Bert based Question Answer Bot (NLP)
This is the project I created for Lincode's Virtual Hackathon 2020.
Features of the Model:
-
π€ Can find precise answers from large paragraphs.
-
π Very Fast Responce Time.
-
π Use of current State of the Art BERT model to Increase preciseness.
-
βΎ Infinite Quick Question-Answering once the model is set on the paragraph(Data).
-
β³ Can also answer Questions in Real-time without having to train on Data Paragraph using google's Word2Vec model.
-
π Ready to Deploy.
Use Cases:
-
Can be implemented in Frequently Asked Questions(FAQ) section for best user expirience.
-
Study Helper: Just give the model the paragraphs you want to extract information from and just start asking questions!
-
Chat-bot: with some tweaks the model can be transformed into a Personalised Chat-bot.
Tech-Stack:
-
python
-
Pandas
-
Numpy
-
Tensorflow
-
Word2Vec
-
HuggingFace NLP
-
NLTK
More-info in README.docx
Thank you! π€