CIKM 2017 — AnalytiCup, BigTransport and An Honourable Mention

Photo credit: Nicolas Lannuzel, licensed under Creative Commons.

As governments embrace new information and communications technologies, the gamut of Smart City initiatives will expand to improve public services and citizen welfare. At the 26th ACM International Conference on Information and Knowledge Management (CIKM) in Singapore last month, some of these emerging tools and research took center stage. Pulse Lab Jakarta presented its research on public transportation and was also pleased to receive an honourable mention for a prototype designed to assess human mobility in Singapore during haze crisis.

Traditionally, CIKM has served as a venue to discuss information and knowledge management, data mining, data analysis, among others. Under the theme “Smart Cities, Smart Nations”, this year’s conference had a focus on Smart Government. The conference convened not only government counterparts, but also technology providers, experts and practitioners from the private sector to explore technology and business collaboration opportunities through open innovation.

Pulse Lab Jakarta was happy to demonstrate two of its ongoing projects at CIKM AnalytiCup and CIKM BigTransport Workshop:

Big data insights to better inform haze crises in Singapore

This year’s CIKM AnalytiCup consisted of four data challenges that were open to members from industrial, government and academic organisations. Pulse Lab Jakarta participated in the DataSpark Mobility Open-Task challenge, which involved making use of datasets from DataSpark and proposing a task and an implementation plan.

Our data science team proposed a human mobility prototype that enhances the information that is currently available in Haze Gazer. Using two types of dataset provided by DataSpark API, our team identified citizens mobility patterns during haze crisis, including their stay locations and origin destinations between two areas.

The other challenges were: Lazada Product Title Quality Challenge, DHL Temperature-Controlled Supply Chain Hackathon Challenge and Shenzhen Meteorological Bureau-Alibaba Short-Time Quantitative Precipitation Forecasting Challenge. The list of winners are available here.

Big data analytics to enhance public transportation service

With public transport being one of the main aspects of Smart City development, the BigTransport Workshop on Big Data Analytics for Enhancing Public Transport focused on highlighting research and innovations that can respond to public transportation issues that have emerged. Some of these concerns include an increased expectation of service quality, comfort and efficiency among commuters; an influx of new commuters working or visiting cities; and an imbalanced supply and demand of services.

In line with the workshop’s focus, Pulse Lab Jakarta presented its analysis of regular commuters’ behaviour patterns in Jakarta. The study investigated commuters’ temporal regularity and looked at how external factors such as weather and traffic conditions affect the behaviours of regular commuting passengers of Jakarta’s BRT system/ TransJakarta.

Some of the other research at the workshop included identifying congestion cascades using only bus trajectory data, predicting train delays and its impact on the railway network by using geo-tagged tweets data, and predicting road network traffic by conducting spatio-temporal clustering based on graph signal processing. The list of the studies are available here.

From conference to collaboration

Mr. Kok Yam Tan, Deputy Secretary (Smart Nation and Digital Government) in the Prime Minister’s Office of Singapore, gave the opening remarks at the conference. He stressed the importance of connecting academics with industry practitioners — and the role of government as a facilitator. As a bridge for example, governments can encourage and support technology research initiatives across different sectors, which can then be useful in the overall scheme of expanding Smart City initiatives.

The Director of Machine Learning at Amazon also struck a similar chord in his keynote speech: while Amazon relies on machine learning algorithms to optimise customer experience, many of their methods have been adopted from published research. With greater collaboration among those in academia, government and the business world through venues like CIKM, more innovative and strategic research will continue to flourish in order to enhance and sustain Smart Cities initiatives around the world.

Our Data Engineer, Imaduddin Amin, presenting during CIKM 2017 in Singapore. (Photo credit: CIKM 2017)

Pulse Lab Jakarta is grateful for the generous support from the Government of Australia.