 If we were to associate education to one of these networks, what would it look like and are we getting the most out of the experience we're giving to learners by perhaps, I would say that a lot of institutions are perhaps more to the decentralized approach of learning and teaching where we've got this wonderful weapon of mass distraction that the U.S. government have done. Facilitated and creating that can connect us not just to people, but to knowledge. And so I think that's something very interesting, which is obviously why we've started looking at that as an aspect of OPDEL. So I'm not going to mention the M word, but in 2012 I started looking at what people like George Siemens, Stephen Downs, David Cromroy were doing in terms of a distributed education and the tools that they were actually using and you recognise a lot of these names, these services. It's quite interesting, Dave White and his colleagues have found some very interesting ideas into visitor and residency in different spaces, but we're creating data, we're doing stuff in these spaces, but how can we actually use that within learning and teaching? And this is something that George Siemens has recently been thinking about. Google have developed the knowledge graph, so this idea of nodes of information that can be connected, so it's like the six degrees of separation of Richard Bacon. Is it Richard Bacon? Am I thinking Blue Pea presenters? Someone, six degrees of something to someone. Kevin, thank you. So this idea of we're present in different spaces, but that potentially builds the picture of that person. And it is about the person, their attributes in terms of things that they already know, the connections that they already have to other people or other information. And this was post-reflection on doing octel, was how can we start doing some of that within the octel course? And as the title of the talk indicates Open Badges seems to be a good way to do that. And I'll detail some of that in a second. And this is just underlying the idea of a personal learning knowledge graph. And I think there is implications, you know, Audrey's touched upon that very eloquently this morning about ownership of data, ownership of your profiles in different spaces. So that's the kind of some of the context in my own thinking around this, so octel. For some more information about octel, it's actually a course that's facilitated by it, but it wouldn't be possible without alt members. So the lead tutors are alt members. The support tutors are alt members. Any tutors or support tutors? We've got Linda, Tracy Madden. So these people are part of the construction. They're developing the materials and they're helping with the delivery of materials. And so it wouldn't be possible to actually run an octel without them. This is kind of the general model that we use within octel. And if you've been browsing around the conference platform, a lot of what we developed in octel actually gets put into the conference platform, so it's this idea of being able to actually pull in some of the data from these different spaces. So this year we're pulling in slide shares, for example, into the conference platform. And the reason for doing this is to then, as a very loose collection of people participating in the course, it's redistruding that information back to them so that they can start making their own connections. It's not just connections to new information, but connections to new people. And having a profile as part of that is key. And also email is key. We can't get away from email right now and it's a very good technology for actually pushing back. So that's what we're doing. We're sucking information into a site, a WordPress site, and then we're pushing it back out by email or RSS feeds or just coming to the site itself if that's what you want to do. Obviously there's implications with using, I've tripped over the RSS word. I'm not going to go into too much detail about that today, but it's a technology for just moving data around places. And it's a technology that seems to be on the wane. So you should be looking at the work. People like Ken Lane, who was mentioned this morning, are doing in terms of another technology, APIs. But again, I won't go into too much detail about that. So the interesting stuff. So the badging aspect. So hopefully you're familiar with digital badging. I don't think there needs to be much explanation about it because I think a lot of people already grasp the idea of what a badge is. Digital badges is exactly the same. It's something that an individual can have. They can display if they want. It's something for them to collect. So as part of OCTEL, we had a number of badges that participants could collect. The course is structured on a weekly topic basis. And so each week of the six weeks, we had a collection of five different badges that people could achieve. And these were the same types of badges each week. Although the activities associated with those would be related to the week in question. We started off with some very basic badges of checking in. So this was a case of coming to the site and clicking a button. Check-in badges, I think, were useful. And you'll see why in a second. Just as well as an entry point getting people used to what they have to do within the system to actually get a badge. Then we were using weekly webinars. So we had a webinar badge. We used an access code at the end of the webinar. So the people that were coming several weeks soon worked out that they could scroll to the end of the video and collect the badge if they wanted to. But we didn't get too