One last question that I had, so you did a lot of verbals in storytelling. You already mentioned personality. Now, one of the things that many people have in mind when regarding go to this computational paradigm, computational social science especially, is that basically, what you work is with language. So language traditionally, that's usually the domain of the humanity. It's like literature, and English major, and reading books. That was very qualitative, almost closer to art. Literature closer to art, then to science. Now, literally, you make a science out of it. You take this rich literature and you put it in scientific. You count it. You put it into numbers in order to be able to teach machines. You talked about many things like personality. I know you did a lot of research on storytelling. Storytelling is something that has been with humankind since we set around. Yeah, that's why I'm so interested. I'm really interested in narrative genre, especially informal narrative. So I've had a number of different projects in that area. One of the things to say, in form of genre, we've taken and we use this in our lesson the last year in order to give the system some of its own content. We have these blogs that people posted someplace like LiveJournal where they told stories and we have a lot of those stories. We've had them annotated with the deep story structure. So we know a bit more about what the timeline of the story is and what the events. So we have more than just the surface string of the storytelling. So our lesson last year, we had a component where it was like, do you want to hear a story? Then, we would retell some of those stories that we got from social media. So that was one case where we've used that, but we've done a lot of work on using that deep representation of story, for example, to be able to change a story that's told in third person, to say one that's told in first person that would make the characters come alive. We're using that now in a system that we're building for National Science Foundation for the Cyberlearning Program where we're trying to improve children's oral language and narrative comprehension skills. So we're targeting five to eight year old who aren't getting enough language input at home. Then, the idea would be that there would be a storytelling agent who would say, maybe model the child's ethnicity or gender. So it'd be an animated character on the screen. They would interact with the child to tell the story, and I have the child be able to ask questions about a story, and that's in collaboration with some folks at UC Davis. So Michael and I think computer science, there is a human body animation expert. Emily Solari is in School of Education and she works on reading narrative comprehension. So there's a lot that I think the storytelling interaction is fascinating and you were talking about the idea of social sciences. There's a lot of things that, like a story at the dinner table, stories told every five minutes, or that people use their storytelling as part of forming their identity. Right. You decide what stories you're going to tell and how you're going to tell them, and that's part of your self-presentation. Your personality is the story that you tell about yourself. Yeah, the stories that you tell about. Also, people say that about culture. A culture is defined by the stories that it tells, what the common threads are in those stories that actually defines a culture. So I love that area. So do you think actually, now that you have machines, also telling stories with us, and you do a science, and you understand rather scientifically, what storytelling is? What makes a great storyteller? What's the personality of it? So is the humanities with this computational social science since we now have access to qualitative information can make quantitative science out of it? Is that growing together? You have a master's in linguistics as well. Well, I have collaborators in linguistics. My collaborators here in linguistics are also interested in narrative structure and I've also talked to various people in literature here. So there is a mismatch of method. So most scientists or whatever you want to, most people who work in the literature area are used to much more fine grained analysis of a single story in something that's more qualitative. You're aware of that. That's what you're referring to. Right. The idea that a machine could provide them with any insights on something, it's still pretty foreign I think to most people in literature, but there has been a lot of work in natural language processing recently on just things looking like, take Jane Austen's novels, for example, and look and build a social network out of which characters talk to who and how much they talk to each other, and try to understand what the social network is. Then, there's been some work trying to look at the different ways that a character in a novel might speak when they're speaking to different people depending on who they're talking to. So Elizabeth Bennet, when she talks to her sister, she might speak, or her father, or D'arcy. She might speak completely differently in all those situations. So there's been some, like the work on personality recognition using some of these tools to characterize the differences a character might speak when they're speaking to different people. So then, you use the words to classify their personality which might be. Contextualized. Yes, it's complementary. It gives you a quick over view of. Right, it could. If you really look at the text as well with your own eyes, so you know that person, you might discover some mismatches between or the machines does and what a qualitative [inaudible]. Right, and that's still a very new area, and I think it might be one of those areas where you need that new generation of literary theorist to open to the idea that the computer could provide them with some interesting insights. Yes. To move that on. Well, on that note, I hope we have a new generation of researchers that will also join. You can find Professor Walker online as well as all the interesting work, and her dozens and dozens of projects that she has been and is running in her lab and her team. Well, thank you very much. All right. Thank you very much for being with us. It's fun. Yes, it was fun. That's not either [inaudible]. Right. Thank you. Okay.