 Thank you very much. So yes, just a quick disclaimer, so Elizabeth Dalton obviously who is heading up the project Inspire within Moodle HQ wasn't able to be here, so that's why I'm presenting on this. So, you know, learning analytics, we heard earlier about learning analytics and how it can work and how it sort of what people are doing and what the concerns are. You know, Inspire is a little bit different, taking approach of it's not necessarily doing the same as every other product out there because it is about something, having something in Moodle. And so learning analytics are about learning. It's not just about the administration. One project that I knew about a few years ago, they were using learning analytics to track if a student logged onto Wi-Fi, went to the library, logged into their email, logged into the LMS and a video platform and they were the five lights, if they didn't turn on in a week, that student hadn't engaged with the college. Now that is analytics, but clearly it's not learning analytics because there was nothing in there about whether the student is learning, although that was the foundation of whether they would be able to because they hadn't engaged. So, if we look at this, we have this four different areas, academic scholar, looking at efficiency, being learner centered and social reconstruction. And you know, most people at institutions subscribe to one or more of these sort of areas. And these different theories, it's so different, everyone has their own different opinion on learning. We heard earlier about the different course types, the different teaching styles that people are doing, the different purpose of their LMSs. And so understanding what learning is, is really important in this regard. And then you sort of move forward into that and you think about, okay, well, so what is the known good? What are we thinking about? This is what we're trying to measure towards. So, for example, we look at mastery of competencies. It was one of the things that was mentioned earlier and it was immediately questioned as, well, how do we know that they're actually a master, that they have mastered that competency or mastered that ability? Increased community influence. So, if it's in a social way, you can look at it from a completely different angle and have they got the right connections? Are they having meaningful discussions? I think Dr. John Whitmer and one of his slides on Monday had looked at the quality of engagement even between people. And then achievements of the learner's own goals. So, if they're stating at the beginning of a course what they want to do, that's important. But how many times do we actually ask that? Are they marching to our tune as institutions? Are we going, you know, you have to get a degree or 70% on this course? That is your goal. Well, maybe understanding is part of their goal. Maybe getting confidence and mastery in some of those areas as well as passing are their goals. And if we're looking at what is learning and what is a known outcome, do we have to take those sort of things into account? And obviously just scholarly advancement. Are they just learning and improving and becoming, having a deeper understanding in that whole area? And that is what it is about. So, that's going to be deep. So, efficiency is more light. But the social side, I think, is about community participation. But then you're getting into what is participation? So, it's reading forums, participation. Who here would agree that just passively those lurkers are participating? Okay, so we've got like five, maybe six out of the room. But yet, I certainly know from my own experience that you might have 80, 90% of people registered and using a forum just doing that level of participation and that's meeting their goals. But we don't, if you go with the left hand side, you go, okay, well, the learner knows their goals and they're getting what they want out of it. But we can't measure it. But they can measure it. So, there's a lot of research into these areas. And this is the whole community of inquiry side of things. So, you've got social presence. And that's where it's about engagement, even being at a moodle moot, supporting discourse, really getting into facilitating those discussions to get that deeper, meaningful debate going. But then you've also got on the other side where the interactions, goals, directions. And that's where you start moving into where it's a bit fluffy and a bit soft on one side into more regulated learning. And the whole regulation and it's exam based rather than discussion and learning based for learning sake. And this whole area encompasses the learning experience. And if we're going to be using learning analytics, to improve the learning experience, we have to start thinking about all of these areas, and how we try and measure them, and how we try and then communicate what that measurement is and that meaning behind it. And this is one of the areas where I know, Elizabeth usually goes into more depth on this area, but it's something which, so you do have a, the teacher has a place here, and we're not talking about using learning analytics to replace teaching. It is about supporting, but you have the teaching presence. And then you have the social, the students, the peers, where the student is learning from and all of their areas around that. But all of these could combine and feed into where that cognitive presence is. And these relationships have been tested in research and in sort of post-course discussions and analysis to see how these come together. And all of this sort of basing on where we're going with Inspire. But if you look at the Moodle course and you start producing reports now, and I think this is one of the reports that was shared with the ad hoc reports on Moodle Docs, you look at what can you see when you look into Moodle. So you've got different events. So Moodle, every time something happens, it fires off an event from most cases, and it records that happened so a user logged in they viewed a course another user logged in there and you can see the different times where it was which course and so on but what does this tell you it isn't actually information you can ask it questions and get information out of it but the data itself is not meaningful which also ties into what was being said earlier in the panel if you just give data to students is that actually helping them what information should you be suggesting that comes from that data or what information should you be giving them and then this is sort of a process going through the whole process around the cognitive depth I'm just going to check the reference here so what we're looking at is the model around and we're working on a number of models at the moment around how to test what this different depth of learning is involved and I'll share all these slides I know that's a little bit small to follow I'm just trying to give the context behind inspire at this point then we look at the social breath that's the cognitive part the social is about the connections between these so