手法を比較

	Reference Publication Year Spectroscopy (RPYS)	Burst Detection (Kleinberg) for Emerging Topics
分野	計量書誌学	計量書誌学
系統	Process / pipeline	Process / pipeline
提唱年	2014	2003
提唱者	Werner Marx, Lutz Bornmann, Andreas Barth & Loet Leydesdorff	Jon Kleinberg
種類	Cited-reference temporal analysis pipeline for historical roots	Temporal burst-detection pipeline for emerging terms and citations
原典	Marx, W., Bornmann, L., Barth, A., & Leydesdorff, L. (2014). Detecting the historical roots of research fields by reference publication year spectroscopy (RPYS). Journal of the Association for Information Science and Technology, 65(4), 751-764.	Kleinberg, J. (2003). Bursty and hierarchical structure in streams. Data Mining and Knowledge Discovery, 7(4), 373-397.
別名	RPYS, Cited-Reference Spectroscopy, Historical Roots Detection	Kleinberg Burst Detection, Citation Burst Analysis, Burst Detection Algorithm
関連	3	3
概要	Reference publication year spectroscopy (RPYS) detects the historical roots of a research field by analyzing not the field's own publications but the publication years of the works those publications cite. Introduced by Werner Marx, Lutz Bornmann, Andreas Barth, and Loet Leydesdorff in 2014, RPYS aggregates all cited references across a corpus, counts how many reference the literature of each past year, and plots the resulting spectrum. Seminal works leave a distinctive mark: the years in which they appeared show up as sharp peaks rising above the smooth background of routine citation. By detecting these peaks — using the deviation of each year's count from a running median — and then inspecting which highly cited references produced them, RPYS pinpoints the foundational papers and books on which a field was built, providing a quantitative, citation-based historiography.	Kleinberg burst detection identifies periods during which a feature in a document stream — a keyword, a phrase, or citations to a particular paper — suddenly surges in frequency, signaling an emerging topic or a moment of intense attention. Introduced by Jon Kleinberg in 2003 to find bursty structure in streams such as email and news, the algorithm models the arrival of events with an infinite-state automaton in which higher states correspond to faster emission rates. A burst is detected when the optimal explanation of the stream requires moving into a high-rate state, with a built-in cost that discourages spurious switching. In scientometrics the method has become a standard way to detect rising research terms and 'citation bursts' — papers or topics whose citation rate spikes — making sudden growth in the literature visible and datable.