Institut Teknologi Bandung

Gemastik (National Students Show in Information and Communication Technology) is a national technological championship that was held by Puspresnas under the Ministry of Education and Culture. The goal is to improve soft skills, technology literacy, data, and human resource. There is 11 divisions that are contested in the Gemastik, one of them is data mining.

Ryandito told us that he has been interested in this competition since the last year Gemastik XIII, especially in the data mining division. Having the same passion on the field, Dito then invited Randy and Fawwaz to be his teammate. They named their team "nasi Goreng".

In the Gemastik XIV data mining category, the participants are required to extract unstructured data, especially sentences in Bahasa Indonesia. The nasiGoreng team utilized Twitter API to collect data on the public sentiment of the vaccination program in Indonesia. They specifically classified which tweets are negative or positive and which tweets are subjective or objective by using POS-Tagging and artificial intelligence.

The team said that they are used to working with tabular or structured data. So, when they started to learn natural language processing (NLP) for Gemastik, it is a whole new experience for them. Randy admitted that he was amazed by NLP because as a real language it is full of context, but it can also be 'measured' by computer.

Team nasiGoreng shared some of their unique experiences. On the qualification and final phase, they are more focused on learning text classifications. But when the case for the final phase was revealed, it was totally different from what they had previously thought. Team nasiGoreng, who does not prioritize this kind of question and answering task, was quite surprised at the time the case was revealed.

However, they did not give up. They keep researching multiple resources that are available on the internet – in Gemastik, participants were allowed to use every resource that they can found. Dito said that his previous data-mining exercise made it easier for him to adapt to new cases. So, what they had prepared was not completely in vain.

Their challenges don't stop there. On their first submission, they had found that their machine learning model wasn't working properly because of the existence of a significant accuracy gap between their internal test and the submission output. They feel overwhelmed by this nuisance, but nasiGoreng said that they were determined to finish what they started.

Their perseverance to keep looking for solutions finally paid off. In the last minutes, around 10-30 minutes before the submission deadline, they managed to find the solution for the accuracy problem. As a result, they come out as one of the teams that managed to build a model with relatively high accuracy.

Accompanying their success, there is this factor of mentorship by their lecturer supervisor who is always supportive of them. Fariska Zakhralativa Ruskanda, S.T., M.T., a lecturer at the ITB Informatics Engineering Study Program. While reporting this, the ITB Public Relations reporter had the opportunity to conduct an interview with her.

When asked about the importance of support by a supervisor, Fariska replied that she believed that ITB students already possess extraordinary cognitive and exploration potential, but they also need to be guided properly. Therefore, a supervisor needs to guide their students to look for the correct resources according to their competition division. In addition, Fariska also believes that a supervisor also plays a role in encouraging their students to keep looking for new experiences and providing moral support to help lift their mental burdens when faced with difficulties.

On the other hand, according to Dito, guidance, and advice from Dr. Fariska helped them to find a solution at the final stage. The guidance from their supervisor also made them realize that there is still a lot that they could learn about, that "NLP is not as simple as it looks". In other words, Dr. Fariska guidance opens up his mind.

From this experience, Randy advised his fellow students to be persistent when faced with unexpected hardship. If at that time nasiGoreng had decided to give up, then it would be impossible for them to be standing on the podium. "Don't ever give up. We need to persevere until the last minute," said Randy.

Meanwhile, Dito advised that we should not be afraid to try something. He said, "Don't be afraid, just do what you want. Everyone, even smart people, starts from ignorance. If we were afraid, we would have lost already."

Finally, Fawwaz added, "Perhaps it is wrong to say that we are not capable of, but we just haven't tried to learn or haven't found the right learning technique. Therefore, you need to find your own learning style, and then just try to learn."

Reporter: Maria Khelli (Informatics Engineering, 2020)
Photo: Interviewee Documentation
Translator: Favian Aldilla R