We have done sentiment analysis as a part of Minor Project -
- text dataset with Fastext, Elmo and Bert embedding, and trained with CNN,RNN,RMDL,HAN and Multi-channel CNN and RNN.
- image dataset sentiment analysis with CNN.
https://www.kaggle.com/jiashenliu/515k-hotel-reviews-data-in-europe (text reviews) https://data.world/crowdflower/image-sentiment-polarity (image reviews)
Here is the colab link of all code-
-
Fasttext - https://drive.google.com/file/d/1M67TUE9OAGZOiPCToDP-2htiaWpwtaOg/view?usp=sharing
-
wordEmbedding + CNN - https://drive.google.com/file/d/1PUO-BMeWw8pQ39OpMDg6H8xpw7_gti7g/view?usp=sharing
-
wordEmbedding + RNN - https://drive.google.com/file/d/1PUO-BMeWw8pQ39OpMDg6H8xpw7_gti7g/view?usp=sharing
-
Elmo + CNN - https://drive.google.com/file/d/1SGElAzr--LelRL0v8rjUzF5ESYYxx9_j/view?usp=sharing
-
Elmo + RNN - https://drive.google.com/file/d/1SGElAzr--LelRL0v8rjUzF5ESYYxx9_j/view?usp=sharing
-
Elmo + RMDL- https://drive.google.com/file/d/1knz1PFJFLhoarVCTxC5MJnCI3AoIlYqe/view?usp=sharing
-
Bert + CNN - https://drive.google.com/file/d/146KXumr0lXjvAFcV8ogCYc_URUPR0Nqd/view?usp=sharing
-
Bert + RNN - https://drive.google.com/file/d/1W2U5PPHZo9yXBEbu8wwbbapUCxx4f9US/view?usp=sharing
-
Bert + RMDL - https://drive.google.com/file/d/1Zmv6TIcSe_qSgHJlXwDkX74gUx_HMHqC/view?usp=sharing
-
Bert + Multi-channel CNN - https://drive.google.com/file/d/1W_rNyR4CFWoqkWNKyiSXtv8e8_L55OGD/view?usp=sharing
- Elmo + Multi-channel CNN - https://drive.google.com/file/d/1HOCEdD4UszgnC5C7vs5E1iJjG5yeWGrV/view?usp=sharing
- Elmo + HAN - https://drive.google.com/file/d/1powJSaC4nDhkEWBDFeObJoyJz1lx_bkI/view?usp=sharing
https://docs.google.com/document/d/1ygtTUgWqI61iMYbUV3zfLPHNA92sIefFnjtltoN0sFo/edit?usp=sharing
Reference Code - https://drive.google.com/file/d/1G5GeZNSXbGlu_f_wPXtbQemz7EIbSZI3/view?usp=sharing