Modelling Epidemic-Related Time-Series Data and Web Search Behaviour

Tee, Jing Ling (2019) Modelling Epidemic-Related Time-Series Data and Web Search Behaviour. [Final Year Project Report] (Unpublished)

PDF (Please get the password by email to repository@unimas.my , or call ext: 082-583914/3973/3933) Tee Jing Ling.pdf Restricted to Registered users only Download (3MB)

Abstract

In Malaysia, there is currently does not have a study discuss the web search behaviour and also the epidemic related information. The purpose of this study was to investigate the correlation between the disease-related Google Trend and the cumulative number of reported cases and after that construct a dynamic forecasting model for epidemic based on online search data from Google Trends. Online tracking system of Internet hit-search volumes, Google Trends was used to explore the web behaviour related to the virus outbreak. Correlation analysis was done between the number of reported cases of Zika Virus Disease (ZKVD) and Hand Foot Mouth Disease (HFMD) and the web search data. The number of reported cases are retrieved from PAHO (Pan American Health Organization), WHO (World Health Organization) and Malaysia’s Open Data Portal. The result show that there is a significance relationship between the Google Search Volumes and the number of reported cases. Linear Regression Model and ARIMA model are used to predict the future cases happens. The comparison between the two model were done to choose the best forecasting method. The result show the ARIMA model was the best forecasting method with the lower AIC value compare to the other model. In conclusion, Google Trends can use as a complementary data for the influenza surveillance. The ARIMA model can be used to optimize and predict the HFMD preventing by prediction on the HFMD cases in Malaysia.

Actions (For repository members only: login required)

View Item