Skip to content Skip to navigation

Frank by OCBC Big Data Rockstars (2018)

From left to right are : Low Zhi Wei, Soo Zhi Kai, Yi Xiaoqin, Shi Chen and Liu Yuanjing.

We are a team that came from diverse backgrounds and experiences but with a common goal to learn more about Big Data in Financial Services and how to apply the analytic techniques.

Together, we went through 13 weeks of classes learning about various analytical tools and the applications. The scope was wide, encompassing various aspects of business – from machine learning for clustering and recency-frequency-monetary segmentation for customer analytics, to shift management and forecasting methods for operational management, as well as analysis of large transactional data sets with association rules and market basket analysis.

Respective techniques provided different observations that allowed us to improve business processes or workflows and provided us insights as to which were the important factors from the business and practical perspective to better prioritise the results.

All these learning was consolidated and tested with the data sponsored from OCBC Bank. Guest speaker, Mr. Donald Macdonald (SVP, Head of Group Customer Analytics & Decisioning, OCBC), introduced to us the sheer amount of possibilities that can be discovered and enabled with Big Data, and shared with us how Big Data grew and evolved to an integral decision making component in OCBC. Apart from the quantitative elements, the qualitative aspect of Big Data is equally important as it crystallises the essence of the quantitative results. A strong understanding of the business problems and lively and visually appealing storytelling are crucial in assisting listeners in understanding your key quantitative findings. Mr. Donald also iterated on “never to use pie charts”!

Our project was centered on a large dataset of credit card transactions provided by OCBC. While it appeared ‘plain’ and relatively simple in terms of absolute number of data variables, the sheer interaction of different levels in each of these variables were numerous. Like how Mr. Donald opened our eyes to the possibilities of big data, our exploration of the dataset using the techniques taught in class created multiple, diverse and potentially deep perspectives. We crafted our overall product utilizing the storytelling lessons shared to differentiate ourselves and was inspired to produce a memorable, relatable, and convincing story to the class and final assessors from OCBC.

Our presentation was focused on customer segmentation and leveraged on the multiple findings in our rule association analysis to uncover various opportunities and presented these to OCBC. Though we emerged as the unanimous favourite among all the teams, the assessors commented that there was an improvement in the overall quality of all findings presented and the class learnt and benefitted from various recommendations. The conclusion of the course signals the dawning of our big data journey into the corporate world.

On behalf of the class, we express our deep appreciation to Professor Ma Nang Laik for her close support and consultation sessions with us, and constant reminders and tips that shaped all our eventual presentations. We would like to also thank Mr Donald for his guidance and memorable takeaways.

Last updated on 17 Oct 2018 .