Classification method for detecting coercive self-citation in journals |
| |
| Institution: | 1. School of Management, Harbin Institute of Technology, 92 West Dazhi Street, Nan Gang District, Harbin 150001, PR China;2. College of Information and Computer Engineering, Northeast Forestry University, No. 26 Hexing Road, Xiangfang District, Harbin 150040, PR China;1. Rathenau Institute, Science System Assessment, Anna van Saksenlaan 51, 2593 HW The Hague, The Netherlands;2. VU University Amsterdam, Network Institute & Department of Organization Science, De Boelelaan 1105, Amsterdam, The Netherlands;3. GRIPS – National Graduate Institute for Policy Studies, 7-22-1 Roppongi, Minato-ku, Tokyo 106-867, Japan;4. Université Paris-Est, ESIEE – LATTS – IFRIS, 2, bd Blaise Pascal, Noisy le Grand 93160, France;5. CNRS – Aix-Marseille Université, LEST UMR 7317, 35 Avenue Jules Ferry, 13626 Aix en Provence Cedex 01, France;1. CNR-CERIS, National Research Council of Italy, Institute for Economic Research on Firm and Growth, via Real Collegio, 30, I-10024 Moncalieri (TO), Italy;2. University of Torino, Department of Chemistry, via P. Giuria, 7, I-10125 Torino, Italy;1. Laboratory for Studies of Research and Technology Transfer, Institute for System Analysis and Computer Science (IASI-CNR), National Research Council of Italy, Italy;2. Italian National Agency for the Evaluation of Universities and Research Institutes (ANVUR), Italy;3. Department of Management and Engineering University of Rome “Tor Vergata”, Italy |
| |
| Abstract: | Journal self-citations strongly affect journal evaluation indicators (such as impact factors) at the meso- and micro-levels, and therefore they are often increased artificially to inflate the evaluation indicators in journal evaluation systems. This coercive self-citation is a form of scientific misconduct that severely undermines the objective authenticity of these indicators. In this study, we developed the feature space for describing journal citation behavior and conducted feature selection by combining GA-Wrapper with RelifF. We also constructed a journal classification model using the logistic regression method to identify normal and abnormal journals. We evaluated the performance of the classification model using journals in three subject areas (BIOLOGY, MATHEMATICS and CHEMISTRY, APPLIED) during 2002–2011 as the test samples and good results were achieved in our experiments. Thus, we developed an effective method for the accurate identification of coercive self-citations. |
| |
| Keywords: | Scientific journal Coercive self-citation Classification |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