hung up about that. And then the final badges on the top row there are a TEL-1 and a TEL Explorer badge. So a TEL-1 was an activity, one activity. If you do any activity this week, this is the activity you should do. And for that, we'd asked participants in OCTEL to submit a link to something they've got as evidence of that activity. As part of the OCTEL site, we did include forums. So it might be a link to a forum post that they've made on the site. But the idea was to encourage them to write in their own spaces, so in their own blogs or in their own Google Plus groups, and then share that URL. So the only stipulation was that the evidence that they were submitting was public, so that we could go and see it anywhere else. Browsing the internet could see it as well. And then we had to TEL Explorer badge, which was, if you wanted to do more than one thing, you could do two or three or four activities. Same again, submit a link to that. And if you did any three of these badges, you automatically got a weekly topic badge. So we did this all within WordPress and the plugin we used was called BadgeOS. It's a free open source plugin. If you're interested in badging as part of courses, I would highly recommend that you go and look at it, and I'll highlight some of its features if I have time at the end. But one of the nice features within it was that you could create steps. So, for example, the topic badge, we just created a template step of, if you did three things, then you got this badge automatically, which helped a lot with the administration of this. So I mentioned the TEL One and TEL Explorer badges. So this was useful for us in terms of an open course. Even though we've got data collection methods for aggregating activity, if you have a blog registered with the site, we'll pull in the data for it. But not everyone wants to take that step. Not everyone has a blog. So the idea of people actually submitting the evidence to us starts filling in some of that personal knowledge graph. So they're declaring bits of information that they want the rest of the world to see that other people on the course might actually find useful. And I think that's quite interesting for us in terms of being the ability to build up a picture of a person. Because we were using a variation of digital badges called Mozilla Open badges, the evidence associated with the work that they were submitting actually gets baked into the badge information. And so when that badge is in the Mozilla backpack, which is a portfolio place where you can store your badges, they've got a link. The person looking at that badge can potentially go back and see the evidence. So we're getting some sort of interoperability here. I mentioned the check-in badge and one of the aspects of that I think we found useful in terms of a personal knowledge graph was one of the features within the system was the ability to show who else had actually earned the badge. So it gives situational awareness, again in an open course context to actually see who else was active in that week. I think it's very useful for other participants. It can be a very lonely experience. And clicking on the person's avatar, if they've been awarded the badge, they can go through to the profile, see the person's profile on the site, see where else they existed in the internet, make those connections off the site. So it's making something that is useful for the individual beyond just the course. I mentioned we were using badge OS plug-in for this and one of the other nice features was its ability for tutors to provide feedback on students' work. So those familiar with the WordPress plug-in platform will recognise some of this interface and it's just using the comments as an opportunity for both the tutor and the learner to actually engage in dialogue around the evidence that they've submitted. For us, this was all in the public. So again, there are opportunities for people to make connections or to learn vicariously from what other people are doing. I'm going to skip out. This is some of the features in terms of the badge OS plug-in, in terms of creating these stepped procedures that would automatically award the badge, which we found very useful. And because we were integrating a social network aspect with body press, we can actually do things within the community. But again, all these points, these are optional things for people to do. These are optional kind of nodes within the graph that people could start creating for themselves and hopefully start making connections. I think one thing to reflect on is a lot of this isn't particularly new and I'm a member of a community called Stack Overflow, which is for coders and again you have this idea of badges and people can award badges to other people. So there's already a lot of experience of these types of systems going on. I think one of the questions asked often about badges is, do they actually count? Do they mean anything? And it's quite interesting as part of the course that even though we had people at the beginning saying, oh no, I'm not going to do badges, they're not for me. Within week two or three I was getting emails about how can I submit this evidence for the badge? I'm sorry, I've been caught up in this and I think there's a real ownership here. People, they want to collect things, they like doing that and so always these badges can count for the individual. They won't display them if they don't mean anything to them. Wherever they mean anything to an employer is another question. I'll skip over some of the r-bots, but I will point out one thing. There is a wider direction