Abstract
Background and Objectives: The purpose of this paper was to investigate the elements of the medical science publications indexed in the DOAJ database using the Creative Commons license with data mining approach.
Material and Methods: The research method was descriptive and included the medical journals indexed in the DOAJ database until early 2016. The required data from the DOAJ database were stored as metadata in CVS format. In order to analyze the data, Rapidminer software was used and Crisp standard methodology and C5.0 decision tree algorithm were applied.
Results: The results showed that 28.55% of medical publications did not use Creative Commons license. The maximum number of licenses in this database was the attribution license. Among countries, Egypt ranked first with 24.18%, Britain ranked second with 16.71%. The highest number of publishers belonged to institutional publishers and associations with 502 titles (50.71%), commercial publishers with 307 titles (31.01%), and academic publishers with 181 titles (18.28%). Among the years, 2011 with 146 titles (13.73%) and 2013 with 122 titles (11.47%) had the highest number. Also, the results of the decision tree of the C5.0 presented a model with a reliability of 75.19% and error rate of 24.81% (accuracy -1 = error rate). The maximum number of permissions used in this template were related to attribution permissions with an accuracy of 80.45 and the model was confirmed.
Conclusion: The use of licensing attribution in medicine by countries and publishers is an indicative of community acceptance for the development and continuation of medical knowledge with more acceleration in cyberspace.