 Thank you. I have my co-author Jesse parent is in the chat and he's going to give you a link to the slides which are available at this link, a tiny URL link and you can follow there. There are a couple more slides actually in that slide deck because I wanted to cut these down for in the interests of time. So this is epistemological directories for research development and education. This is our lab and Jesse and I are the authors. Okay. So this talk is going to introduce a couple of instances of this thing we're calling epistemological directories and they come in different flavors right now. It's something that we've been kind of developing on multiple fronts so you'll see some varieties of that as we go along. But this is also rooted in the evolution of the web and I wanted to bring this up because I'll talk about it in the talk a little bit in terms of like what the antecedents of these directories and then the feature of these directories. So the first part of the web was a web 1.0 and that happened from sort of the beginning of the graphical web browser to about maybe the early to mid-2000s and then involved broadcasted information and things like annotated bibliography. So it was quite static. Then we had web 2.0 which evolved with the advent of social media platforms. So from the early 2000s on to more or less today. So we had social media. We had version controls and other technology that enabled things. And then we're moving towards web 3.0 which involved even more advanced technological innovations like virtual worlds and the semantic web. And so with that framework I want you to think a little bit now about history. What is a historical milestone and what is the relevance of a historical event? So I have these really abstract diagrams but these are supposed to represent events in time. So these events might have qualitative significance and explain a certain amount of the unknown. So it's like some innovation as we were talking about in the last slide or some paper or some person that did something great and that's up to the person interpreting it. But can we do this without subjective judgment? So can we do this in an objective way? But even if we can't does this lead to group consensus or gatekeeping? So who's telling the history and who's presenting it? So there are a lot of common themes across scholars and learners and if you look in the full slide deck you'll see a bunch of these questions. The one I wanted to focus on here though was what are the most important things we need to know in order to be competent? And so this is kind of a digression from the last slide but you'll see how they stitch together in a minute. And so the answer to that is we can use some criterion to say this is what you need to know to be competent in some area. And so this has been addressed by the physicist Leonard Susskind who along with his co-author wrote a book called The theoretical minimum. And so their concern was if you took an informed novice someone who had the capacity to learn like all the you know basically with maybe an undergraduate degree or maybe a high school diploma you know what if they acquired if they had some basic skill level what is the most efficient way to understand the contents of a certain field? So in this case they're interested in teaching someone basically everything you would learn in an undergraduate physics course but to do it in a self-motivated way or a self-educated way. And so we've you know this is this is their answer you you address these different issues if you read the book you'll see they have a couple of key things that they present and it's supposed to give you everything you need to know to be somewhat conversant in physics. Now I work for an organization called the Openworm Foundation and we've observed this very issue. We have people who come in with skills in like you know heavy computer science skills maybe people who come in with mostly biological skills and they need to contribute to this community but they need to understand a little bit of both and so they need to get up to speed in one area or another and this is one approach that kind of used to do this but there's got to be a better way to do this a more systematic way. And so that's kind of where we're going with the epistemological directories. Now epistemological directories have their origins in web 1.0 and so this is back and from like the late 90s early 2000s this is an example of a directory from that era. And so you have it was put together by the Conrad Lorenz Institute and they had this website where you would go and you could do you know learn all sorts of theoretical concepts in these different fields. So you know developmental biology economics cognitive science you could click on one of these buttons and you know it basically presented papers with annotations with some other information but it's very static and you know it's it's it was a predecessor to Wikipedia but it did have this sort of you know centered around an academic field flavor to it. So why don't I introduce one instance of one of these epistemological directories and that is called the knowledge space and so we've implemented the knowledge space in the area of cybernetics and systems and so this is the GitHub repository here and you'll see if you look at the full slides you think you can actually click on it and go to the GitHub repository. So this is a GitHub repository and it has a bunch of folders. These directories are topical stubs of you may be able to see here there are different topics and there are different granularities you have like irreducibility, EGRT, general interactivity. So there are all these things that like as we've come along through this topical area we've thought okay this is a good thing to put a stub in this directory for and so you can see there are different granularities and they're different they may be a little bit vague as to what the topics contain but you click in through to that directory and you could read the readme file and you can get a taste of what those are. Version control encourages edits from the community in general public so hosting it on GitHub we can use version control to bring in people's contributions and then the directories edits and content can all be proposed through pull requests. So it makes it quite a bit very democratic in that way. So the idealized structure you know we have when