Automated Prediction of Customer Hotspots to Taxi Drivers Using Clustering Techniques and Web Scrapin

Abstract

Taxi service is one of the most important service in our society. There are many mobile application to customer to book a taxi. But the problem is that applications are not utilized properly to find taxis to customers in a city area or a busy environment because the demand for taxi exceeds the supply. Purpose of this research is to predict the taxi travel demand in a city area. So that mobile applications can utilize properly to guide taxi drivers to give a proper service to customers. The prediction is done clustering the historical data such as time, weather, location using clustering techniques like k-means and DBScan we can cluster the data and cluster hotspots of customers can be found. There are websites that shows details about the upcoming events. In that websites we can find event location, time and other details about the event. So using web scraping techniques we can scrape those data to get that event data. Using those data we can notify the taxi drivers about the nearby events. So they can easily find more customers who are attending to those events quickly. By this method the time and money of both taxi drivers can be saved. So the profit of taxi drivers will be increased.