Sentiments (positive, negative, neutral), emotions (joy, anger, sadness, fear, disgust, and surprise) and values (power, achievement, self-direction, stimulation, benevolence, universalism,hedonism, security, conformity, tradition) expressed in text may convey important information for policymakers.

By applying such analysis to news (mainstream or unverified), political speeches and debates and social media conversations, we aim provide to Policy Analysts a near-real time summary of topics, narratives and viewpoints from a large set of stakeholders around given policy issues.

Activities include:

• Development of two sets of sentiment and emotion detection and classification system exploiting machine learning and AI
• Implementation of these automated analysis modules in Europe Media Monitor (EMM) and the Conference on the Future of Europe (CoFE) IT systems. 
• Media Sentiment and Emotion analysis newsletters focusing on Covid-19 related issues. 
• Support activity in case of requests coming from other units or DGs. 
• Applied Research related to value, emotion and sentiment detection.