After detecting news of interest, which in our case means political news, we have to identify the key elements appearing in these texts, also known as "Named Entities". In this workshop we will present an approach to train a Named Entity Recognition (NER) system to discover different entities in the texts such as the "Action" (i.e. what is being done), "Subject" (people, organizations, etc... performing the action, or being a subject of the action), "Claim" (i.e. "The reason/motivation related to the action") appearing in the analyzed news. During the workshop we will also discuss other entities of interest that the AI system can learn to discover in the texts.