Fuzzy sentiment analysis model on social media texts

BACKGROUND

A company has a module to classify messages in social media according to their professional sector and they need a module to detect “dangerous” messages in social media.

Problem

A company has a tool that captures and classifies thousands of messages a day from different types of social networks in relation to their client’s professional activity. To add value to their tool, the company needs a sentiment analysis module to provide their clients with insights about how their product is evolving in the market as well as trends in their industry.

Benefit

Affective Norms for English Words provides a set of normative emotional ratings for a large number of words in the English language and three emotions measures for each term. It is used as a basis for describing a fuzzy model for analyzing opinions which leads to improvement in business activities.

METHODOLOGY & results

Architecture: Cloud architecture based on AWS services and databases, using RedShift.

Developing language: Java.

ML techniques: NLP, opinion mining and fuzzy logic

Results: The module detects the sentiment of each of the messages segmented by professional sector based on a fuzzy classification with an accuracy of 85%.

This project was carried out before AI Shepherds’ foundation by its team members.
View More Projects
Scroll to Top