Natural Language Processing

Natural Language Processing is a machine learning technology that gives computers the ability to interpret manipulate and comprehend human language.

What this repository contains

This repository contains several concepts for Natural Langage Processing and common task solved using different NLP techniques and tools with Deep Learning.

Building Chatbots

Building different kinds of chatbot like rule based chatbot, intent based chatbot, slot filling chatbot, generative chatbot and Modern RAG based chatbot.

• Simple Chatbot: similarity based and rule based

Text Classification

Text classification using different Machine Learning Technique like traditional machine learning algorith, MLP, RNNs and modern Transformer based model like BERT.

• Atis Intent Classification MLP
• Simple Binary Classification Example

LSTM Neural Network Architecture

Using LSTM network for different task e.g language modeling, text classification, question answering, summarization.

• Binary Classification Many to One: Tensorflow LSTM
• English to Roman Nepali Translation Word Level Encoder Decoder: Tensorflow
• English to French Character Level Translation Encoder Decoder:
• Next word prediction LSTM Many to One: Tensorflow
• Next word prediction LSTM 2 Many to One: Tensorflow
• Math Text Addition LSTM Many to Many: Tensorflow

NLP with NLTK library

Library for text processing.

Sequence to Sequence Learning

Sequence to Sequence Learning for different variable length sequential task.

• Sequence to Sequence LSTM for Reverse Number Many to Many: Tensorflow
• Encoder Decoder Architecture for Reverse Number: Tensorflow

Text Representation

Different text representation techniques like bow, embedding, one hot encoding, tfidf, etc.

• Bag of Words
• N-Grams
• One Hot Encoding
• Tf-Idf
• Word Embedding: Word2Vec