ChatBot

From my interest in Natural language processing I intend to make a chatbot using python and NLTK

Motivation

In this modern era digital marketing and onlile shopping is increasing day by day.Real time customer support is a growing need for all type of buisness holder. From this my interset grwon and wanted to make a chatbot.

BlogPost

For detailed overview, here is the accompanying bloga link

Pre-requisites

NLTK(Natural Language Toolkit)

Natural Language Processing with Python provides a practical introduction to programming for language processing.

For platform-specific instructions, read here

Project files short description

Data

For making a chatbot it's necessary to have a large dataset according to ones domain. I had my own dataset where one column shows the queries and another coulmn displays correspondent intents. Another data file contains intents and theie replies.

Data Preprocessing

data_loader.py

This file contains all sorts of data preproceesing.For starters:Tokenization,stemming.Bag of words,lower casing.

Models

I tried many models to get better accuracy for better results.For instance:

base_model.py

cnn_bigru.py

cnn_bilstm.py

cnn_lstm.py

tflearn.py

To call them I used factory method

factory.py

Train

For training and prediction and results:

training.py

Configurartion

For configuring which model,batch size and path to use upadte config.py file:

class Config:

    @staticmethod
    def get_model_name() -> str:
        return "CnnBilstm"

    @staticmethod
    def get_batch_size() -> int:
        return 16

    @staticmethod
    def get_epochs() -> int:
        return 200

    @staticmethod
    def get_dataset_path() -> str:
        return "../../data/All_intent.csv"

**Deployment **

Run from terminal

python deploy.py