Article Preview
TopIntroduction
Computer Science subject domain is developing rapidly, and many researchers find different types of new inventions. These researchers search for better and effective ways to share those new inventions with the research community. When considering sharing new research findings with large peer groups, most of the scientific authors focus on publishing their findings in internationally recognized journals. At present, mainly with the development of the Internet and the increase of open-access publishers, a large number of journals have come into operation. But, most of the journals available today are low-quality and predatory ones. Most of them only focus on earning money by publishing papers without maintaining quality. Predatory publishers claim that they provide good editorial services and peer reviews, but in truth, they will publish almost everything that they receive (and paid for) without proper quality control process. Because of the unavailability of a proper quality control process, poorly conducted and false research studies could be published in many journals. That causes greater consequences because these papers are cited and used as the basis for further studies. Another cause for concern is that honest researchers are sometimes duped into believing that journals are legitimate and may end up publishing their valuable efforts in them. Most of these journals have false editorial boards. People with a good reputation are named as editors without their knowledge.
Therefore, measuring the quality of the available journals and the development of suitable methods and metrics to measure the quality of journals has become a crucial need in the present academic society. Many researchers have paved their attention on finding ways to identify the legitimate journals out of the predatory ones. As a conspicuous solution, Jeffrey Beall (a library scientist at the University of Colorado, Denver, USA) has compiled a list of predatory journals and publishers, which can be used by authors to identify predatory journals out of the legitimate ones in different subject fields. But, only identifying predatory journals was not enough. Researchers examined what quality metrics can be used to measure the quality of journals. Many indicators, such as Journal Impact Factor (JIF), SCImago Journal Rank (SJR), Eigenfactor, H-index, Source Normalized Impact per Paper (SNIP), and Article Influence (AI) were developed in order to measure the quality of journals.
As well as by considering these quality metrics many reputed and prestigious databases such as Scopus, Scimago Journal and Country Rank website, eigenfactor.org website, and Web of Science (WOS) were developed to rank and categorize journals. But most of those databases have used a limited number of quality attributes for ranking journals. As examples, Scimago Journal and Country Rank website have used SJR for ranking, WOS has used JIF for ranking, and Scopus has used SNIP for ranking and categorizing journals. Mainly this study aims at developing a new categorization model for Computer Science journals according to their quality by considering the values of their quality attributes such as JIF, SJR, Eigenfactor, H-index, SNIP, and AI. Further, this focuses on finding the most suitable metrics for measuring the quality of Computer Science journals out of the selected six attributes and developing a more accurate data mining model by considering those quality metrics.
The remainder of this paper is organized as follows. Section 2 discusses related work. Section 3 describes the proposed clustering approach. Section 4 discusses our experiments and evaluation. Section 5 is the discussion section. Finally, Section 6 concludes the study.