A Method of Automatic Detection of Pseudoscientific Publications
Currently, pseudoscientific theories are actively promoted being published in a large amount of papers. They appear in mass media, in patents and even in scientific journals, and it is rather difficult for non-expert to distinguish scientific paper from pseudoscientific. A method for identifying pseudoscientific publications based on automatic text analysis is proposed. At first, the text is partitioned into small fragments consisting of several paragraphs. Then feature extraction occurs using an automatic linguistic analysis and classification of text fragments is implemented by support vector machines. Experiments show that the method divides scientific and pseudoscientific publications into different classes with high accuracy.
KeywordsIdentifying of pseudoscientific papers Intelligent text analysis Support vector machine
Unable to display preview. Download preview PDF.
- 1.RationalWiki, http://rationalwiki.org
- 2.Aleksandrov, E.B.: Answers to Questions about Pseudoscience. Journal (In defense of science) 8 (2011) (in Russian)Google Scholar
- 3.Gitelson, I.I.: Necessity of Government Protection of the People from the Onslaught of Fake Medicine. Journal (In defense of science) 2, 52–55 (2007) (in Russian)Google Scholar
- 4.Sceptic Society, http://www.skeptic.com
- 5.Skeptical Inquirer, http://www.csicop.org
- 6.Journal (In defense of science) 12 (2013)Google Scholar
- 8.Osipov, G., Smirnov, I., Tikhomirov, I., Shelmanov, A.: Relational–situational method for intelligent search and analysis of scientific publications. In: Proceedings of the Workshop on Integrating IR Technologies for Professional Search, in Conjunction with the 35th European Conference on Information Retrieval (ECIR 2013), Moscow, Russia. CEUR Workshop Proceedings, vol. 968 (2013)Google Scholar
- 11.LIBSVM – A Library for Support Vector Machines, http://w.csie.org/~cjlin/libsvm