You are working on a text classification problem where short customer comments need to be labeled by sentiment. The messages are informal, often include typos, emojis, and mixed opinions in the same sentence. You have a labeled dataset from past reviews, and the main challenge is turning noisy text into a reliable classifier that can separate positive, negative, and neutral feedback.
How do you discuss a project where you had to classify text or sentiment?
