Innovations

Data Mining

Code

PTA.6

Title

Data Mining

Case study/reference

Mytransport.sg (2016a). Apps Zone [online]. Available at: http://www.mytransport.sg/content/mytransport/home/appszone.html [Accessed 8 February 2016]

Mytransport.sg (2016b). DataMall [online]. Available at: [ionhttp://www.mytransport.sg/content/mytransport/home/dataMall.html [Accessed 8 February 2016]

Recommendations for the design

Public Transport Operators and Authorities need to consider the following issues when designing new services or products:

1) Identify needs/objectives you are trying to satisfy. What data do you need?

2) Are data required reliable ? Which are the operative procedures, responsibilities allocation and control tools used to assure data quality ?

3) Are there any legal restrictions that could hamper implementation? If so, how can these be overcome?

Data mining algorithms are strictly dependant on which are the objectives of data mining analysis to be adopted.

Recommendations for the implementation

Set up appropriate procedures to check continuously the data quality and assure data mining processes are running on reliable data. Adapt operational procedures to support control procedures of data quality

Brief description

Set up of new data mining procedures and tools to gain cross-related knowledge from data produced by ITS systems (fleet monitoring, e-ticketing, APC). Results of this processing will be e.g. where a shelter is better to be installed ? How selling network can be improved ? When/which lines/services are most promising to be considered for enforcing tickets control?

Full description

Data mining solutions contribute to co-develop new markets that will support the stakeholders’ operations and increase the Public Transport’s customer base, economic efficiency, safety, improve public health and life standards. It can be based on the collection of Big Data or the adoption of targeted data centres. Data coming from different ITS are collected and cross-related to: 1) improve service planning, 2) improve knowledge about service operation conditions and optimize procures for service control, 3) provide added-value services to end-users. Feedback data provided by users can be considered as another data sources to collect data.

CIPTEC score

4.5