Project Overview

Objective

This data science collaboration team’s objective is to leverage data, analytics, data visualization, and machine learning techniques to implement and enhance the experience on Project CUBS’ main deliverable: a language learning platform. We aimed to collect and analyze user data to identify incorrect answer types based on language region, provide customized curricula based on language acquisition abilities, recommend individualized learning activities, and visualize the collected user data through dashboards. Additionally, our team developed a machine learning model for handwriting recognition, created an algorithm to score user inputs, and connected learning suggestions to the learning scores for improved learning abilities.

<aside> 💫 Team Goal(s):

• This DS team aimed to develop a language machine-learning algorithm for teaching Korean as the initial language on the platform.
• Continuously evaluate and improve the performance of the language machine-learning algorithms based on user feedback and data analysis.
• Collaborated with other teams to seamlessly integrate developed machine-learning algorithms into the platform and support the overall objective of creating a differentiated language-learning experience. </aside>

About Client

This project aims to produce a differentiated language-learning platform for MyKstar. An up-and-rising startup focused on revolutionizing online virtual language learning. With a commitment to expanding and enriching the language learning experience. We are collaborating to incorporate innovative technology features such as karaoke and machine learning handwriting recognition features to enhance learners’ engagement and overall progress.

Collaboration

A Graduate-Student group team at the University of Colombia provided us with extensive research for enriching and improving our collaboration project. On a yearly basis, they specifically select a startup company to provide their consultation expertise. For this project, MyKstar was the selected client to collaborate with and provide the respective deliverables.

<aside> 💡 Detail:

Columbia 292 Consultation Team

</aside>

2023 Spring

(Jan 9th - May 7th)

<aside> 🎓 In collaboration with:

Columbia 292 Consultation Team

</aside>

Topic Highlights

Crawling

Selenium

OpenCV

Machine Learning (CNN)

NLP

Deliverables

OCR - Optical Character Recognition

Selenium Driver

MyKstar: Platform Application

Untitled

Next Steps

• User Data Collection and Analysis: Gather and analyze user data to gain further insights into learner’s behavior and language acquisition patterns to improve the platform.
• Language Machine-Learning Algorithm Development: Continue to work towards improving algorithm accuracy upon successful deployment on using Korean as the initial language. Finetune language machine-learning algorithm and handwriting recognition model accuracy to enhance application user experience.
• Implement Personalized Curricula: Leverage collected user data and language acquisition habits to create customized curricula tailored to individual learners.
• Communication with Client: Maintain open communication with the client to align the data science objectives with overall vision of the differentiated language learning platform.
• Development and Improvement Documentation: Thoroughly document process of data science and algorithm development to share findings and insights gained with client for continuous improvement and project success.