AI Assisted Journalism

Access to heroku app :

https://auto-nrv-cntr.herokuapp.com/

Overview

It takes a village to do structured journalism. Coordinating among so many disparate activities is a serious barrier. We believe that, if done right, smart automation tools can help. Our aim is to use intelligent automation to coordinate developer activity, content production, and content presentation into a working structured journalism ecosystem. which will accelerate the process in getting news published since it requires multiple level support and multiple actors involved. We also integrated fuctionality to create on the go transcript of interview using audio to text conversion method. Creation of rich content for specific audience is something untouched in majority of newsroom and we tried to present it.

We used Node Red as our platform to replicate and orchestrate different services to have an end-to-end workflow of Newsroom.

1. The application is IoT based and can be used to include further advancements.
2. Application will outline actors of Newsroom in form of Finite State Machines.
3. Use Azure cognitive services to have transcript ready on the go.
4. Use sentiment analysis to provide multi feature story around a topic.
5. Application will create story content ready for consumption and publishing within minutes.
6. Seamless communication between actors, Sponsor, Editor, Journalist and Publisher, inside newsroom.
7. Production ready and distribution ready application through Docker

Tools – Technologies Used

1. Node Red
2. IBM Watson
3. Finite State Machines
4. Sentiment Analysis
5. Docker
6. CI/CD
7. MS Teams
8. Python
9. Angular/JS
10. HTML CSS
11. Mongo DB
12. JSON