BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//UNIFR/WEBMASTER//NONSGML v1.0//EN
CALSCALE:GREGORIAN
BEGIN:VEVENT
DTSTART;VALUE=DATE:20231206T151500
DTEND;VALUE=DATE:20231206T151500
UID:14774@agenda.unifr.ch
DESCRIPTION:In this thesis, we primarily introduce two ranking methods in terms \nof scientific networks and markets, respectively. First, considering \nthe exponential growth in the number of academic researchers, \nidentifying the highest quality papers is a very demanding task for \neditors of scientific journals. While several measures exist to \nevaluate a paper's impact post-publication, the challenge of \ndetermining the potential impact of a manuscript during the \nreview process remains an understudied issue. In Section \n\ref{sec:4}, we propose a reviewer-reputation ranking algorithm to \nidentify high-quality papers based on paper citations, where a \nreviewer’s reputation is computed from the correlation between \ntheir past ratings and the current number of citations received by\nthe papers they have evaluated. During the review process, \nreviewers with high reputation scores are given more weight to \ndetermine the quality of papers. We test the algorithm on an \nartificial network with 200 reviewers and 600 papers, as well as on \nthe American Physical Society (APS) data set, including in the \nanalysis 308,243 papers and 274,154 mutual citations. We compare \nour approach with two existing methods, demonstrating that our \nalgorithm significantly outperforms the others in identifying \nmanuscripts with the highest quality. Our findings have the \npotential to enhance the impact of scientific journals, thereby \ncontributing to academic and scientific progress.\n\nSecond, We focus on a centralized platform in online markets that \nhelp buyers and sellers find each other and reduce information \nasymmetries. To better understand the role of an intermediary on \nmarket outcomes, we propose a new platform design model \nwhose foundation rests on the tools developed by physicists \nworking on complex systems. In this model, the platform can \ndecide whether to rank the visibility of products based on the \ncriteria of higher-quality products or higher fees paid by \ncompanies. Our framework allows us to study the influence of \ndifferent platform strategies on player payoffs in a market with \npartially informed consumers. We find a fundamental market \nfailure: the optimal platform strategy minimizes social welfare. \nTherefore, consumer search within the platform must be driven by \na sub-optimal algorithm that solves the trade-off between the cost \nof fees charged by the platform and a high transaction volume.
SUMMARY:Ranking in Science of Science Networks and Market: Algorithm and Analysis.
CATEGORIES:Soutenance de mémoire/thèse
LOCATION:PER 08\, 0.58.5\, Chemin du Musée 3\, 1700 Fribourg
URL;VALUE=URI:https://agenda.unifr.ch/e/fr/14774
END:VEVENT
END:VCALENDAR