 So, this talk was prompted by – we often get asked what's the difference between a vocabulary and an ontology, and so there's a bit of a pun going on there in the title here, but vocabularies and ontologies, similarities and differences, definitions and structures. So there's a continuum here, I'm going to talk about that. The mini-agenda, just that Simon will introduce a bit of theory about vocabularies and ontologies and their similarities and differences, and then we're just going to step through five implementations of vocabularies. You see a very impressive one just then, we're going to go through five that are on that spectrum, and then we'll just conclude in a minute or so. So, Simon, over to you. Yeah, and Nick and I were preparing this this morning, so it's not necessarily going to be that slick. Anyway, yeah, it's great that we've just had the presentation that we just saw. We're probably going to be looking a little bit more behind the scenes here, and yeah. So, I'm seeing slide two. Great. Just to motivate it, first we'll take a look at a simple vocabulary. There's two approximately similar versions of the – well, of almost the same thing here. If you'd select that first link. Thanks, Nick. Can people see the web page that I'm now showing? Probably not. No. Oh, dear. I think I'm showing – I'm going to do – I'm going to cut back to the browser, and we're just not going to have to see things in presentation with that. We'll look at it. Okay. Yeah, I don't have any transitions, so it's okay. Yeah, so you should now see the web page. Yeah, so if you just scroll up to the top just so we see the headers there, this is a list of so-called agent roles. The list – this list I think comes from the doubling core list. So, if you scroll down now, you see that there's a whole set of the kinds of functions that a person or a job position might have in respect to some information resource. So, all those terms on the left would make sense. If you just scroll up just so we can see the headings here. What we're looking at here is a very flat SCOS vocabulary where each of the items you could click through, but you won't find much on those, has just basically got a label and a definition or description with no hierarchy here at all. Thanks, Les. Looks like the doubling core set. The other link that we had back in the presentation there is a related set which comes from the geographic metadata standard. You'll see almost all the same words. There may be a few more in there – funder and mediator and these things as well. But again, this is just a flat list principle investigator. There's a cool one, right? So, this comes from a more scientific domain. So, this is just – this is formatted just in SCOS as we heard from the previous – if you go back again, Nick, to the presentation. As we heard from the previous – that's the one I wanted to see. The previous presentation, SCOS gives you a basic hierarchical vocabulary structure. In the left here, this is literally the definition of SCOS concept as it comes in the SCOS standard in the RDF representation of that. On the right, you see a sort of a pseudo-UML picture of the class SCOS concept and all the properties that might be from that. And note for those people who are already familiar with the SCOS vocabulary standards, SCOS is not a full implementation but is certainly inspired by or the intention is to provide an RDF-compatible implementation of the well-known ISO thesaurus vocabulary standards. I've noted there that two seven double eight and five nine six four, which everyone kind of knows and loves or at least is vaguely aware of, has relatively recently been replaced by two five nine six four. So, a SCOS concept – that's the definition of SCOS concept. Actually, flip through to number six first, actually, Nick. So, here's an example from a vocabulary which I've been managing, working on for a number of years now, got papers written about, was published through the CISVOX service which we developed at CSI our own path. And there you see in all its glory the definition of something called the Cambrian, which is those of you from the natural environmental sciences would be aware that that's something important in geology. It's a time period which started 542 million years ago and ended 488 million years ago. So, you know, well before any of us were aware but it's very much a central part of historical geology. You can describe the Cambrian using SCOS and that's an attempt doing it there. And you see there's quite a lot of relationships there, narrower, broader, broader, transitive. That's the next step out. This particular one is taken from a resource which is the International Stratigraphic Chart from 2016. Also, we've got lots and lots of labels for that in different languages including some non-Latin languages which you can see down at the bottom there. Okay, so flip back to number five please Nick. We've worked on the gory details of what a geochronologic era is which is, the Cambrian is one of those, and you can design a much more semantically rich information model for that which is shown pictorially on the right there. And on the left, that's the the turtle encoding of the RDF implementation of that as an hour class. And you see that in bolded in the middle there, I've said we're going to say that a geochronologic era is a SCOS concept. It's a subclass of that. All geochronologic eras are also SCOS concepts. But we've also got a whole bunch of additional geology specific properties added in there. And if you flip down to number seven now, Nick. So that same individual Cambrian, which before I showed just as a SCOS concept, when you add in the additional properties which come from geology like the rank of this thing, that it's got a beginning and end which comes from temporal topology basically and that it is an interval and contains other ones and its end is coincident with the end of something smaller and its beginning is coincident with the beginning of something smaller and it nests inside all those things. That's all those bolded properties that are on the right hand side. So the beauty of this technology platform, the RDF platform is that if you only want to think of it in terms of SCOS, you can do that, which is what I showed on that slide six. But if you want all the additional semantics that you get when you really head into doing it as a full on geochronologic era, which is also as you see there a proper interval from the time ontology, then those can be added in without getting in the way of viewing it just as a simple concept. And in the past, I've published these through vocabulary services with the idea that when viewed purely as vocabulary fodder, if you like, you can query it using all of the SCOS properties. But if you are a geologist and you know more about what this concept really means, then you can query it or view the information about it using the richer properties. And in the previous presentation, we saw flipping within the pool party