Prediction of future citation count with machine learning and neural network

Abstract

The ability to forecast a paper's future citation impact is gaining traction in the academic world. To simplify the process of predicting future citation counts, we choose a binary approach in this study. Job that requires categorizing. The research relies on data collected from 2,600 physiology-related publications found on the Web of Science. Only eight bibliometric parameters of papers cited in the first three years following publication were considered. There are three machine learning models and a neural network developed to see how well these features predict future citation counts. The experimental outcome demonstrates the utility of the chosen characteristics in predicting future citation counts. Predicting future citation counts is a challenging task, but machine learning and neural networks can help.