of travel that I'm quite interested in following here. I'm not going to go into too much detail but you should Google domain of one soon. See what University of Mary Washington are doing in this space because I think that will be a real eye-opener for you. Audrey is the advocate of that project. I trust and respect what Audrey says. If you don't trust me, trust Audrey. There's obviously privacy issues so I'm just going to throw that up there for a few seconds. Mainly so you know that I know there are concerns which maybe you'll ask questions about and finish with a thank you. Do you have anything to ask any questions? Do you have anything to ask is said to wait till the mic gets there because the audience participants can't hear you. Anybody have any questions please raise your hand and the mic will travel to you. If you have any questions online then press question here. Then raise your hand. Thanks. Moira Mayly. Hi. These digital badges flag the beginning of some sort of international cross-platform degrees or qualifications of some sort and usually standard systems follow those initiatives and they can sort of slow them down and it gets caught up in a bureaucracy. Is that the way you think that badges will go? Do you think more generic competencies are going to be indicated by them which could be added in? It's early on. I think a lot of people are... I was talking to Fiona Harvey from University of Southampton who's starting to look at badges as well. She was saying that there's issues for us, for all, as an organisation. We don't have institutions like the Quality Assurance Agency looking over our shoulder all the time. For institutions like University of Southampton, there's obviously very different implications. Talking to Fiona, I think for them it was trying to tackle it in two different ways. There would be unofficial badges awarded as part of some of the courses that they're doing, but that might then lead on to a university recognised badge. Whether that slows down the development, I don't know, would probably be my honest answer. It's creating those opportunities, so I could start creating my own badges. I could award myself. There's nothing stopping me to do that, so it's the opportunity of people creating badges for them to achieve. I don't know if I answered that. I get the impression just from the last three days that badges this year, at least for me, are this year's kind of hot topic whereas Moog was last year's hot topic. This year's kind of got creative juices flowing. I noticed that the conference website this year is WordPress and BuddyPress, which seems to be exactly the same as your octel course. I'm wondering if you'd had any discussions with the rest of the association about whether you're going to be awarding badges for next year's conference. It's funny you should say that. It was a discussion that... It was something we wanted to do and there were a number of aspects that we could potentially badge in terms of if you're a presenter, if you're an attendee, and aspects of your interaction on the conference as well. Unfortunately, we were beaten by time, so I think that's one of the useful things of doing octel is it's an opportunity to experiment with these things and then deploy them in a different context. That's our development arc, is octel. We then put that into the conference site. Then whatever we learn from the conference, we'll put it back into octel. At the same time, we're looking at thinking about how we can support the wider octel community as well. Obviously, octel, seymol are the aspects of that that we can start including badging. Octel has a number of special interest groups. Are there parts of that that we can start badging? We've got badges. We've got badge sunglasses on. Everything is badge-tinted. That's our last question from Bob at the top. A reflection, that it felt there were two distinct levels of badging. What you were showing us was really quite fine-grained badging about participation in maybe just a week of the course, which felt great in terms of motivating the learner, et cetera, but feels sort of way too fine-grained, for example, an employer to age with. I'm wondering if you've thought about aggregating those fine-grained badges together to almost automatically feed into higher-level badges, or is that very complicated or what? It was very much the reason that we had, as well as the more fine-grained badges each week, this topic badge. That would be something. At the end of the course as well, we did some special badges. If you got all the topic badges, then we had a gold badge, an octel gold participant badge. I think that was useful in terms of allowing the individual more control over what they wanted to display, but also keeping that motivation going throughout the course. I think that's one of the things with badging is that it's micro-accreditation, and there's some really interesting work in terms of creating different pathways or different collections of badges that actually mean something else. I think there's going to be a lot more work in that area. OK. Now, between you and Lunt, we have one more really interesting talk for a speaker I'm going to introduce in one moment, who's, I think, ready to go. Are you ready? Fantastic. Excellent. Thomas has joined us from IBM TechTrade, and will introduce himself and what he does a little bit more, and tell you a little bit more about data before lunch. So I'm going to hand over to Thomas, and please give him a warm welcome. So thank you very much, very well. I'm keeping you from lunch. I'll try and keep it to something that