IEEE Computer Society team
Winners have been announced for the 2021 Global Student Challenge (GSC-21)! The first round of the competition started in May and it was about solving student teams two challenge problems. Each team was evaluated on their implementation and reported on their results. This year, 487 students participated in 250 teams representing 195 universities from 41 countries.
Want more technical news? Subscribe to the ComputingEdge newsletter today!
First place: Taicheng Guo, KAUST University, Saudi Arabia
On the second place: Allan Henrique Kamimura, Universidade de Sao Paulo, Brazil
Third place: Chi Zhang, Zihao He and Yue Li, Xi’an Jiaotong University, China
Special Computer Society Award: Fahmi Noor Fiqri, Universitas Pakuan, Indonesia
Audience choice: Bohao Liu, Zhenfei Luo and Lulin Liu, Xi’an Jiaotong University, China
- Taicheng Guo and team [Video] [Slide deck]
- Allan Henrique Kamimura and team [Video] [Slide deck]
- Chi Zhang and team [Video] [Slide deck]
- Fahmi Noor Fiqri and team [Video] [Slide deck]
- Bohao Liu and team [Video] [Slide deck]
Slide deck with program highlights [View]
Words from competition chairman Saurabh Bagchi
The competition aimed to engage the global student organization in solving two real-world data analysis problems. The issues were based on data sets collected from real-world computer systems (Challenge #1) and from Tweets related to the pandemic (Challenge #2). While failures of computer systems are a problem under normal circumstances, the outages or even delays in the middle of the pandemic with remote working that put more strain on such systems have been particularly annoying. Therefore, the challenge problem asked the participants to create data prediction models to predict when such errors were likely. The system administrators can then take measures if the probability of failure is particularly high.
The second challenge problem asked the software to classify people’s emotional state based on their Tweets. With isolation forced upon us by the pandemic, social media posts promised an important way to gauge people’s emotions. This challenge sought to push the boundaries of how accurately software can do that, despite the nuances of expression that depend on culture, nationality, and other factors. And it turned out that the best data analysis software could do this surprisingly well, even if it had less than 140 characters to weave its magic on.
On a less technical level, the competition was a powerful way to bring the global student community together, despite many being isolated from one another, creating a level playing field to some extent. Student teams had the freedom to innovate, learn the skills they lacked to solve the problems, and put their heads together to create entrepreneurial, highly accurate solutions. What was heartwarming was that some of the winning solutions exceeded the accuracy achieved by the original authors’ models. To the organizers, and to all who watched this competition unfold, this was a reminder that human ingenuity is a precious commodity and obeys no boundaries of country or status.
Regarding the participation of the organizers, we had a very effective collaboration between academics (faculty members and researchers at universities) and industrial practitioners. The different parts of their views and the overlapping parts helped bring out different facets of the submissions.
Challenge Problem #1: Computer System Error Data Analysis
With the growing scale of supercomputer systems, scientists are now able to solve challenging computer problems in a matter of seconds, which would take hundreds of years on a personal computer. However, with increasing scale (and complexity) the probability of application failure, whether due to hardware or software errors, increases. Such failed applications not only slow down scientific progress, but also lead to a huge amount of wasted resources, both in terms of time and energy consumption. If we can predict when an application would fail due to a system error or due to software errors, preventive mechanisms such as checkpointing can be initiated to store interim results, reducing the amount of wasted computation.
This challenge problem involves predicting failure of application executions (referred to as “job”) on Purdue’s central computer cluster. The teams get data on the health of each compute node and resource usage by all tasks running on the cluster.
Challenge Issue #2: Sentiment Analysis of COVID-19-Related Tweets
This challenge involves sentiment analysis of Tweets related to the Covid-19 pandemic, a task for classifying text with multiple labels. Since the outbreak of the pandemic, it has affected more than 180 countries, where there have been huge losses in the economy and jobs worldwide and about 58% of the world’s population is confined. Researching people’s feelings is essential to ensure positive mental health outcomes and to keep people informed about Covid-19.
In this competition, the released training data contains 5000 labeled tweets, while the released validation data contains 2500 pieces of unlabeled tweets. The classes for each tweet are: Optimistic (0), Grateful (1), Empathetic (2), Pessimistic (3), Anxious (4), Sad (5), Annoyed (6), Denial (7), Surprise (8 ), Official Report (9), Joking (10). Tweets can be tagged with multiple classes. The participants should automatically label the tweets in the test dataset.
About Global Student Challenge
The Global Student Challenge is a platform for students from all over the world to create innovative solutions to data analytics problems. This 2021 judging panel consisted of IEEE Computer Society volunteers, academics and industry representatives.