Detection of Predatory Publishing Using Mixed Dataset: Proposal of a Data Model

Research Publications typically showcases new findings and conclusions. To ensure the reliability of publication, various authentic databases have been curated such as enlisting scholarly articles, research papers, etc. One such key database is Scopus. It is one of the largest scholarly directories of peer-reviewed literature in scientific, technical, medical, and social science. In this work, data from varied authentic publishing sources such as Scopus, DOAJ, Beall's, and Open Dataset for Predatory Publishers is collected, merged, and transformed into a single schema, and validation of publishers and their titles is analyzed. The publisher and their titles being classified into three categories: green, orange, and red nodes based on their title's presence in different authentic sources. All the green nodes publishers have every single title placed under the legitimate category in all our focused data sources, orange ones have either of their titles in the suspicious category and red ones are completely illegitimate ones. The results show the change in the percentage of legitimate publishers individually and produced single schema with a significant difference. With the proposed strategy the possibility of finding the legitimation of publishers becomes more accurate for further scientific assignments.