interests you. So when I was giving this opportunity to Alton, I can only thank you for doing that. Why should you listen to me? Because, obviously, especially after Audrey's presentation this morning about all big companies just want you to essentially buy from us. I don't think that's the case from me here. So as you will have just seen, I'll press the right button. There we go. Dame Wendy Hall spoke last year. I actually did some research in the conference as well. And she mentioned here about when, in her web conference, which I found very interesting, big data at the bottom down there. She's talking there about how the web has enabled is pushing a lot more data out to everyone. And that's part of what I do. So I'm actually part of our big data and analytics team at IBM. And then also to top that off as well, on Monday, just to help me, Jeff spoke about what the priorities for education were. And in that, he mentioned as well, just in the middle there, a data-driven world in analytics and predictive. So it's not just me telling you that I think this is an interesting topic, but there's some key people in education who are saying this as well. And there's another person from IBM, actually Caelan Hargrave. You'll see his Twitter handle at the end, but if you follow him, he's got some really interesting things to say around this as well. And then the final thing is, is there anyone here from London South Bank University? No. Okay. It's invested with IBM a lot of money. But the reason behind that is because they're really trying to empower the student and work with the student. This is actually from a press release on their website that if you Google, you can find. And I'm going to talk a little bit more about LSBU a little bit later. So what I'm going to talk about, so I'm just going to give you a quick overview of what IBM thinks big data is. And that's come from the reason why we are, I guess qualified, is we invest one billion a year in R&D for big data analytics. So it's quite a big tranche of what IBM are looking to do in the coming years. Some case studies, so to go back to Audrey again to actually talk about reality and not fantasy and tell you some stories about where it's actually been used and then actually talk about where it's been used in education because obviously that's what everyone's interested in. So what is big data? Well, according to Gartner, there's three major views. You may have heard of these. So they are volume, velocity and variety. And that, to qualify to be a big data problem or just be big data, you've got to have two of those. So essentially, high volume, lots of data. Velocity, how quickly that data is coming at you and whether you need to analyse that quickly. And then high variety. So it's a fact that there's now structured data and unstructured data. Traditionally, we've always dealt with structured data and we've dealt with it very well, especially from an IT company like IBM, but how do we actually deal with this new unstructured data, whether that be social media, log analytics, those kind of things. And then there's some interesting quotes here. So one from Clive Humbie. Data is the new oil. Data is just like crude. It's valuable, but if unrefined, it cannot actually be used. I think that's a really interesting point. So if you've just got data, data is just data. It's just a number or some text. Essentially, once you can do that analytics and define that value and where you want to go with it, then it becomes of some value to you. And then another one from John Naysbyt, who's an author and speaker on Future Studies. He's actually pointed out that it's one of the few resources that is not only renewable, but also self-generating. And data will continue to grow and grow and grow. Now, obviously, as we've talked about with Audrey, previous security and privacy issues do become a part of this, and that's something that everyone needs to think about when they're dealing with data, but essentially it is going to continue to grow. And there's nothing new about big data. Data has always been growing. That's just been the case for the time. What's new about now is the technology that's available to actually do something with that data. I would argue. So this is IBM's point of view on the characteristics of big data. So yes, we agree on volume and velocity and variety. So volume, as it says on the slide, by 2020 there will be 35 zetabytes of data. That's a hell of a lot of data. And velocity, there's 30 billion, and this is over a year old slide now. So I don't know whether that number is still actually true. It's probably more. In fact, Jeff on Monday spoke about then putting RFID tags on toilet rolls, so they can follow him around. I don't know where he got that from, but I found that very interesting. And then variety. So that's a key one for us, effectively analysing that different variety. So, again, going back to that structure and the structure. The extra one that we'd add is variety. So actually, do you trust that data? Because it doesn't matter if you do some analysis on data you don't trust, then you're not going to trust it. I'm sure you have it in every day. The job you do, I have it, I get a spreadsheet, I question a few numbers, and then I do the analysis and I'm like, well, I don't trust that analysis because I don't trust the data it's come from. You've got to have that. And then these bottom two have a little star on them that are my personal. They're not IBM, so using the disclaimer we talked about before, this is my personal view. Understand the value of that data source. If you don't know what