Tag Archives: citation network

CNetS faculty lead two prestigious DoD Minerva projects on the science of science

July 31, 2019News, Santocitation network, collaboration network, deep learning, excellence, funding, graph embedding, impact, neural networks, novelty, science of sciencesanto

Two CNetS teams were awarded prestigious awards from Minerva, a research initiative of the Department of Defense that supports basic social science research focusing on topics of particular relevance to U.S. national security. One of the two awards will develop Science Genome, a new quantitative framework to investigate science of science using representation learning and graph embedding. The $4.4M project will take advantage of the availability of digitized bibliographic data sets and powerful computational methods, such as machine learning with deep neural networks, to tap into hidden information present in complex scholarly graphs. The project is led by YY Ahn and also includes Staša Milojević, Alessandro Flammini, and Fil Menczer (more…). The other award aims to understand the fundamental laws ruling science dynamics: the description and prediction of the evolution of scientific fields, how to define and measure the novelty of a scientific work, how to assemble successful teams to solve a specific task, and how to define and measure the impact of scholars’ research. The $5M project is led by a consortium of seven prominent science of science experts in four US institutions, including CNetS professor Santo Fortunato (more…). Both projects have potential applications in policy-making, for institutions and funding agencies.

Nature story on universality of impact metrics

November 6, 2013Fil, NaN, Newscitation network, scholarometerfil

top-scholars-widget — You can embed this top scholars widget from Scholarometer

A story in Nature discusses a recent paper (preprint) from CNetS members Jasleen Kaur, Filippo Radicchi and Fil Menczer on the universality of scholarly impact metrics. In the paper, we present a method to quantify the disciplinary bias of any scholarly impact metric. We use the method to evaluate a number of established scholarly impact metrics. We also introduce a simple universal metric that allows to compare the impact of scholars across scientific disciplines. Mohsen JafariAsbagh integrated this metric into Scholarometer, a crowdsourcing system developed by our group to collect and share scholarly impact data. The Nature story highlight how one can use normalized impact metrics to rank all scholars, as illustrated in the widget shown here.