Towards Emotional Authenticity in News Presentation: A Machine Learning Approach

In the contemporary media landscape, maintaining emotional authenticity is essential for television presenters to establish trust and forge meaningful connections with their audience. This study introduces an innovative machine learning model that utilizes advanced facial recognition techniques to enhance media communication. Unlike existing methods, our approach specifically identifies television news presenters within broadcasts and analyzes their facial expressions to assess their emotional authenticity. By focusing on the congruence between the presenters' emotions and the content of the news, our model aims to ensure that the emotional delivery aligns with the tone of the news. Preliminary results indicate that our model not only accurately identifies presenters and interprets their emotions but also offers novel insights into improving emotional congruence in television news. This research highlights the potential of machine learning to significantly enhance the quality of media communication by ensuring emotional consistency, thereby enhancing viewer trust. Tests have shown a notable improvement in the accuracy of matching expressed emotions with the expected emotions according to the context of the report, underscoring the effectiveness of our approach.