 and it looks like it's in presentation mode now. Thanks, Megan. Okay, thank you. So, this presentation's coming at things from a little, a very different view, I guess, of a vocabulary user, myself as a soil ecologist. Some of the challenges we've encountered. So the application challenges of applying control vocabs in standardising soil data. I had no faith in the internet either, so all slides will see how we go. So this work's taken place under a number of soil data information related projects that the team here at Surrey Fair Junior partners on, including capture management, discovery, portals, delivering data through ARDC supported Agriculture Research Federation founding project and the focus of this talk will be the work that's being done under the Soil CRCs visualising Australasian Soils project. So the aim of the Soil CRC visualising Australasian Soils project, that's BAZ, is the creation of an interoperable spatial knowledge system for the Soil CRC participants and the broader agricultural community with access to data information and knowledge on Australasian soils. So the focus of the vocab work to date has been to assist with the standardisation of grower group and capture management data being made discoverable through the platform. So just in short, soils related data is made discoverable through the VAZ platform from a number of sources that today are focusing on the data of farmer groups and capture management authorities. Now this data comes from providers in Excel format in non-standardised structures and so it undergoes an initial assessment. We map it in using an Excel template and this is loaded into a database and from there it is published and reviewed by providers and made available through various web applications. So the application of controlled vocabs helps deliver on most grower group and capture management authority use cases in some way including a range of topics around search and discovery, communications, providing context and trust such as assisting with access rights and building interface functionality. So as data service providers and researchers our role is as users of controlled vocabularies and when necessary we're creators and managers. So the focus of this talk is how we're using control vocabs to standardise data and the challenges that we've encountered so our current approach and solutions and would really appreciate any it's a good opportunity for any thoughts and feedback from the vocabulary community out there. So this is a simplified observation and project content data model and it highlights the vocabularies we're currently using to help standardise soil data. They include properties being observed, the procedures that are being used to make the observations and what is being measured, so the specimen. So we use vocabs to describe the results where there are certain values to describe the features and there are the relationships between the features as well as some project specific information such as organisation type and commodity. And we're also using units of measured control vocabularies in a number of ways as highlighted here. So here's just an example of the data view through the interface. You can see observations at depth which are selected from a certain location from that. So hover over and you'll see the control vocabularies used to describe the observable properties, the procedure used to measure those observable properties, units of measure and the results. So here's just a bit of an excerpt of some interface functionality which is in development. They include search, filtering, styling, observations at depth and working on trends over time and reporting templates. So now here I'm going into some of the various challenges we've encountered as vocabulary users using vocabs to describe the soil data and some of the approaches that we've taken. So one of the first questions we come up against is is it appropriate to use international terms to describe our soil data? So in this example, is it appropriate to use these terms to describe our soil features? This is a question to the Australian soils community. We did, we've used terms from Inspire and the examples that we've applied or are applying in some cases as shown in blue in this diagram. As a vocabulary user, I found it quite challenging to find and interpret vocabularies for the describer of relationship types and the roles of the relationship roles between the features. So eventually you found them and we've applied these terms from a relationship ontology which is intended for use with biological ontology and they're shown in orange on this diagram. So in regard to sampling, spatial sampling features, sampling methods, these vocabularies were quite challenging to find and when they were found, we found them under GSosaurus. The definitions of the concepts read more like sample types rather than sampling methods. So we're not 100% convinced that we've got the right vocabulary terms here yet. So a question we frequently encountered asking all the time is how do we know that the vocab we're selecting and using is persistent and trusted? How do we know it's actively governed and preferably by an authoritative body and who is an authoritative body in some cases? We've had to make a lot of best judgment calls. We tend to select vocabulary terms which appear to be from formalised, versioned and governed ontologies rather than project or experimental base. So here's an example where we've avoided using experimental OGC soil data interoperability vocabularies which are experimental, obviously. Yeah, so when vocabulary terms seem to describe the same or similar concepts, we're challenged with, which identify to select. So we're possible, we've tried to use vocabularies which are referring off from the source. Such as here, we've had to choose between an experimental term in linked data registry for an observable property when we can see that it, we can now see that it's in the environmental ontology from the chemical entities of biological interests. So we should probably be using that identifier rather than the linked data registry one. So I found in general that I found it challenging observable property vocabularies. So finding those in relation to soils. We've been using mostly experimental terms creating as part of a project that's available in little CSI-related data registry at the moment, which is really not ideal but that's what we're using for now and when we have found observable properties out there the question is that are they the ones we should be applying to soils data? So for example, we have concentration of ammonium in soil here that's from the phenotype and trait ontology. And here you can see we've used a stable concept for observable property of sulfur from e-briefs. So you can see however that the unit of measure is in milligrams per litre. So what implication does this