Text Mining and Classifications in Yunduoduo
Date:
A natural language processing project applying BERT for sentiment analysis on social media posts.
Key Contributions
- Collected data from “Yunduoduo”, a social platform for college students
- Fine-tuned the BERT model to predict post sentiment
- Compared various machine learning algorithms (Random Forest, Logistic Regression, Linear SVM) for accuracy and efficiency
- Proposed an automatic classification application prototype