in forum posts John was referring to some research earlier where you do have these stronger binds between students so you might have for example in forum posts where he was talking about grading forum posts you can have things like where a student might be very busy on a forum saying hi John how are you hey I'm great we were last night and you might have some social interaction but are you looking at the meaningful amount of infam of text that's going in there is it actually purposeful are you detecting confusion like you're not understanding what it is so really getting into the quality of those connections and not just the volume of those connections and that's a really important area to look at but a page if you just have content it doesn't really allow for that social interaction so then this limits which activities and in some ways when you look at all of the different activities in Moodle some of them can be more useful in one of these areas than another and so you might have messaging you might have forums you might have chat or commenting on glossary and database where this kind of action or interaction can be tracked and also then assessed and if you categorize all of these you can see here that exactly I was saying with the different activities you can go okay so on the top left you can have the book you can have folders you can have pages you can have score and objects although arguably that can also appear and in different areas and how the different activities can go much and can have a wider social breadth and a deeper cognitive usage and so when you go down you've got assignments we're actually creating materials or chats where they're discussing focused discussions and forums and wikis where they're co-creating content so you've got this whole matrix of potential with the different activities and that's one of the things which is hard and that's why you have to try and assess that because people use them for different things I know some lecturers who use wikis for example instead of a Moodle book because they like writing in wikis and they like the way it navigates but its usage in their course is the same as just a static web page where others heavily use it for this collaborative learning and so when we're looking at all of these and you're trying to analyze well how are they using this wiki what are they using it for what's a teaching purpose is there a teaching purpose is an assessment and see when you take all of these areas you take the social aspect you take the depth you take the assessment approach you take whether you're going for a regulatory and sort of exam based or whether it's a knowledge based or whether it's mastery and you go so this is where AI can really start getting into helping and machine learning so what we're looking at is collating all of this into a central into a central brain where our machine learning will start turning these predictions and taking them together to look for patterns to I mean this technique is widely used and the goal is to predict student success automatically based on previous records earlier was mentioned that volume of data isn't necessarily going to be just as useful however it is something to create that baseline and with Project Inspire we're looking to have this API which will allow Moodle to be able to push this data and predict two teachers and two students to help support them now before we go through this process obviously we're doing a research phase where we're looking to anonymize data and then later on it'll be something where you have this actually within your own within your own system but we're also making it pluggable like with any of the Moodle system so you could maybe use a different machine learning tool as well now there is a working prototype does anyone here ever go to the prototypes page on Moodle okay so please go there and have a play around it's just indicative it's just some examples at the moment interfaces and final at all but we are learning at the moment so please go and have a look at it and indeed are just some examples of the kind of interface pages there this is identification of a student at risk of dropping out and this one here is about all the different kind of indicators that are being taken into account so we talked just even about those four areas but there's lots of indicators there and the idea is that the machine learning will try and interpret the results of the data and also within your own Moodle site of what is a good course and how the courses work and what the outcomes are expected from this format of course because obviously there's been good research I believe at the OU as well about how learning analytics and course design really do work hand-in-hand so it's important to take those sort of things into account and then it's about the different models and you can choose which model because you might have different models based on the different teaching subjects so whether it's mathematics or whether it's language the same type of course design might be used in a completely different way and so this is another one again you can go in and have a look at the prototype and this is obviously a teacher being able to see students some who are at risk and some who might not be and you can then you can sort of drill down into what is the predictive analysis behind that and then here you can see well they 50% of their profile is completed I must admit I don't know any student who ever completes their profile does anyone here complete their profile fully on Moodle.org even okay eight people okay you all fail so but this is the this is the basic behind of what we're talking about courses with regular opportunities for volleys of discourse and for cognitive engagement but then you have to have instructors who are going to provide regular prompt detail feedback maybe that Jill Watson can help with the regular part but publicly in private and students who actively and regularly participate and not just lurk so now we're talking about that 10% of people and so one size doesn't fit all and which we have already sort of touched on that it is so many different versions so we are looking at all of that and there has been a lot of research and that's why we're doing more now so we can understand and we are it is going to be completely open it is going to be completely possible for you to see what how those algorithms are working and what basis is and you might turf them out and implement your own instead what we'd like you to do is to get involved now we have a course on Moodle.org where you can go in and you can get information on the project in more detail you can also then choose to participate take the survey let us know share your anonymized data with us and then really it's it's about engaging and working together it's like John said you know let's start trying something now let's see what the data can show us and learn from it both learn as institutions and learn as a community and Moodle itself to see what we and what is possible within this because it's time to get your feet wet thank