So as Rene already told you about, we are interested in the relationship between events that are happening in the real world and certain news frames that follow these events. So as you can see on this slide, for example, we recorded spikes at certain news themes. Themes, for example, could capture anything from vaccination to terror to immigration. Now, if you look at this slide, you can see a huge spike in the terror theme around March 23rd which coincidentally happened when the Brazil terror attack happened. We see similar spike in terror themes on June 13 when there was the Orlando nightclub shooting. Now, look at the theme of immigration for example, we can see that news coverage, news outlets talk more about immigration when there was the Brexit. Incidentally also, this theme got a bigger trajectory during the US election. The green graph shows the theme cyber attack. Now, whenever you look at the spikes in this theme, this doesn't necessarily mean that there actually was cyber attack. It rather means that during this period of time, there was an increase in news articles that reported about cyber attack. For instance in this case, you can see that the theme, cyber attack, was especially salient during the statement on Russia's involvement in the US election. There are two other cyber attack spikes that you can observe. One, there was during the second presidential debate which, as you can see, followed the statement on Russia's involvement. So we could maybe make the argument that part of the US presidential debate was the discussion of cyber attacks. Lastly, we can see a big spike in the cyber attack theme whenever Harry Reid call for James Comey's resignation. So with a push of a button, we can now detect the digital footprint, if you want, that news articles leave out there for us to detect, for example, the political climate of a society right now. What are current topics that are highly discussed at a certain point in time in a given society? Maybe it's about terror, maybe it's about immigration, maybe it's about cyber attacks. So I explained to you our procedure how we extract more information from texts. Now, let's talk a little bit about how do we measure moral information that is presented in the brain. How do we measure how all information is decoded and encoded in our brains? So to do this, as you can see on this slide, we set up an fMRI experiment in which participants, while they underwent fMRI scanning, were reading about different moral violations. For example, in this case, you can see that participant red, you see a girl ignoring her father's orders by taking the car after her curfew. Then we have participants simply rate, well, is this extremely morally wrong, is this not so wrong? After doing this, we ask ourselves, "Okay, while people read these so-called moral [inaudible] , their examples for specific moral violations, how do we find out which brain regions were activated by people who read these opinions?" Here again, computational approaches help us to find out which regions corresponded to the certain stimulus, in this case, our moral violation examples. We train a so-called machine learning classifier. This classifier essentially learns, "Okay, here's a brain activity pattern. What was the stimulus that the participant was presented at this specific point in time?" So the machine learns, "Okay. Here's some brain activation, and here's a stimulus." Now, we teach the machine this association across a certain proportion of our data, and once the machine has learned this association, we then test how good does this machine classify new brain activity. In other words, can the machine figure out by just giving the machine brain activity, which stimulus someone was showed? If this classification is above chance, so above 50 percent, then we say this was a successful classification. Now, I know I kept you all waiting for this answer. Still, where is small information represented in the brain? What we did is we used what's called a searchlight to look for certain regions of interests, or ROIs, R-O-Is in the brain that our classifier classified correctly as morally relevant. They were associated with a certain moral stimulus. Now, if you look at this slide, you can see for example, that the precuneus and the medial prefrontal cortex were especially likely to be activated whenever someone was reading a moral violation. On the right side, you can also see that there was some activation of the insula and in the temporal lobe. Now, we did one last analysis. We can group moral values into a category that focuses on group cohesion, hierarchy structures, and cleanliness, and these moral values are called binding moral domains. There are also the individualizing moral domains, which is care and fairness. They stress the importance about individual freedom. We know from the literature, for example, that conservatives and liberals value these two different categories differently strong. So we know that conservatives value binding domains more strongly, and that liberals value the individualizing domains strongly. So it's really interesting, if you think about it, that conservatives and liberals have these differences in what they perceive as important moral values. Conservatives value the binding domains more. What does binding mean? Well, over evolutionary history, it was really important for us that we stay together in a group. So these moral values enforce group cohesion. They enforce that we stick to our tradition, to our hierarchy structures, but that we also stick to cleanliness, for example, to avoid the infection with pathogens. So what about the moral values of liberals? Well, it turns out that liberals value the individualizing domains of caring, and fairness, which primarily are about upholding the freedom of the individual, that we care for one another, that we exchange and we engage in fair exchanges. It turns out that liberals value these domains way more than the binding domains of loyalty, authority, and purity. Now that we know about the differences that underlie conservatives and liberals in terms of their moral value systems, we can look if these differences also become apparent at the brain level. Can we see that the binding domains are activated under different patterns compared to the individualizing domains? We actually found out, as you can see on this picture, that the binding moral domains are more strongly activated in the medial prefrontal cortex and the temporal parietal junction. This is really helpful and interesting for us because we can, on the long run maybe, use this information to better understand ourselves essentially. If conservatives rely on different moral values and liberals rely on different moral values, well, maybe we can change the way we communicate in such a way that we all understand each other better. Then we will see a less polarized political system. How can we frame messages, for example, that they appeal to certain groups that might have a conservative ideology? How can we frame messages to people that have a liberal ideology? Finally, I'd like to share some of our future ideas that we have for this lab with you. So for example, you have learned about MoNA. Of course, we want to fine-tune and extend MoNA. We want MoNA not just to code moral information in newspaper articles, movie scripts, but maybe even in political speeches, or in books, in novels somewhat. Further in the future, we want to extend our MoNA system to not only just look at film scripts or narratives, but also books. We also want to learn about the relationship between events in news frame and the bigger picture. Lastly, we're interested, how does the brain come in here? How does the brain explain how we process information at a certain level? How can we use computational tools to answer these kinds of questions? So learning about computational methods can seem really intimidating, but trust me, I started coding a year and a half ago and so far, I've learned a lot, and it's been really really fun. You think of problems in new ways. You can tackle amazing questions in different ways. It's really, really fun. With some effort, you can definitely change a lot. So I hope you enjoyed our little introduction into some work of our lab. If you want to find out more about our research, why don't you go online and find us in the Internet. Our website is medianeuroscience.org. As Freddy said, some of the things we do here sound difficult. I'm not saying they're not difficult, they are difficult, but you can do it, you can learn it. Who knows? If you're interested about this research and you find it relevant and important, someday you maybe a graduate student here in our lab. So I encourage you to go to our website, and find out more, and think about a career in communication, media psychology, and cognitive neuroscience. That's what's the media neuroscience lab bundles together as a new, and our perspective, exciting new research..