This is a Data Science project that will follow a Data Science Framework for Big Data Analytics, aligned to the Scientific Method described by the UK Science Council. The hypothesis is to detect threats to the global food distribution and direction by defining a global food distribution's operational chain, identifying & modelling the Operational Risks using Operational Research Methods and implement a Machine Learning game theoretic algorithm monitoring to predict future threats proactively.

This project is solving the Food Directions challenge.


Can we ever solve the problem of World Hunger and World poverty. If World Hunger could be eliminated, can it be said that World Poverty would follow suit? And in turn would it enable World Peace. The Food Direction Challenge is an ambitious attempt to solve the problem of World Hunger. Smart Farming and Agricultural use of land globally could be the answer to eliminating World Hunger. Feeding everyone may reduce if not eliminate World Poverty and enable World Peace. A bold statement, you may say. This project is going to look at what is stopping this from happening right now and what stepping stones can be put in place to make this bold statement a reality, using Data Science and Big Data Analytics.

In this project, objectives are to visualise data to show the direction of food, today, describing the global food distribution and direction as an Operational Chain that involves people, processes, environment, technologies, policies etc . Exploratory data research will attempt to highlight all the contributing data points historically to date contribute to this Operational Chain. These data points will range from individual and collective people(s), policies, processes, environments, technologies and with as many data points as possible analysed. The results of exploratory data analysis and research of the data points will help to create a story, to be told through infographic data visualisation, and show how that story can change in real time, in future, as new and current data is added.

Having told a data visualisation story of what the current state is and how it got to where it is, the next objective is to implement game theory machine learning algorithms, absorbing all the data points collected, to monitor HOW that story changes in real time and WHY and then use a game theory machine learning algorithm to predict the changes before they happen, indicating HOW and WHY.

Being able to show HOW and WHY changes will happen will enable improved and smart decision making, minimising if not eliminating the impact food direction has on the hungriest and poorest people in the world, and putting those stepping stones in place to enabling World Peace, by using Data Science and Big Data Analytics.


In our journey of analysing the data it became clear that there were clear patterns and imbalance between the amount of land being used for agriculture per country and that value that agriculture was generating for those countries, that indicated there were opportunities to smooth out that imbalance by highlighting the factors interrupting the distribution and direction and tackling those factors head on.

There were also early indicators that some factors were not man-made i.e environmental disasters and disease epidemic which suggested that using current predictive models already commonly used for monitoring the environment and predictive models for monitoring disease epidemics and combining it with agriculture and farming data would provide governments and farmers and everybody in between to put in place damage limitation and mitigation steps that could be taken if warned early enough.

There were also a number of man-made factors i.e. policy changes due to trade agreements, conglomerate food companies' demand versus supply, use of agriculture land for non-consumption purposes i.e biofuel farming interfering and upsetting the balance. Again combining data already captured across these man-factors would enable smarter decision making to limit damaging impact when challenging these man-made factors.

Another discovery, given the fragmented data sets was the time it takes to cleanse data, profile the data and extrapolate to create new data points to plug the data gaps was smarter data capture. Rather that mining the data that there is and hope for the best, here is an opportunity to be smarter about what data points to look out for and capture and actually be proactive through hindsight to go looking for specific types of data and reduce the time to market for identifying and visualising the intelligence in that data. Using the Lavastorm Analytics Engine to audit trail all the cleansing, munging and profiling of the data before modelling allowed third parties to review the process of data preparation and explorative data analysis, and concur with the results quicker as well as enable a reflective review of the process and think about doing it more smartly by improving the data capture at source.

Finally, the last discovery which was highlighted by the varied skillset of the team, was the importance of being able to tell the story that is emanating from the data. Not everybody can see just by looking at the numbers or even at the odd pie chart or line graph. However the decision makers are often the ones that can't see the story and yet they have to make that crucial decision.

Enabling effective communication and storytelling is a crucial part of the Data Science and Analytics. A individual should able to look at the data visualisation (the same way they would look at a picture) and get what they are looking at within 20 seconds. Why should an expert always be needed to explain the numbers, You don't have to hire a specialist every time you look at a picture to explain to you what you need to see?

Project Information

License: GNU General Public License version 2.0 (GPL-2.0)

Source Code/Project URL:


Data Science Framework for Big Data Analytics -
How Does Your Garden Grow Demo Website -
Using Lavastorm to audit trail the analysis -
The Final Project Pitch -


  • Kennedy Bhagwandeen
  • Lovisa Inserra
  • Wah-Kwan Lin
  • Sayara Beg