Forecasting disease outbreaks by using internet access and search data

Tracking source of the disease and the forecasting the disease outbreaks are vital topic for human life. In recent years, with the rapid development of technology and wide usage of the internet, the amount of data that can be collected to extract meaningful information from the data with data scientists. Data scientists began to seek a relationship between the internet search data and hospital reports. Results of the studies have shown that, there is a relationship between people with internet searches, and their visits to hospitals. Using this relationship, significant amount of research is introduced to predict disease outbreaks. The two objectives outlined in this thesis as the basis. The first objective is, forecasting the disease outbreaks by using frequency data. Second one is to determine the relationship of diseases that share similar symptoms and decide whether this relationship is of importance on forecasting disease outbreaks. Firstly, in this study, Wikipedia, Google Flu Trends and models that are created by the combination of these data sets to predict influenza in the United States of America was tried. According to the results, the models are quite successful in predicting the flu epidemic were created. In the second study, in addition to Wikipedia and Google Flu Trends, Google Trends was also used. In addition, this study does not only cover the influenza disease, but also tries to forecast other disease which have similar symptoms with influenza. Moreover, in this study, the relationship between disease and improvements of the usage of similar disease data sets together were examined. The proposed method reveals the success of the resulting outputs.