My main research agenda is developing and applying the computational methods and tools needed to study media effects in a fragmented digital media landscape. As people consume news from an increasing number of traditional and social media sources, the field urgently needs to develop better ways to measure news consumption in order to study the effects of each person’s individual media diet. My research directly contributes to this development by furthering our pioneering work on data donation, where (fully informed and consenting) respondents donate a cleaned and minimized version of their social, mobile and digital news browsing history. A second key methodological part of my research agenda is the introduction, adaptation and validation of methods from AI and computational linguistics into communication science research. Recent years have seen enormous gains in the development of attention models in deep learning, but these are only very slowly being used in the social sciences.