have for us when we're applying it to sulfur concentration which we're reporting in multiple units of measure and none of which are milligrams per litre. So where vocabs have not been discoverable we have to create them and we are constantly asking the question are we sure they don't exist out there or have we just not found them? Never 100% sure. So we have needed to create cell chemical procedure vocabularies and assertive values that we receive from providers. We have gotten up with Simon and Jonathan's help through the agri-fed project, the green book. So that's strain standard for chemical methods and when we do create vocabularies at the moment we're using the Excel to LDR tool and the user interface of CSIRO link data registry. So some considerations here in this approach where we come up against this upscaling of staff, time and effort, particularly for example the green book getting that up as a control of vocabulary, selecting or creating URIs which are unique and reflect the governance arrangements and considerations of the ongoing governance maintenance and support and including potential role for research vocab destroyer going forward for us. So you'll see here the registers that have been created and the collections made available to help describe soil starter in the CSIRO link data registry. So Simon and Linda Gregory have gotten up the Australian soil land survey field handbook at current the landform classifiers, soil profile classifiers. There's a soil chemical methods I spoke about and there's a register that... Sorry. A register that was created for... I just realised, can you... You might be able to see yourselves on the screen there. The register we created for EDU. So created and maintained by Australian higher education researchers and registered for ourselves Federation universities. So note the registers and the URIs create, reflect the governance arrangements. So some challenges in applying soil chemical procedures when we got that book up as a control vocabulary and data comes in Excel format from data providers, not necessarily using the standard terms. Here are some of the challenges we encountered. So on the left, this is the kind of information you see from a data provider. Here they've got more... Example one, they've got more than one code. So the pH example, we've had to create a generic procedure code. The ammonium one at the top, the provider clarified that it was indeed 7C2B. So I've had to go back to the provider. Should we have included species on the end of those procedure codes? We're not 100% sure, but that's what we've done. So example two. This is an example where you get multiple carbon procedures coming in and it's just not clear what type of method is used or whether it's a green book code. So again, we've had to go back and clarify with the provider. So this is challenging. It's not always possible with historic data to get that clarification. We've got combinations of procedures and calculations within the one term. So we've had to create an identifier with the description that indicates this procedure is a calculation case-based. This is an identifier saying that the procedure is a calculation based on an underlying method. So example four is it's just not a green book method at all. So we've had to create a new vocabulary for the Bromfield method. And then example five, we've got extensions of the green book. So where the method was applied a different way to what it was intended and extended out. So some challenges we've encountered when applying the Australian standard for soil profile descriptions. So we've just begun this work. So describing soil profile data. An example, so the main challenge is that observable properties of the green book, of the yellow book, which are up, the definitions tell us something about how the observable property was actually measured. In example one, it says that field texture was measured by using the Northgate triangle. So it implies the method in there. And we have results come in where you don't know the method that the provider used. And some of the results are not classifiers identified in the yellow book. The example of colour, for example, they say the procedure is kind of implicit in this observable property of colour as well. They use a monsel colour system. We get results in that all results that come in are not in accordance so far with the monsel colour system. So how do we apply these observable property identifiers from the yellow book and the classifiers? So that's very much a work in progress. It's kind of the most recent work we're investigating. So the question is do we need to create identifiers for procedures? Do we need to create new observable properties to describe the data from providers which do not match the yellow book observable property or measured in a different way? We've got the classifiers of the yellow book, which are up, but we have data coming in which use classifiers or assertive values that do not match the yellow book. So we have them to get up, classifiers there. And just some more random thoughts for Australian soil scientists, chemists, ontologists, is it's quite a challenge linking the observed properties in some cases to the species for the green book procedures. For example, extractable copper, this method 12A1. So should we be associating it with copper cation, copper group molecular entity, which one at the moment we're doing this in our database? The definitions for the procedures do not exist. The definitions within the green book are really lengthy. They're pages long. So that's a role potentially for the governing body of that and authors of their green book future versions. And so appropriate governance and extension of this vocabulary is a question. And some more thoughts for, in particular, regarding the Australian soil and the yellow book is... So what is being measured? So the species are not available as controlled vocabularies. And in most cases or in a lot of cases we need Australian specific definitions, for example, for which are available but just not as controlled vocabularies. They're buried within the yellow book. So the definition of the species gravel, clay, sand, silt. And this slide's just really an ending slide of recognising that we don't exist. This work doesn't exist at all in a bubble. So we need to remain informed of the work and the roles of the various national states, territories, their partners, the global community. So here are just some to paint the bigger picture. And there's all the... Thanks to the contributors, these are the farming systems and catchment management authorities data that we've been working with and the university partners on other research partners on vans, which I believe actually some of them might be on the call, Brandon and David. So Andrew from FEDGE needs on this call as well. So any questions? That's it. Thanks.