Unemployment in the Czech Republic and Job Search on the Internet
|Author:||Bc. Ondřej Zacha|
|Year:||2015 - summer|
|Leaders:|| Mgr. Petr Polák MSc. Ph.D.
|Work type:|| Bachelors
|Awards and prizes:|
|Abstract:||This thesis examines the relationship between Czech unemployment rate and job search related behavior
of Internet users. The study uses a simple autoregressive model and augments it with search query data
from two most popular Czech search engines, Google and Seznam as well as data on numbers of job
vacancies and reactions to them from job search portal Jobs.cz. Our results show that data on number of
job vacancies can moderately improve short-term forecasts (“nowcasts”) of Czech unemployment rate in
terms of RMSE and MAE, whereas search query data from Google and Seznam failed to improve
predictive ability of the baseline model.