Predicting Airbnb user destination using user demographic and session information

In this report, we develop a model to predict the Airbnb user’s booking destination country based on their demographics and session data. This model is very helpful in providing personalized recommendations and targeted marketing to enrich user experience and optimize business conversion. The dataset we used was provided by Airbnb’s user information. The given problem is modelled as a classification problem and found random forest classifier pruned with importance of features to be a very good model to predict user preferred destination. The proposed model has an accuracy of 88% which is better than the baseline model and decision tree classification model. Keywords—Data Analysis, Customer Segmentation, Random Forests.