EKSPLORASI INDEKS GOOGLE TRENDS PADA PEMODELAN DATA TIME SERIES

Abstract

The use of big data such as the Google Trends Index has been widely developed in various fields. One of them is as a complement to BPS official data in time series data analysis. The purpose of this research is to collaborate with Google Trends index data and BPS data on the analysis of the Open Unemployment Rate in Indonesia. This study will also compare the best analytical methods between classical statistical methods ARIMAX and Artificial Neural Networks in modeling Unemployment datasets in Indonesia. The research period is February 2005 – August 2021. The results show that modeling unemployment rate data in Indonesia by adding a predictor variable in the form of the Google trends index gives better results than modeling without the Google trends index. In addition, the analysis using the classical statistical method ARIMAX gives better results than the Artificial Neural Network method.