DataSpark Pte Ltd’s Mobility Open-Task Challenge at the International Conference on Information and Knowledge Management (CIKM) AnalytiCup 2017 attracted researchers from around the world to find creative and meaningful solutions to crucial problems in Singapore. With the incentive of a $17,000 SGD prize pool, participants found solutions for a wide range of industries, such as transportation, urban mobility and environmental issues using DSpark’s Mobility Intelligence APIs.
Coming Singapore, Indonesia, Italy and UK, the top 5 teams presented their solution in person at the CIKM AnalytiCup finals in Singapore, on 6th November 2017. The team at DataSpark Pte Ltd were impressed by the creativity of the finalists in using DSpark’s Mobility Intelligence APIs to help build models and visualisations to solve various problems and industries in Singapore.
The finalists and winners were:
Predicting Taxi Demand-Supply Mismatches by S. S. Jha, M. Milaj, S. F. Chen & A. Misra
Taxi operators are facing growing fierce competition from on-demand ride sharing services. The winning team spotted an opportunity to improve the efficiency of the taxi system, by developing a better predictor of accurate picture of true demand and supply using Mobility Intelligence data. Tested on real taxi drivers, their winning solution reduced idle time between taxi jobs by 10-20%.
A visual network-based tool for transportation planning by M. Ferretti, L. Pappalardo, G. Barlacchi & B. Lepri
With the unpredictability of the weather and faults in the public transportation system, transport planners need to understand the achilles heels and resilience of a transport network. Combining data about where people are moving and data on the public transport network, the team from Italy has created an impressive visualisation and simulation of the impact of transport lines on people’s mobility throughout the city.
Pattern Analysis in Smart City Mobility by Camelia Elena Ciolac
Creating an impressive Big Data architecture platform, this solution processes Mobility Intelligence along with other environmental sensors, painting a holistic picture of the rhythms and human experiences throughout Singapore.
Better Situational Information on Haze Crises in SEA by Imaduddin Amin, Muhammad Reza & Jong Gun Lee
With the haze crisis sweeping over Singapore twice a year, how does it impact people’s mobility? This team from Pulse Lab Jakarta creates a solution to identify where people are during haze, so authorities can plan and respond accordingly.
Detection, Localisation and Characterisation of Transient, Urban Events by K. Jayarajah, N. Athaide, V. Subbaraju & A. Misra
The dense population of Singapore leads to many events, both expected and unexpected, from small to large scale, constantly out throughout the nation. This team from SMU combines DSpark’s Mobility Intelligence data along with social media data to identify events as they happen.
The AnalytiCup was a part of the larger CIKM 2017 conference (International Conference on Information and Knowledge Management). Incorporating 4 diverse challenges with $60,000 in prize and travel money, the AnalytiCup attracted thousands of teams from all over the world. Other corporate sponsors included Shenzhen Meteorological Bureau, Alibaba Cloud, Lazada Group and DHL. Read more about the competition here.
The winners for DataSpark Pte Ltd’s challenge were decided on and awarded their prize money on the day, with the judging team from DataSpark Pte Ltd consisting of Aloysius Lim, Director of Product Management, Giuseppe Manai, Director of Data Science and Keen Dang, Senior Data Scientist. Judging criteria included data modelling methodology, validation and measurement of results, real world testing and the live demo of the application.
Watch our interviews with the finalists: