Building the big data eco system

Posted on July 9, 2014  /  5 Comments

Yesterday, Sriganesh Lokanathan and I spent the day at the inaugural event of a business analytics group, pulled together by the Sri Lanka Ministry of Higher Education and the School of Information and Business Analytics at Deakin University. Seven Sri Lankan universities were represented by faculty as were several Sri Lankan IT firms.

When the issue of university-industry collaboration came up, Srinath Perera of WSO2 said that they were already collaborating effectively, including through the three-way relationship with LIRNEasia and University of Moratuwa.

Clearly, partnerships are critical in a a field like big data where we are all operating at the frontier of knowledge. The data resides inside firms and government organizations. Without access, there is nothing to conduct research on. The infrastructure and tools are being developed by IT firms. Without that, it is impossible to work. The demand for data scientists is so high that any organization, especially in the non-profit sector, would be lucky to hold on to good people for more than a few years. That means that a pipeline of well trained young data analysts is a critical precondition. The knowledge is fast moving and is being developed by those who grapple with big data. Without those people, universities will not be able to offer a useful education.

Contrary to the emphatic protestations of Tim Unwin, the geographer turned ICT expert, this is a metaphorical eco system. The different components need each other. They cannot function alone.

Yesterday, a good foundation was laid. The Ministry of Higher Education, especially Secretary Sunil Nawaratne and Advisor Vishakha Nanayakkara are to be congratulated on their excellent initiative. Deakin is playing a superb role in providing the initial momentum. Now it is up to the local actors to build on the foundation. We will do what we can.


  1. Rohan, I think you meant it when you say data scientists .. you need people with domain knowledge, people who understand inner working in the domain.

    1. As i belive domain knowledge is not far enough to handle this kind of senarios. We need to have ppl who really has the ability of analytical thinking with practicle useage on domain.

  2. Rohan Samarajiva

    Yes, big data demands team work. We need people who know how to analyze the data and people who have good domain knowledge (i.e., people who know urban policy as it plays out in that particular city, if we are using big data to gain a better understanding of a city). If played right, this could be a way to break down the silos within universities and elsewhere.

    What Secretary Higher Ed said yesterday about ability to work in teams and lifelong learning are highly appropriate in this context.

  3. This is the future. If Sri Lanka, especially our universities and IT sector can work with Deakin together we can take our IT/BPO/KPO sector to the next level. We have all the core competencies: capable people, cost effectiveness, conducive environment ( 3Cs) what else? Synergy Effect is the key word.. Thinktank can play a vital role in making it happen. Subha Gaman!

  4. The training for computational scientists has to be expanded if they are going to become effective data scientists. I never learned much about statistics or modeling all the way through many years of EE/CS education and I had to teach myself many of these skills. This has not changed much.