we go into one of these folders or these directories we can do things like render equations in a very nice manner using markdown we can also incorporate things like Jupyter notebooks and co-lab notebooks so if you're doing machine learning or if you're doing like you know coding you want to include some code in your descriptions or your content you can do that and it's very easy to do in GitHub. So the second instance is called a knowledge map and a knowledge map is a little bit of a departure from the last instance I showed these are graphical representations and so these are timelines and historical summaries of key events so this is why I talked about history a bit and knowledge maps can be used to summarize directories in a knowledge space so we want to summarize things that happen how did a field unfold and what are the key events and then we'll talk about in the third instance epistemological maps and we'll get to that slide in a couple slides but this is going a little bit further in that graphical direction so this is an example of a knowledge map this is a machine learning knowledge map it starts in the 1800s and the granularity is a bit you know you know most of the stuff happens in the last 70 years but that's you know you can add different categories you can add different events different people and that's what makes it dynamic but you can also do something like this have a in the field of artificial life instead of machine learning and here you have a little bit different approach this is where we have a lot of events people experiments software platforms things like that and they're all kind of like you know they look very chaotic here but you can you know go to one of these events and in this case like open problems in artificial life this was a published conference proceeding so you have names here you have a date and then you can find out more about each of those events so one of the thing couple of the things about knowledge maps to keep in mind is that you have a number of milestones per duration of timeline so depending on your pedagogical aim you want to include different types of milestones and different amounts of milestones so you saw with comparing the machine learning map and the artificial life map that the artificial life was much more dense and that's important if you're trying to teach you know maybe a lot more about the field or you know more intimate history or something like that so a small number of milestones may take a minimal amount of information approach but the artificial life map might take a different approach which is to sort of be inclusive in terms of the history and then milestones can stress common themes across learners so learners can propose different miles or different milestones or nodes into this into these timelines and you know they can modify them as they need as they see fit whereas maybe someone who has some knowledge of the background of that field sees fit and you can stress obscure events say that build the foundation for later advances so in the artificial life map we have things like core war and theses which were software programs that weren't maybe very big at the time but they led to advances later on and so you'd want to include those in a map so that those are those are knowledge maps now epistemological maps which is the third instance of this takes us a little bit further this is a again a visualization and this is a theoretical sort of sketch of it and you have two different categories basically of facts one is what do we know so the knowledge maps that I showed you are things that we know this has that has to do with the history of the field but these are things that we know happened in the history of the field well what do we need to know is the other or what we don't know is the other category and so this goes maybe a little bit beyond the history of the field but it could be in the history of the field we don't know maybe some of the history of the field it may be pretty obscure unless you're a historian doing a lot of research on on you know topics like in computing we have people who made contributions and were forgotten about and we'd like to bring them back into the fold and so that's what these little bars are here under what do we need to know so we can include events that were previously obscured or maybe things that we find through experiment or some other research mode and so we can then build a distribution of those facts I talk about it very quantitatively but you can do this in terms of a historiography and you can build a better knowledge map or history of the field and so this is all for education educational purposes but there are a couple of things we need to think about here we need to think about top-down control and the emergence within a community and I think we know we talk about community standards we want to enforce community standards but we also want to allow people to build a knowledge base over time so we don't want to be too over you know heavy-handed with community standards we don't want to use an admin to say wash things that we don't think are important but we also don't want people just to propose things that aren't really critical to the learning experience and so we need to clarify what is significant and so finally I get to the fourth instance which is frontier maps and this is something my co-author has been working on this is an instance of his frontier map and so this incorporates some of the elements of the of the first instance what he's done is he's read kind of reimagine this a bit and he's built a broad taxonomy outside of traditional disciplines so he uses these macro tags techniques discoveries and ideas and these are these serve as like directories here and those are you know ways that you can like think about it in terms of like going from the general to the specifics you go from ideas and discoveries to techniques or you can use ideas and then techniques and you can jump to discoveries it's it's that sort of synergy that you can build into these maps you know you have different categories of thing and you can build from that but you also want to cover transdisciplinary topics so you know you want to frame the map as a learner that grows our educational resources so right now Jesse has artificial intelligence and systems and ethics in his frontier map but later that might transmute into different fields