CIKM AnalytiCup

Participate in DataSpark Mobility Open-Task Challenge

Phase 1 - Proposal

In this phase, your task is to propose a problem that you would like to solve in the area of Mobility—for example, transportation, health and fitness, urban planning, tourism or others. Your proposed solution could be a mobile app or a Web app, and must use at least 2 data sources: 1 provided by DataSpark, and 1 other publicly available data source.

Please submit a 2-page proposal with the following outline:
● Problem Statement
● Data Sources
● Proposed Analysis or Use of Data
● Proposed Mobile or Web Application
● Potential Impact of Solution

High-quality proposals received in Phase 1 will be shortlisted as semi-finalists for the Phase 2 of the competition. The evaluation criteria used in the shortlisting process include: relevance to a mobility problem, creativity in use of data, rigour of data analysis or modelling, creativity of solution, and potential impact of the solution.

Proposal Submission Deadline - 12:00 UTC+8, 1 May ’17
Semi-finalists Notification - 18:00 UTC+8, 29 May ’17

Phase 2 - Prototype and Solution Report

In this phase, you have to build a working prototype of the web or mobile application that addresses your proposed problem statement. The application must call at least 1 data API provided by DataSpark (details will be provided if you are shortlisted). It must also include at least 1 other public data set.

You will need to demonstrate not only how your solution works, but also your data analysis or modelling methodology to solve the problem. We are looking for solutions that are not only creative and effective, but are also built on strong data analysis fundamentals.

Please submit a 4-page solution report with the following outline:
● Problem Statement
● Data Sources
● Analysis / Modelling Methodology
● Accuracy Validation & Methodology
● Impact on Problem
● Web or Mobile Application Design
● Comparison with other state of the art solutions
● A link to a video demonstrating a working solution (Max 8mins)

The evaluation criteria in this phase include: analysis and comparison with state of the art solutions that solve the same problem, skill in data analysis or modelling, accuracy of analysis or modelling, impact of solution, design of web or mobile application. It is mandatory that the working prototype makes use of the required data sources, including calls to DataSpark APIs.

Report Submission Deadline - 12:00 UTC+8, 24 Jul ’17
Finalists Notification - 18:00 UTC+8, 4 Sep ’17

Phase 3 - Final Judging via Live Demo

In the finals, you will be presenting in person to our panel of judges a live demo of your web or mobile application, which is expected to be an improved version of the prototype submitted in Phase 2. The winning team will be one that has a strong solution that combines multiple data sources, applies rigorous data analysis, solves the stated problem with a well-designed application, and achieves real world results.

At least one representative from each finalist team must be present in person at the CIKM 2017 conference in Singapore for judging on November 6, 2017. You will be allotted a 15-minute period to deliver a presentation covering:
● Data analysis / modelling methodology
● Validation / accuracy measurement methodology & results
● Real world testing & results
● Live demo of web / mobile application

The evaluation criteria in the finals include: reasoning and decision for accuracy measurement method choices, statistical analysis of the fused data, replicability of the application and analysis across other geo-locations, design of web or mobile application, validation of impact with real users, and incorporation of user actions in feedback loop to improve/augment the DataSpark data.

Final Judges
Mr Shaowei Ying, COO, DataSpark
Dr Ying Li, CTO, DataSpark
Dr Yunye Jin, Senior Data Scientist, DataSpark
Mr Edgar Hardless, CEO, Innov8

Final Judging via Live Demo - 6 Nov ’17


Manuscripts (a proposal in Phase 1, and a solution report in Phase 2) should be PDF-formatted using the ACM camera-ready template, and submitted electronically through the EasyChair online submission system.

Team Size

  • Teams may comprise up to 4 members.
  • All team members must be listed in the EasyChair online submission in Phase 1.
  • Submissions with more than 4 authors are automatically rejected.


If you have any question regarding the DataSpark challenge, please join our Google Group, where we host a Q&A forum. Before you post a new question, please check if your question has been previously answered in the forum.