/data-mining-proj

A text classification task using google's bert

Primary LanguagePython

data-mining-proj

Baoqi Yu, Bochao Wang, Chenxing Zheng

A text classification task.

Sentence embedding: TF-IDF weighted average of glove word vectors / Google's bert sentence embedding

Classifier: simple 2-fc-layer neural network

Result: 72% dev. accuracy using my simple NN && 85% accuracy using bert's built-in classification model.