The research project aimed to develop a natural language interface system called Lydia that can be easily extended during a conversation between a user and a computer. This system can learn new concepts and rules by being told in natural language and can understand the meaning of new words based on their context in a sentence. The goal was to make the interface more flexible and user-friendly, reducing downtime and frustration for new users. The key methods used were learning by being told and inferring word categories from linguistic context. The findings suggest that this approach could improve the adaptability and usability of natural language understanding systems in various applications.