LashaBuxo / SentEmo-Analyzer

Library which analyzes texts by emotion

Welcome to Sent-Emo Open Source Project!

Sent-Emo Analyzer is open source library, which analyses given English text by emotion and its sentiment values. Below are the example projects and the materials used in project.

How it Works?

using SentEmo;

SentEmoAnalyzer analyzer = new SentEmoAnalyzer();
analyzer.Initialize();
EmotionState result = analyzer.DoAnalysis("I am happy that you came, Thanks for that.");
Console.WriteLine(result.ToJson());

And result looks like:

{
  "Sentiment": 0.7,
  "Anger": 0.06,
  "Disgust": 0.11,
  "Fear": 0.09,
  "Joy": 0.47,
  "Sadness": 0.06,
  "Inclusion": 0.13
}

Example Projects

Based on the sent-emo library there is 3 project, which aims to show the sent-emo library in action:'

• 1st project - completed.
• 2nd project - soon...
• 3rd project - soon...

Example Project 3: Sent-Emo Chrome Extension

empty

Example Project 2: Sent-Emo API - localhost

empty

Example Project 1: Multi-Language Translator

In spite of a fact, that library only works for english text, in this example there is Multi-Language analyzer which uses Google Cloud Translator API for retrieving corresponding traslation.

Note: The Quality of Sent-Emo library evaluation decreases for non-english text scoring, but still reflects "correct evaluation". For more information about see last section

How accurate are the scores?

empty

3rd Party Resources

• Apache OpenNLP - We use OpenNLP library for dividing each sentence, by speech parts and its semantic values
• English words data organized with emotion score by saif.mohammad