HB + HackOverflow

• 

  A screenshot of bot's reply

Inspiration

• Imagine you when you unlock the phone, you are greeted with around 100 unread messages. Wouldn't it be useful if we don't have to read all of them and not missing out from the group conversation? With our bot, you don't have to.

What it does

• analyses sentiment of each message.
• Out of n messages specified by the user, output a filtered list of messages grouped according to positive and negative emotions.

How we built it

• used telegram bot java library to create a working bot
• used nlp library to perform analysis of the messages in the chat

Challenges we ran into

• Problems in using Google Cloud NLP library.
  • Problem in importing the dependencies to our project.
  • Problem in configuring Google Application Credentials for authentication.
• Problem in using telegram bot java library
  • Due to the lack of documentation of the API and our lack of experience in creating telegram bot, it took us a long time to figure the basics of how to use the library.

Accomplishments that we're proud of

• successfully created a working bot that is able to filter out

What we learned

• how to create a simple telegram bot.
• A slight feel of what NLP is and what it can do.

What's next for HackOverflow

Built With

• google-cloud-nlp
• telegram-bot-java-library