environment. Well, actually, no, it was in your interface, wasn't it? Or maybe it was pool party, looking at it through the SCOS interface. But then, as you said, you had all the additional clinical properties associated with the drug, what it was designed, what conditions it was designed to treat, which are obviously go way beyond the semantics that you just have in the simple knowledge organization system or SCOS. So we're really in the work that we're doing is trying to preserve the anchor back to SCOS, while also using the possibilities that you get from Al and ontologies. And of course, those concepts I showed in slide five, the definition of geocronologic error, itself would need to be published in some kind of register of ontology definitions. And that could a list of things like geocronologic era and stratigraphic section and geocronologic boundary themselves would form a richer kind of vocabulary of vocabulary of those geologic time terms. So pass over to you now, Nick, and you can flip through those examples. All right. So just before I do that, I'll just mention that SCOS itself, for those who are unfamiliar, is actually an ontology. It just happens to be an ontology about vocabulary. So knowing the slide over on the screen for the moment, things that are formalized in ontologies can be formalized in SPOS and other ontologies because SCOS itself is one of those sorts of ontologies. So I'm going to skip over a couple of slides because we're a little bit short on time, but I'll just talk to them very quickly. So one of the first examples of the first example here is the Global Agricultural Concept Scheme Core. And that's just a very big, flat SCOS vocabulary. So it's got a lot of terms in there. It's delivered with a tool that's similar to, but it's a different tool to the pool party tool that we saw before. And it's definitely in the vocabulary land. So it's vocabulary, not ontology there. The next one, and so of course, people can click on these links in their time and you'll see the interface that's provided. It does have machine readable versions that has HTML web pages and so on, but it's definitely in vocabulary land. Now, geologic timescale, Simon's actually talked about that. So we'll skip on over, but just a reminder that it's both a vocabulary plus. So it's got a couple of extra fancy bits. Another similar one is an observed property vocabulary that Simon's also worked on. And I will click on this one because it's going to, if I do it right, show us a slightly different tool. So this is the linked data register, which Simon actually did show before. And like both the pool party tool and others, it's got various tooling help get around vocabularies and so on. Now, in this case, this vocabulary is a vocabulary plus. It has some other properties. And I have to look a slight to remind myself what they are. Yes, so we can look at. Any of those soil moisture ones would get me. So here's the information about deep soil moisture. And there are some other things like features of interest that are related to this term that are clearly not Scots things. Okay. Now, this next example, example four is the one that I think best highlights vocabulary and ontology differences, because it's actually both. We've got two things there. So the first thing I'm going to show you is a vocabulary that is used to classify physical samples. And that's the vocabulary bit. So you can see in the name here it's got the word voc for vocabulary. And then separate to that, we have an ontology as well. And I'll talk about the similarities and differences. So looking at the vocabulary, this vocabulary is got terms like blast. The blast is a method for extracting samples. And it's got a whole bunch of other terms in here. DIY is irrelevant to samples, but maybe you want to use them for identifying samples. Dredge, that's a sample getting thing. And this is just a list and it's derived from an HTML code list. So in the normal way of doing business, you can define a code list. And then you can take that code, link data and scos, and now you've got a scos vocabulary. So that's the code list. Now the ontology, very different beast. The ontology is data model around things of interest to samples. And I'm using an ontology viewing tool here to actually turn the ontology into HTML and text. And it tells you about the classes of things that you might want to know about in relation to this sample or to samples generally. So it's like the ontologies that Simon's just mentioned and there's some pictures somewhere or other in the documentation that will show us what graphically what the classes of interest we think of interest around a sample are. But one quick look at these two is that the ontology actually uses the code list. And I'm going to show an example from a live system here which is GA's sample catalog. So what this is doing is it's just a landing page for metadata about a sample. And I'm now going to click on the RDF or turtle view of this. Now I'm wondering if this is going to actually be visible. It just is a land load, Nick. Yeah, I know. I don't know how to change that. Go to your download page. I can do it manually. I think this will force the download still. Hopefully. Yeah, it's going to force the download still. Okay, I can't show you folks but I can describe it. You can see here this sampling feature. So that's a property or that's a class that's related to the sample and it has the term for whole. So that's that's an ontology property. There's no nothing scoss about that. That's a special thing that you have to know about samples and ontology and describe. But the sample type. It's got the word core here and cause a URI and that URI actually links to a term within this code list. So here we have the ontology specifying some relationship in this case, the relationship of sample type and then it's using a term from the code list to actually instantiate that. So here we have ontology using vocabulary. They're both available for use to describe samples and the from a management point of view, the update cycle is very different. We expect the ontology to be updated very rarely, whereas the code list we would expect to be added to quite frequently. So the last one I'm not going to demonstrate. I'm just going to mention example five. This is the probe ontology. So it's a core W3C ontology for describing provenance and it is not at all a vocabulary. It is a data model and there are no individuals or instances of things that you can refer to directly in Prove. So if you look at the Prove documents online you'll see types of you'll see classes that you can use. So there's a class called entity and you can describe your data items as an entity and link it to people which Prove would describe as an agent but there's no sense in which there's a list of individuals that you could refer to. So it's right at the other end of pure ontology not vocabulary. Okay we've definitely blown our time but that's it. So I will stop sharing and hand back to Kim.