Skip to content

carolineapp/what-is-it-watson

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

42 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Media Roundup v1.0

Check out my app here https://media-roundup.herokuapp.com/

Analysing the news with Artificial Intelligence

APIs used - IBM's Watson and News

Why?

Recently, it came to my attention that one could play around with IBM's Watson, an Artificial Intelligence platform, which I think is pretty neat 😎

How?

I decided to utilise its Natural Language Understanding functionailty. This API takes either a URL or plain text and then returns the following features:

  • Emotions (joy, disgust, anger, fear, sadness) as a numerical value
  • Sentiments (Positive or Negative)
  • Entities (People or Places)
  • Concepts
  • Keywords

In my app I have decided to use the emotions feature (for now!).

For the content, I have used the News API which takes a search term (user input) and a source (1 of 4 options). This returns the most recent articles relating to this topic from which I choose the first.

N.B This returns the most recent articles containing the search term, so may not be 100% relevant. However this is supposed to be lightheated only and not a true, balanced representation of the topic!

User Journey

As a user I can;

  • Enter a topic of my choosing
  • Choose a news source (Daily Mail, Guardian, Reddit and Buzzfeed - more to be added)
  • Get back an emotion (joy, fear etc)
  • Decide whether you agree with that or not!

Planned Additional Features

  • Twitter. Analyse tweets on a given topic, due when Twitter decide to give me an API key
  • More news sources from News API
  • Add a Postgres database to add URLs and topics
    • Can then plot trends over time by search term and/or source
    • Users can up/down vote as to whether they agree with the Watson analysis of the article

About

Analysing the news with News and Watson APIs

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors