Sentiment and media analysis in Twitter

Twitter sentiment analysis allows you to keep track of what’s being said about your product or service on social media, and can help you detect angry customers or positive mentions.

Comment

Inspiration

I wanted to build solution that help developers to use AI services and automate the way team can receive insights from social media in teams chats like Microsoft Teams and Slack using Postman Collections and Monitors.

What it does

  • Retrieves 100 recent tweets along with images based on provided query.
  • Describes what is on the images using Microsoft Azure Computer Vision service.
  • Detects in what language these tweets are written using Microsoft Azure Language Detection service.
  • Retrieves sentiment of tweets using Microsoft Azure Text Analysis service.
  • Combines result in a report.
  • Sends report to Slack channel via webhook.
  • Sends report to Microsoft Teams channel via webhook connector.

How we built it

Using Postman Collections, Pre-request scripts, Tests, Workflows, and Postman Monitor.

Challenges we ran into

  • At one point I wanted to use Postman Visualization of API Data ( https://www.postman.com/api-visualizer/ ) for visualization, but it seems that it's not supported when running collections.
  • Some of the API require multiple calls to process the data. That's why I've used a combination of pre-request scripts and tests with workflows to mitigate that.
  • When using free tiers of Microsoft Azure Cognitive Services some of the requests need to be throttled.

Accomplishments that we're proud of

  • Fully automated twitter sentiment reports in Microsoft Teams and Slack

What we learned

  • Postman Collections
  • Pre-request scripts
  • Tests
  • Workflows
  • Postman Monitor

Built With

Share this project:

Updates