Sentiment Analysis of Tweets from Twitter

The project investigates the use of a multimodal feature learning approach, using neural network based models such as Skip-gram and Denoising Autoencoders, to address sentiment analysis of micro-blogging content, such as Twitter short messages, that are composed by a short text and, possibly, an image.

A novel architecture that incorporates neural networks is proposed and tested on several standard Twitter datasets, showing that the approach is efficient and obtains good classification results.