Approaches and Methods for Predicting the Trend of Scientific Outputs: A Scoping review

Considering that the volume of publications is growing at an increasing speed, forecasting the research trend and identifying emerging issues is of particular importance. It should be noted that choosing the right method for accurately predicting the research trend is very important, which has been the focus of many researchers in recent years. In this regard, the present research aims to present the findings of research conducted in Iran and the world regarding the most important approaches and proposed methods for predicting the trend of scientific outputs in the future.

The present study was conducted using the scoping review method. The implementation of the current research includes 5 stages: 1) Identifying research objectives, 2) Identifying related research, 3) Selecting research, 4) Data extraction, and 5) Summarizing, discussing, and reporting the results. To identify relevant research, international databases in English (ScienceDirect, Emerald, Scopus, Web of Science, and ProQuest) and Iranian databases in Persian (Magiran, Noormags, Civilica, SID, and Irandoc) were searched without considering the time limit. In this research, the PRISMA diagram was used for the sampling and data selection process, and the JBI evaluation tool was used to check the quality of selected sources. Finally, 117 effects were analyzed.

An overview of the studies carried out in forecasting research trends shows that these studies have attracted the most attention of researchers in the world and in Iran during the last two decades (especially from 2012 onwards), and the increasing trend in conducting these researches is evident. The most important studies conducted concerning the future of studies were the studies that identified trends and emerging research topics to determine the future direction of studies. In different periods, various approaches have been used to determine emerging issues and predict future research trends, which can be divided into 5 main categories: scientometric, quantitative and statistical, qualitative, mixed method, text mining, and machine learning. A review of studies showed that the most common approach used to identify emerging topics and predict future research trends was the scientometric approach. However, in recent years, due to the limitations of quantitative analysis and scientometric methods to determine the direction of future research, and with the increase in the volume of scientific production and the problems resulting from the analysis of large volumes of data, advances in computer technology and word processing tools. Text mining and machine learning approaches have been used to identify emerging areas and predict future research trends due to their high power in big data analysis along with traditional scientometric methods. The most important disciplines that have paid attention to the problem of predicting the trend of scientific outputs are related to sciences and engineering. It seems that paying attention to this issue in the mentioned fields can be because the speed of developments in these fields is higher and as a result of the necessity of conducting studies to predict the future developments of studies to synchronize and deal with them correctly before other fields and more has been The most important sources to be analyzed to achieve the future path of researches are the articles published in journals. The reason for the focus of these studies on journal articles can be that in any scientific field, articles are usually the result of research projects, theses, and other research experiences, which due to the limited access to these sources, can allow researchers to quickly access the results of these studies. and provide more convenience. On the other hand, scientific publications publish the latest scientific achievements and research findings in the shortest time, and this causes researchers and those engaged in scientific activities to be aware of the latest and most reliable scientific and research achievements. Therefore, to study the future trend of scientific outputs, articles have been considered more than other sources.

Different models and approaches have been proposed by researchers to follow the evolution of scientific products in the future. But it seems that a combination of quantitative and qualitative approaches is needed to make accurate and reliable predictions and overcome the limitations of each of these methods. Be used simultaneously. In addition, utilizing expert opinions can be considered as a complement to scientometric analysis to predict the future. Having a future research approach in scientific policy-making and research management can play an effective role in charting the prospects of scientific development and provide the possibility of policy-making and planning for the future for researchers and stakeholders in various fields.