How to create a simple information bot using Tensorflow

Tensorflow

Tensorflow is a Google-powered open-source machine learning libraryfor high-performance numerical computations.

To get into the basics of Tensorflow, go to https://www.tensorflow.org/tutorials/.

This tutorial is all about creating a very simple information bot using Tensorflow that will classify the intent of user from the pre-defined intents and provide the corresponding information to the user.

Follow these steps to create this simple bot -

1. Clone the github repository from https://github.com/prayas26/Simple-Chatbot-Using-Tensorflow
2. Install the requirements using the command pip install -r requirements.txt.
3. Open the intents.json file and create a dictionary with the intents (an example is already present in the file).

• Tag - it is the intent of the information
• Patterns - sentence patterns for our neural network text classifiers
• Responses - it is the answers to the corresponding intent
4. Now open terminal and enter python and hit enter.
5. Enter the following lines of code -
• import nltk
• nltk.download('punkt')
6. Run the command - python bot.py
   You can see your bot being trained. After the training is complete, you can see many new files added to your folder.
7. Now, just run the command - python response_bot.py
8. Enter any question with the defined intents and check the reply.

Other posts - Creating a SlackBot in Python using Flask, Slack App, AWS Account & MySQL Workbench

Author - Prayas Mittal

12:59