Social Data Science Topic: Predicting Unemployment:
By Andy Lam, Mette Lange, Ida Lykke Kristiansen and Bob Kruithof.
Group 13.
Project Description:
We would like to estimate and predict the unemployment rate in Denmark.
In order to do this, we are going to use data from two different sources, Statistics Denmark and Google Trends.
We are going to examine if we can find certain correlations between the unemployment rate and the search trends for Danish words for ‘Unemployment benefits’ (‘dagpenge’, ‘kontanthjælp’) and words revealing job search, e.g. ‘jobindex’ (a webpage for jobsearch).
Our hypothesis is that an increase in searches of those kind of terms will be correlated with unemployment. Since we have earlier access to the search results, we would be able to predict unemployment in Denmark.
Our inspiration is the flu prediction model, where the flu outbreak is predicted based on google search terms.
The approach on this project
Denmark’s unemployment rate as our target
a. Our team will extract the data from Statistics Denmark as a basis for unemployment.
Use of Google trend for correlation
a. We think that we can find a correlation between unemployment rate and certain Google search words such as ‘Unemployment Benefits’ (in Danish), within specific period.
Why we want to do this project
Unemployment rate as a good economic indicator
a. As mentioned above, Unemployment rate is an important economic indicator to show the condition of Economy and Community problem. It will be great if we can predict the unemployment rate in the future for better policy making.
Technology Advancement
a. In the age of Big Data, it is not such difficult to gather data from Internet to do statistical analysis and develop the model to do the predictive analysis. For example, in our project, we will just use the data from Statistics Denmark and Google Trend so that we can make a correlation between the word searches, e.g. ‘Unemployment benefits’ with the actual unemployment rate during specific periods. By showing that we can make certain prediction models, people might be able to use the models in way that can be beneficial for everyone.
Social Data Science Topic: Predicting Unemployment: By Andy Lam, Mette Lange, Ida Lykke Kristiansen and Bob Kruithof. Group 13.
Project Description: We would like to estimate and predict the unemployment rate in Denmark. In order to do this, we are going to use data from two different sources, Statistics Denmark and Google Trends. We are going to examine if we can find certain correlations between the unemployment rate and the search trends for Danish words for ‘Unemployment benefits’ (‘dagpenge’, ‘kontanthjælp’) and words revealing job search, e.g. ‘jobindex’ (a webpage for jobsearch). Our hypothesis is that an increase in searches of those kind of terms will be correlated with unemployment. Since we have earlier access to the search results, we would be able to predict unemployment in Denmark. Our inspiration is the flu prediction model, where the flu outbreak is predicted based on google search terms.
The approach on this project
Why we want to do this project