A Bibliometric Analysis of Input Parameter in Artificial Neural Network Approach for Groundwater Level Prediction

Abstract Rapid growth in the Artificial Neural Network (ANN) approach in groundwater level prediction literature calls for an assessment of the trajectory and impacts to identify key themes and future research directions. In this paper, reported a bibliometric analysis of this literature that focuses on examining input paramater uses, focus of research, and research forward. We used Elsevier’s SCOPUS database, Dimensions, and Google Scholar to search for publications from January 2000 to May 2020 on the ANN approach in groundwater level prediction, and analyzed the final sample of 101 publications using RIS file from Mendeley and Vosviewer software tools. Thematic analysis of abstracts revealed a strong focus on groundwater level prediction with artificial neural network approach. The co-occurrence network map showed the hydro-climatology parameter like precipitation, temperature, and groundwater level connected with a large number of frequently used for input in ANN approach, while the evapotranspiration, evaporation, humidity, river stage, runoff parameter demonstrated much weaker links. Reflected on how these findings may useful for better understand and ultimately be able to use the other hydro-climatology input paramater on groundwater level prediction with artificial neural network approach. Keywords: Artificial neural network, bibliometric, thematic analysis, hydro-climatology, prediction. I. INTRODUCTION1