CITATION COUNT PREDICTION USING DIFFERENT TIME SERIES ANALYSIS MODELS
Received: 25.04.2021; Revised: 28.06.2021, Accepted: 05.07.2021, Published Online: 30.07.2021
Priya Porwal
Amity University, Mumbai 410206, India
priya.porwal20@gmail.com
Manoj H. Devare
Amity University, Mumbai 410206, India
Abstract:
The paper helps to predict the future citation count of a fresh dataset of research papers considering the past values of the citation count of paper using univariate time series analysis models and evaluate its performance through various evaluation metrics. It is important to predict future citation count as it helps to assess researcher’s achievements, promotions, Fund allocation etc. This research is addition to past research where for prediction different parameters like content of paper, author details, venue impact etc. were considered. The real and original data is extracted for dataset from google scholar profile of top ranked authors. Three models of time series, Autoregressive Integrated moving average (ARIMA), Simple Exponential smoothing (SES), Holt Winter’s Exponential Smoothing (HWES) applied to observe the result variations. The models obtained error metric values for complete dataset. All four evaluation metrics were calculated. The best results for the predictions for citation count was obtained from SES and HWES models, whose values were almost the same for all evaluation metrics because of almost no change in formula. Among all four error metrics mentioned in design, MASE gave sensible results with almost all values less than 1. The results showed similar graphs for both SES and HWES for actual and predicted value of citation count as there is negligible difference in formula.
Keywords: Citation count prediction, Holt Winter’s Exponential Smoothing, Mean absolute scaled error, Simple Exponential Smoothing, Time series Analysis