Meet the winners of the GIZ English-Kinyarwanda Text Cleaning Hackathon
Meet the winners · 8 Dec 2022, 08:37 · 5 mins read ·
4

Meet Team KF2R, the 2nd-place winners of the Kinyarwanda Text Cleaning and Augmentation Hackathon and Competition, a physical hackathon run in partnership with GIZ and the Digital Transformation Center Rwanda (DTC). The challenge attracted 125 participants from Rwanda, all vying for a $3 500 prize pool.

The objective of this challenge was to create a text-cleaning script for Kinyarwanda. The participants were tasked with finding and fixing issues in the provided parallel corpus consisting of English-Kinyarwanda sentence pairs, and finding additional data to clean through other means like data augmentation or web scraping.

Team KF2R - Happiness Rose Mary, Rwanda

Please introduce yourself.

My name is Rose Mary, an enthusiastic data and research analyst with an interest in business, health care, agriculture, and education. I am also interested in using and learning new innovative technologies by applying research, data analytics and business skills to solve real-life problems.

Tell us a bit about your solution, and the approach you took.

We first explored our data using both Excel and Python, to know the nature of the sentences we had. After exploring the data we started preprocessing, we removed all non-useful data like numbers, single characters, and stop words for both English and Kinyarwanda. For every task we performed, we had a function to it and all functions were reproducible.

What set your winning solution apart from others?

What set our winning solution apart from others was using the judging criteria as our guide and following the rules of the competition. We also did our best to make sure that no useful data was lost while cleaning. In addition, we collaborated as a team and ensured our solution ran in a short time.

How do you prepare for a challenge?

I first read carefully the instructions and note down all the tasks to be performed. After reading the instructions, I reach out to my colleagues to find someone to team up with. Then I collaborate with the team to achieve the common goal of winning the challenge.

Do you have any words of encouragement for others or advice that has helped you?

I would like to tell my fellow data scientist that they should not underestimate themselves, but rather always challenge themselves because that's what makes you grow. Also, making teams with other data scientists will help you not only learn something new but also increase your chances of winning.

What do you like about Zindi?

Zindi provides a platform where you can collaborate with other data scientists to grow both personally and professionally.

Team KF2R - Kefas R. Galadima

Please introduce yourself.

My name is Kefas R. Galadima; I hold a Bachelor of Science Degree in Computer Science and am currently enrolled for a Master of Science Degree in Data Science at the African Centre of Excellence in Data Science (ACE-DS), University of Rwanda. I am a researcher and a member of Mbaza NLP Community and Nigeria Computer Society (NCS). I am self-motivated, goal-oriented, a team player, and passionate about problem-solving, writing, programming, and analytical thinking.

Tell us a bit about your solution, and the approach you took.

Understanding the problem and adhering strictly to the instructions stipulated in the challenge was what drove our success. Additionally, in our solutions, we took the following steps:

• Proper structuring and documentation of our solutions using comments and titles

• Using OOP and proper naming convention of reusable functions, and

• Proper naming of data frames and variables for easy readability.

What set your winning solution apart from others?

I think what ret our winning solution apart is making sure we strictly followed the steps involved in NLP data preprocessing, such as data exploration using tools like Microsoft Excel and Jupyter Notebook to assess the quality of the dataset; normalising the dataset to have a uniform casing; and checking and removing unwanted characters, stopwords, emails, amongst other techniques to have a cleaned dataset.

How do you prepare for a challenge?

Getting acquainted with the expectations of a challenge by using online resources, and seeking the counsel of colleagues is the first step to take. Secondly, undertaking the intensive practice of the skills required and getting a proper evaluation from others helps me to identify my flaws and rectify them on the go. These measures have always enhanced my performance in any form of assessment.

Words of encouragement for others, or advice that has helped you?

Be open to challenges as it is the best way to quickly learn, create connections, and put what you have learned into practice in solving global challenges. Also, always seek advice and guidance from fellow professionals to better understand a problem.

What do you like about Zindi?

What I like about Zindi is that it provides a great platform to learn, collaborate and connect with experts, further bringing opportunities to professionals and connecting them to industries.

Team KF2R - Byakunda Roger

Please introduce yourself.

My name is Byakunda Roger, I am a student at the ACE-DS at the University of Rwanda

Tell us a bit about your solution, and the approach you took.

We did Exploratory Data Analysis, and then Text preprocessing steps by following the rubric

approach:

- Team discussion to understand the given dataset, especially Kinyarwanda

- Task allocation to team members

- Feedback on the given task

- Discussion on the challenge and what’s next

What set your winning solution apart from others?

We followed the scoring rubric by specifically creating the functions that accommodated the preprocessing steps.

How do you prepare for a challenge?

Gathering available materials related to the topic, by looking for efficient ways to handle the problem.

Words of encouragement for others, or advice that has helped you?

Great teamwork will enable you to achieve set objectives and enhance knowledge sharing.

What do you like about Zindi?

Through Zindi, I am connected with people and current trends in the field and motivated to find solutions for the community.

Back to top
If you enjoyed this content upvote this article to show your support
Discussion 0 answers