Creating Special Issues Automatically for Papers Accepted in Journals

Abstract

Significant increases have been observed in the number of scientific publications in recent years. Every year thousands of researchers participate in academic life. Researchers share their studies and results with other researchers. This sharing is very important for academic progress. One of the most important platforms where publications reach other researchers is journals. Journals periodically publish articles. Some journals also publish special issues. These issues can be about a current topic, event or important person. In our study, a method for creating a special issue from the articles that are not published in the regular issues is proposed. In some cases, too many articles are sent to journals and these articles accumulate because all of them cannot be published in regular issues. The aim of this study is to determine the ones on similar topics and to create a special issue. The similarity of the articles to each other, i.e. the probability of being on a common topic, is found by the Doc2Vec method which is one of the deep learning methods. A special issue is created by selecting a certain number of articles with high similarity. The title of the special issue is determined from the common topic of the articles. In this way, both the authors' waiting time for publication of their articles is reduced and a special issue is created from the available articles in popular topic. The articles in this issue, belong to a current topic published by many researchers, will get more attention and be highly cited from other researchers. © 2019 IEEE.