Oomck-Bot

Getting Started

Dependencies

• Docker
• Python 3

Setup Project

1. Ensure that Docker is currently running on your machine with an up-to-date version.
2. Install python dependencies with: pip install -r requirements.txt
3. Run docker-compose up to start elastic search.
4. In a new terminal, run python main.py.
5. Enjoy!

Test

1. run python -m unittest to run all the tests

Run sockets with other bot

1. Install python dependencies with: pip install -r requirements.txt.
2. Run docker-compose up to start elastic search.
3. Set the host address and port at the top of python sockets/main.py.
4. In a new terminal, run python sockets/main.py when the other bot is running as host.

Features

1. GUI:
   • We implemented a GUI to improve the overall experience for our users.
2. POS Tagging:
   • We implemented POS tagging to assign labels to tokens in the user's input. This helps improve the quality of the bot's responses.
3. Synonym Recognition:
   • We implemented this in our bot to recognize synonyms within the user's input. This improves the bot's understanding of message sent by the user.
4. Interaction with Another Bot via Sockets
   • We connected our bot with another group's chatbot through sockets in order to allow communication between the two bots.
5. Automated Unit Testing:
   • Unit tests were implemented to ensure that the quality of our code is maintained throughout the development process. This ensures that our bot behaves as expected at all times.
6. Handle Spelling Errors:
   • This was implemented to ensure that there are no spelling mistakes in the user's input that could affect the quality of the response returned by the bot.
7. Extra Default Responses:
   • To keep conversation going, there are now more, and on topic, default responses. This keeps the conversation going even when the user is confusing the bot.

Features of caht bot for an API

The following features of our bot could be put in an API for others to use.

1. Easy connection with another bot through sockets.
2. Tokenizing user input with NLTK.
3. Processing and loading of data from a JSON file into an Elastic Search instance.
4. Answer questions regarding the Fast and Furious franchise.
5. The basic structure of our chat-bot allows it to be used with a variety of datasets according to a developer's specific needs.

API calls

The following features of out bot uses APIs to come up with an answer.

1. Wikipedia
   • Start the sentence with "What is the definition of" and the bot will try to find the definition from Wikipedia.
2. Wolframalpha
   • Start the sentence with "Calculate" and the bot will use wolframalpha api to create an embedded image of the result in the chat.
3. Google Translate
   • You can chat with the bot in any language that you want. It will respond in the same language. Don't believe me? Give it a try :)