wang et al 2 to represent the co-author network features for relationship identification, which has better performance compared with other state-of-the-art methods. Daud et al 21 apply machine learning techniques to predict the features in co-author networks. Masum Billah and Gauch 22 designed and trained a support vector machines (SVM) classifier to identify and predict emerging researchers.…”

Section: Related Workmentioning

confidence: 99%

Interactive temporal display through collaboration networks visualization

Ming¹,

Li²,

Zhang

2018

Information Visualization

View full text Add to dashboard Cite

Visual analytics play an important role in understanding complex datasets. The bibliographic database is often visualized as a collaboration network to illustrate the connections between researchers. Static networks, however, barely reveal any information when the dataset includes temporal variables. In this article, we propose an embedded network visualization to display the temporal patterns hiding in the data and use intelligent filters to avoid occlusion. We examined different graphing styles, such as the temporal display and the time direction, to find the best way to present the temporal features. In addition, we demonstrate the utility of our approach with case studies and evaluations of real bibliographic databases.

show abstract

Using machine learning techniques for rising star prediction in co-author network

Cited by 77 publications

References 32 publications

Analyzing Promising Researchers Using Network Centralities of Co-authorship Networks from Academic Literature

Analyzing Promising Researchers Using Network Centralities of Co-authorship Networks from Academic Literature

Practice what you preach: Credibility-enhancing displays and the growth of open science

Interactive temporal display through collaboration networks visualization

Contact Info

Product

Resources

About