Analysis of Public Awareness on Global Warming: Forecasting using Google Trends and FB PROPHET Joko Handoyo, Indra Gunawan, Suluh Jatmiko
Sekolah Tinggi Teknologi Ronggolawe
Abstract
Environmental preservation, especially global warming, is the main topic that is the focus of discussion in all countries in the world today. Considering that the earth^s temperature is increasing over time, this topic becomes important. Machine learning is a derivative branch of artificial intelligence, one of whose capabilities can be used to make data/dataset-based forecasting/predictions. The aim of this research is to analyze the causes of the decline in public awareness of environmental conservation, especially global warming. This research uses a dataset obtained from Google Trend for 2004-2023 using 15 keywords related to global warming and 9 contradictory keywords. The results obtained from this research are that in general it can be said that the world community^s interest in the issue of global warming will decrease in the next few years. Another finding is that there are two top keywords related to disbelief in the truth of global warming which are actually increasing. It can be concluded that the decline in awareness is caused by people^s distrust of the true facts of global warming. The two contradictory keywords received Mean Absolute Percentage Error scores of 0.16 and 0.15, R2 values of 0.91 and 0.87. The next finding is that, by adding a new column, namely max (taking the largest value from another column) in the dataset, it can be used to combine several related keys, if there is a lack of data among these keywords. The score obtained by the additional max column is, Mean Absolute Percentage Error value 0.5, R2 value 0.86.
Keywords: Forecasting- machine learning- google trends- global warming- caused of global warming-