the value... Some data, admittedly, is not going to be valuable and you don't want to be analysing that. Some data is very valuable, and only you will know that for your institutions or where you work. And then visibility. Can you see that some of that data? There's a lot of organisations that we talked to actually can't... They can't see it, but they can't see it and they can't get to it in time for when it actually matters to them and what they want to do with that data. So that's what IBM thinks are the characteristics of big data. So let's talk about some case studies. Now, these out from education are going to go through, but they give you an idea of what the art of the possible is. Whenever I talk to different customers about big data, I always try and open up with these just because it gives an overall idea of what the art of the possible is. So the University of Ontario Institute of Technology that's using it for babies and it's a topic that everyone could understand, so it's one that I like to use. Every time I'm premature, and I don't know, I don't have experience of this so I can't claim, but every time I'm premature, baby is born, they are very likely to have a seizure. And if the cure is out there for this seizure, it's just you have to give that cure the right time and they'll be fine. And most people aren't aware of this as a problem. So the way that they traditionally do it is they have a nurse walking around the ward looking at the babies, looking for the specific symptoms of those babies in the ward. Now, what they've been able to do with big data is actually take all the analysis and all that data feeding off the baby from all the tubes and wiring which I don't find particularly palatable, but obviously it needs to be there. Essentially, all that comes out and they're analysing that on the fly, so as it goes, and they're able to identify 24 hours sooner than a nurse walking around the ward actually when a baby is coming towards a seizure. As well as that, they've also identified a pattern to actually be able to work out a little bit better why those seizures happen. So that's helped save lives in this ward and it's something that we're looking at from an IBM perspective at what we can do there. Now, going back to the security and privacy one, I was debating about taking this out this morning after Audrey's talk, but I've kept it in because of what it can do. Now, Terecos, they are managing they manage a lot of secure places. So this is actually in America they are managing some of the nuclear education sites. So obviously they've got nuclear particles on, they don't want people coming near the site because obviously there's the protective around that. Essentially what they're doing is around perimeter fence, you would just have usually a guards in the guard tower in the middle looking out and send out people every now and again. What they're now doing is they've got sound detectors all around the outside of that perimeter fence and they're using the last two to five years worth of data to actually tell them when someone is approaching. So they're analysing the sounds that are coming in. So this is not data in its traditional format. Structured logs or even text data. This is actually sound data coming in so they can tell the difference between a fox walking towards that perimeter fence or a tree falling over 500 yards away to someone actually coming and walking towards that facility. And all they do then is they send out the guard to that area and the guard will deter that person from coming any closer because obviously they realise it's a secure facility. Now, I'm not saying that Big Brother is coming to watch you with this technology. There is very specific use cases and where you want to use it but it's just an example of what that are and the possible is. And the third one is a very simple one. Essentially infrastructure and managing infrastructure. So actually we understand that a light bulb has gone before it's even gone or that you're managing those energy bill forecasting those kind of things. All kind of all available because we're using the massive amount of data that is out there. We're leaving it in its current format. We're sticking it in something called Hadoop in this case but different tools in different instances and using that to actually drive out some outcome and insight that individually these are customers have asked for. And the final one that I'll go into because I find it quite interesting anyway is the smart crime prevention. So Memphis State Police we're having a 3% rise per year of crime in their state and obviously as we all know with cuts et cetera the US is just the same they were struggling with they needed 500 more police individuals to actually manage that so they actually looked at a different way to approach it. They're using predictive analytics to tell them based on what they know from the patterns of how criminals work working out the patterns of that so they can send a police car to the place where they think that crime will happen. That's actually caused a 30% cut. Now you think that's surprising because once the criminals work out that they're working by a pattern they'll start doing some anomalies but actually the way the human brain works you're still working to a certain pattern even if you're trying to avoid the pattern apparently. So I find that interesting. There's a really good advert actually that I'd be able to have done for that if anyone else wants to look up at that. So the more interesting part for you guys is what have we actually done in education? Are we going to