Web Sentiment Analysis deals with the auromatic identification of the sentiment/opinion of social web content. The available plethora of social web text allows the application of natural language processing techniques for the classification of a piece of social text into sentiment categories (e.g. positive, negative, objective, subjective etc.). HILab has been working on mining the opinion of the web, especially for performing Political Sentiment Analysis.

Politicians, analysts and campaign designers have been focusing their strategy lately more and more on the sentiment regarding leaders, political personalities and parties that appears on the social web. HILab has performed a study to

• identify the web sentiment prior to and after the Greek parliamentary elections of May 6th 2012 by analyzing Modern Greek tweets posted a few days before and a few days after the elections, and to
• investigate the alignment between actual and web sentiment, and how these affect/are affected by a major political event like the elections.

The datasets are available for experimentation: Parties before the Elections, Parties after the Elections, Leaders before the Elections, Leaders after the Elections.