Sentiment Analysis Plus

Sferanet logo

Solutions Gallery > Sentiment Analysis Plus

Solution overview

Web reputation is becoming more important. The quantity of data that has to be analyzed is enormous; hence a big data solution is needed. Sentiment Analysis is the process of automatically categorizing opinions within a piece of text such as a tweet (i.e. positive, negative, neutral). An Artificial Neural Network (ANN) uses different layers of neurons to perform learning tasks. Depending on the number of layers, the ANN can be said to be shallow or deep. A Recurrent Neural Network is a network where connections between neurons form a directed cycle. 

Our data scientists developed a model based on Recurrent Neural Networks with a Long short-term memory layer. The model has been presented at the International Summer School of Deep Learning in Bilbao, 17-21 July 2017. The solution uses flume to connect to Twitter or any other mainstream social network. The content is aggregated and sent to Spark Streaming. Then the data is sent to Keras and TensorFlow, where opinions are produced. Successively, opinions are sent via flume to a Spring Boot back end. Finally, opinions are shown in a dashboard built on Angular.


We provide a unique combination of the latest Deep Learning algorithms and the best big data software.

Key highlights

Customer Analytics

SFERANET was born by the merger of two company "Sfera"​ and "Puntonet"​. Our Mission is to combine knowledge and technology to get the best solutions for our clients. 

Main areas:

  • Big Data with special focus on Cloudera Ecosystem
  • Software Development
  • GIS
  • Infrastructure (HW-SW)
  • Deep Learning Lab
  • CyberSecurity

Positive Business Outcomes

  • One point to keep track of many different social networks

Metrics and Proof Points

  • Less time required to react to sentiment changes

Sentiment Analysis Plus: Web reputation made easy

Learn more about the solution from our partner

Your form submission has failed.

This may have been caused by one of the following:

  • Your request timed out
  • A plugin/browser extension blocked the submission. If you have an ad blocking plugin please disable it and close this message to reload the page.