go on? There we go. And I apologise for the pictures from my holiday to Australia because I couldn't find any relevant pictures for this. So this university in Australia has six campuses around a major city unfortunately not all our reference cases allow us to name them but they have a lot of students and they focus on fostering academic performance to gain their maximum benefit and they actually created an office of strategy and quality and wanted to look at what they could do to actually improve the students who were at the university on how they go through that course and have their experience all the way through that university and so they had theories on correlation on why students were dropping out or showing less interest drawing a course. These were range from the level of language their gender, their entry scores and their socio-economic status so they decided to measure and actually monitor with student opinion and course work and performance and actually work out if their hunches and what they believed to be true was actually true. And what IBM have been able to do with this predictive analytics tool using all that data that they've collected from previously identifying the patterns to that. They can now identify risk factors quickly and accurately and can intervene proactively to reduce class failures and drop out rates. Now, I'm not at this point saying that we removed the tutor or anyone who is involved on the face-to-face basis. I agree with Audrey this morning when she talked about actually I remember when I was at university the face-to-face was what really drove the value for me. But if you can before you're actually aware of it from an individual perspective be warned to a percentage of we are 75% certain this person is dropping out because of X, Y and Z because these four people in the past have shown that correlation of behaviour then you can identify to that tutor or individual to maybe it might just be send them an email to say there are these funding courses available because we know that they're from they might be struggling with finances it might be the tutor just goes and have a word with them and say are you struggling but that decision is still made from the institution. That decision is not made by any tool that we tell you you've got to go and do this. We give you what the what we think the reasons are why and we can tell you we can help you make that decision but we will not make that decision on your behalf. I told you I'd talk about London South Bank University a bit more so they took all of what we call sorry about that our exceptional student experience we have these leaflets in our stand outside if you want to come and grab one but essentially the key part of that exceptional student experience for London South Bank University was all around predictive analytics and being able to predict when students drop out they have quite high rate in their personal view of drop outs to reduce that. They're doing that for several areas including collaboration and how those students collaborate with each other, their tutors etc but the key benefit to them is reducing reducing drop outs. American Public University are the open university in America and they again and this is going back to that point where we're not telling you it's definitively like this we're actually telling you we can predict with a certain amount of certainty and actually on the tool you know you can build it out but essentially it's telling you we are 75% or we are 80% certain of this so we're not saying yes it's not a yes or no answer it's a this is how confident we are and you might say well we'll ignore it if we're not over 60% confident that's fine, we're not telling you otherwise but we're actually giving you essentially towards whatever decision you might make and that's again increased retention again which to our State University and I know we're now going to go into four actual US examples but as Jeff said on Monday US trends do come across the waters we've found and I know that Audrey talked about some of them we don't want to but I think this is actually one that's worthwhile and then Hamilton County is actually using it for their I don't think this is as relevant potentially but they are using it for their teachers and the training courses that they go on so they're actually working out from their schools that their class has and the teacher feedback schools actually which courses that they go on are most relevant to them and which one they're gaining most benefit from and then finally Michigan State University they're actually measuring and using this tool with predictive to say Tom went to Aston University he likes sport I can see that through social media and things when I'm looking for a donor to donate towards a new mini bus for Aston University Tom is a good person to send that email out to because he's more likely to donate as an alumni essentially again it's just predicting it's not telling you it's definitive but it's saying this might be better to go to that person so that's the end of my presentation I'd just like to say we are out and stand six outside that's the actual picture from this morning you can get these if you want to talk to any of us Cailin who I spoke about this morning earlier is the education leader at IBM and he's done a lot of these talks unfortunately I'm standing for Cailin so it's a shame you can't get anybody on holiday but he's heavily on Twitter and then there's all our emails including my colleagues from TechTrader in the middle of the building and who can also answer any questions that you have on this