Parking lot availability database for pattern mining

• Main Author: Yap, Hui Fung
• Other Authors: School of Computer Engineering
• Format: Final Year Project
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/59034
• Institution: Nanyang Technological University

As of 2013, the car ownership rate in Singapore is roughly about 12 cars for every 100 people, which makes up to 600 thousands car being owned in Singapore. Every single car has to be parked somewhere at any time when it is not moving. In a small country like Singapore, the need to efficiently manage parking availability becomes a tricky problem to handle. Failure to do so may result in heavy congestion or even accidents within the parking facilities.[1] To resolve the parking issue, this project has been implemented to carry out the necessary features to analyse the parking pattern of the popular car parks located in the central part of Singapore. Parking Lot Availability Database for Pattern Mining (PLADPM) is a web-based project that capture real-time data from the LTA Data Mall and to understand the parking lot pattern through a chart visualization which would aid drivers in Singapore to understand the parking lots behaviour at during the period of the day. A component called the ‘Car Park Lot Availability’ was created and was added into existing project PETRINA. PETRINA stands for Personalizing and Analysing Spatiotemporal Traffic Information, a web-based application that captures real-time data and provides related traffic information such as incident monitoring and analytics services. Weeks’ worth of Car Park Lots Availability (CPLA) data is collected from the Land Transport Authority Data. The data comes in string of texts in the form of Extensible Mark-Up Language (XML) and are streamed every 5 minutes before storing it into the database. This project is a web-based application, whereby the user can view the data collected. Upon user’s query, relevant data will be retrieved and plotted onto a chart to visualise the parking pattern. Two approaches will be used to analyse the pattern to find out the average parking lots available in the car park. Results will provide information like the latest lots availability of the car park based on every 5 minutes (data streaming) and also visualization chart of the availability of parking lots at specific timing (through user filtering). Such information can aid the user to know which period of the day have the most lots available.