This project is a Kashubian translator developed as a research project at Gdansk University of Technology. The aim of the project is to create a machine translation system for the Kashubian language, leveraging machine learning techniques, specifically neural networks. Additionally, we aim to develop a web application that enables to translate sentences to Kashubian language easily.
The motivation behind this project stems from the challenge of creating a translator for under-represented languages, such as Kashubian, which have limited available data on the internet. By addressing this challenge, we aim to provide a solution that can be applied to other minor languages facing similar issues, serving as an example for future projects in this domain. Moreover, we aim to facilitate communication and cultural exchange by providing a platform for translation between Kashubian and other languages, initially focusing on Polish and potentially expanding to English as there is a significant number of Kashubian minorities for example in Canada.
We are adopting a machine learning approach to machine translation, where we will develop a neural network model trained on the available Kashubian language data. Through the utilization of neural networks, we aim to overcome the limitations posed by the low amound of resources for under-represented languages. Concurrently, we are developing a web application that will serve as an accessible platform for translating sentences between Polish and Kashubian languages. Future considerations include expanding the translation capabilities to include English-Kashubian translation.
- Develop a neural network-based machine translation system for the Kashubian language.
- Address the challenge of low resource data for under-represented languages.
- Create a user-friendly web application for translating sentences between Polish and Kashubian languages.
- Consider expanding the translation capabilities to include English-Kashubian translation to cater to Kashubian minorities in Canada.
Contributions to this project are welcome, especially when it comes to providing additional data for training the translator.