App that Analyzes website/doc text and returns sentiment

December 1, 2018

Hi everyone!

I've been working on a small side project that attempts to analyze the sentiment of a website's/document's text and returns that text's total sentiment. I've recently added a tool to help increase/decrease the score of that text.

Give it a try and let me know what you think!!!
http://www.sentimentors.com/

The Website is not yet mobile friendly and works best on laptop/desktops.

Kobe David Raypole

Say something nice to KDRaypole…

Post Comment

1

Interesting idea, I have a similar one sitting on my bootstrapping ideas trello board! Some excellent NLP tooling in this space, in particular in Python and Java. I'm curious what tools you're using?

I ran your service on a news article for the death of George Bush senior and it came up neutral, and some of the negative words were definitely not used in a negative context. This context is what makes things so challenging.

Do you filter out secondary content on pages, like header, footer, etc?

Jeremy

·
8 years ago
·
Reply
1. 1
  
  I'm a ruby developer so I'm using custom built tooling based off a sentiment gem.
  
  A couple of things about results:
  
  some of the negative words were definitely not used in a negative context
  The tooling does not take total context into mind, only the values associated with that word based on a dataset of ~150,000 words each ranked between -1 and 1. Analyzing total sentiment of each sentence would be very challenging.
  Do you filter out secondary content on pages, like header, footer, etc?
  It's not possible to determine the target content of a page based on a the url. So no, but the secondary content doesn't have too much bearing on the total score.
  
  Thanks for the feedback and questions!
  
  Kobe David Raypole
  
  ·
  8 years ago
  ·
  Reply