Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

Using Big Data to combat dengue: Teradata interview

AvantiKumar | March 27, 2015
Teradata Malaysia's Craig Morrison talks to Computerworld on how Big Data will help tackle dengue outbreaks via an MDeC Big App winning project involving the Ministry of Health and MMU.

- Rainfall, wind velocity and direction, min/max temperature, solar radiation, humidity and thunderstorm indicator

  • Construction Data

- Quarterly construction statistics by state for the last 5 years

  • Atmospheric Data

- Country level data

  • Survey Data

- Altitude data

Simply, the Dengue Index is:

  • An application that predicts the likely occurrence of dengue outbreak
  • Based on advanced analytics of weather, construction, dengue statistics and other data to produce an index for each locale in Malaysia

The Index examines key variables that predict a dengue outbreak such as weather factors (temperature, humidity, altitude) and others (density, construction)

The Index consists of a Web app incorporating Decision Support System and Dengue Index

  • Standalone app
  • Predictive models using Decision Tree and Bayesian Network
  • Empirical proof of predictive models

What's the next step?

As it was a competition, it has not yet been put into production, we would need more detailed datasets and a regular data feed.  We are very keen to work with the MOH to put this into production to enable Malaysia to better manage dengue and also as a planning tool to manage the distribution of the dengue vaccine which is planned for release later this year.  It's a relatively cheap solution to a very expensive problem, if we only save a handful of lives it will be worth it.

This should now lead to the next step, which is to put the Dengue Index into production.

As well as public health benefits, what other returns do you envisage if Malaysia commercialised this solution?

Many countries are impacted by mosquito-borne diseases so the Dengue Index can be sold as pre-packed solution or operated as hosted service for less well-off countries.

Other monetisation examples: If delivered as hosted services, this solution has the potential to help establish Malaysia as a global Centre of Excellence in this important area (all the data collected could be exploited here).

In addition, the methodology can be extended to tackle other diseases, while the data collected can be used to enhance vaccine efficiency. The data can be also be packaged suitable to be offered as medical research to institutes and companies.

Most of the data to support this application already exists and is readily available.  In fact, we could apply the same methodology to other countries and hence offer it as a service to governments in other countries.  This would be aligned with the Prime Minister's goal to make Malaysia a regional Centre of Excellence for BDA.


Previous Page  1  2 

Sign up for CIO Asia eNewsletters.