so it's very constructivist in that way frontier maps should provide minimal information to be competent in the field so we go from broad ideas and assumptions and methods and techniques to systems and then the theories we can you know we have this sort of staged complexity of discovery you start at sort of the foundational aspects of a field and you build up to specifics and then nodes can lead to other nodes so you can sub-reference different topics and you can you know kind of go down a rabbit hole but in a controlled way so in the future Jesse would like to build visualizations of themes and individual nodes he'd like to incorporate more things you know jupiter notebooks and tutorials and presentations and then a lot of people that choose paths would you like to be an individual scholar or would you like to be heavily into the collaborative space and so that's that's something that talks about you know learner mode or learner style and so that's all part of his vision for that. Finally I think one of the applications is that you can integrate this with you know digital courses with course materials that are in your organization so in our organization we have courses that we host this is one of them this would be an example of putting these knowledge spaces into the course curriculum I don't show a really clear example that here but you can imagine that we can use these to teach this topic in different various topics in this curriculum and we can incorporate it with other technologies so we use GitHub as the host for this we don't have to conversely we have all these other forms of education in our digital course stack so we have GitHub but we also have hypothesis which allows you to make notes in web pages we have Gitter which is an interactive chat we have micro badges that we use to you know impart you know very small scale lessons to students and then we have YouTube which allows us to stream video of topics so all those things people are learning in our organization and but we want to also have this sort of these sort of epistemological directories one is in terms of like basically a reference so you can go back and look at different topics and store things in your own topic directory and you know compare notes but also just to have these assets as part of these directories to say you know use YouTube videos as a stub or use a hypothesis note as a stub and so those are all possibilities as well. Finally I would like to talk about how we envision sort of interdisciplinarity and so we think about this in a sort of a cybernetic model so we think about this in terms of like the evolution of these type these epistemological directories so we facilitate new contributions from allied fields on a regular basis of people with different skill sets and they want to learn about other areas that may be relevant to an organization and in the process they're making all these cross-disciplinary discoveries so they're making their own contributions based maybe on what they know or what they'd like to know and then they identify key papers and historical milestones and then there are other people who maybe have expertise in that area and they can verify whether that's important or not or maybe that's a good thing to follow up on and so this model of interdisciplinarity is based sort of on the practice of transdisciplinarity so it's kind of like building this emerging community of you know transdisciplinary scholars they're all kind of helping each other learn they're all learning that kind of at their own pace and then you're building this resource for future learners and it's improving with each generation so the community builds a model of knowledge rather than simply consulting or borrowing from another field and I say that with the caveat of what we mentioned a couple slides back which is that you know even if you have this community of people you know it's just one community of people there are other perspectives out there that might be you know that might be able to give you more information so they have their own biases their own perspective their own interests it's you know we really want to be as inclusive as possible to make these really successful and so finally I'll put in a plug for the data reuse initiative and this is our initiative that we do for data reuse and data sharing so if you're interested in a bunch of topics in this area you should check this out and we're accepting collaborators in our in this project and thank you very much thank you so much Raleigh and it looks like there's one question is what kind of information is with what kind of innovations with further enable greater interactivity between users and curators I think maybe building on to the github functionality so building on like you know things that are more effective at interaction between maybe learners or maybe like you know moderators and learners people with expertise bringing them together so you know we as github and github as a social component but we really do need to have a more explicit sort of interactivity you know an interactive aspect to this and it's that's what's missing among a couple other things but I think that's the biggest thing cool and another question is this is interesting and parallels a lot of basic frameworks for archived libraries especially with regards to transmissive engineering and into this into this interior work have you thought about learning from or partnering with librarians archivists who have similar goals well yeah we're at an early stage in this so we would definitely be interested in input from archivist librarians and other people who have done similar types of work but yeah we're looking for people who would like to collaborate so cool um thank you so much Bradley for this lovely presentation are there any more questions from from the studio audience looks like there's a lot of conversations about what software might be best and a lot of sharing of said software of different timeline type software well then um if there's no other question I am I'm sure Bradley and Jesse you may around on Slack during the rest of CSVConf and perhaps in the future as well if anyone has any more questions feel free to drop them a line in the Q&A in Slack Q&A we're about one minute away from closing ceremonies so feel free to hop over to the closing ceremonies URL and uh and hopefully see everyone there thank you so much