ARTICLE Romero_RCS_2021/IDIAP Topic analysis and tracking from Mexico's President daily press briefing Romero, Luis Armando Arias Ramírez-de-la-Rosa, Gabriela Villatoro-Tello, Esaú EXTERNAL https://publications.idiap.ch/attachments/papers/2021/Romero_RCS_2021.pdf PUBLIC Journal of Research in Computing Science 2021 Language is a very useful tool to politicians. Several areas such as linguistics, sociology and communication consider important the study of political discourse analysis. In this paper we present a method for topic detection using LDA in 534 daily press briefing of the Mexico's president: Andrés Manuel López Obrador. Subsequently, through a web system, which we call DICTA, a person can quickly view the topics discussed in these briefing. Through experimental evaluation we found the highest coherence value when using 18 topics