 Yes, and my name is James Hulfray, I'm the Information Manager at Health Direct Australia. And I'm going to show you about our Australian Health Dessaurus, which I manage, and how I'm going to show you what it looks like and how we use it, and what are the current developments we're going to be applying it to as well. Okay, so first of all, who is Health Direct Australia? We were established by COAG in 2006. It was actually Tony Abbott when he was the Health Minister in the Howard government who set us up to try to release the burden on Medicare by providing telehealth services. We were known initially as the National Health Call Centre Network, and we ran the after hours GP helpline service and nurse triage. And it was only about five or six years ago that we actually moved into digital health services, and we took over the management of the Health Insight website. So we have a mandate to work with the federal and state jurisdictions to address their key priorities and challenges within the health, aging and the social services sectors. We have a number of products and services. The main one is our Health Direct website, which was initially known as Health Insight. This site contains a lot of our health information articles, which we produce ourselves. We have links to partner content as well. So we have 30,000 links from peak health bodies, such as Asthma Australia, Heart Foundation, Beyond Blue, and organisations such as that. So we act as a portal to their content. We have a symptom checker service where you can see how the severity of your symptoms. And you can also go and find services in your local area, such as doctors, hospitals, emergency departments, and also specialists. And we also have medicines information, which I'll get onto a bit later on. Another website is the pregnancy birth and baby website for those who are pregnant or are having babies. That also contains health information articles and links to our partner content as well. We also manage the MyAgeCare website on behalf of the federal government that has similar services on the site. We have a care gateway service as well for those people who identify as carers of others who have disabilities or are frail or are aged. And another service is the national health services directory, which is used for our service finder. This is where we have the database of all the GPs, pharmacists, emergency departments, et cetera. Now, the main reason I'm showing you those sites and products is because the source that I manage is used on all of those sites and helps to contribute to it. I don't know if you can see that slide. The video of the faces is covering over the vital statistics on here. The thesaurus is a medical and health and human services-related thesaurus and has to be human services-related to Kata for the aged care and the care gateway. The statistics on the site just covering up on the website, on the video there. There are basically over 5,000 concepts in the thesaurus. There are 27 concept schemes and there are nearly 9,000 synonyms or alternative labels. The concepts reflect the current Australian health and human service environments. And we use it because it's integral to the search and discovery of our content on our websites and services. It's used to improve the user's search experience. But the main thing about it is that it is consumer focused. Most health terminologies and taxonomies are clinically related. So therefore the medical profession or for researchers. But we have a consumer focus and that causes raises a number of challenges for us. So I'll let you have a look at the thesaurus at the moment. So on the left hand side of the screen you can see our concept schemes. And you can see we have ones such as anatomy. There's chemicals and drugs. We put our medicines information. We have diseases and disorders. Equipment and supplies. Facilities. If you work your way down, we have symptoms and techniques. Things like that. So they're the main concept schemes within the thesaurus. If I select one of those schemes, you can see on the right hand side there, I've selected the diseases and disorders scheme. And I've worked my way through the hierarchy through digestive system diseases, down through digestive system cancers. And the one I've selected on the screen there is gastrointestinal cancers. That's one in orange. It has a number four after it, which means that there are four other concepts, narrower concepts below that. Why don't you select one of the concepts? You're actually looking at the software that we use. We use pool party to manage the thesaurus. And you can see what it's showing you at the moment is the SCOS view, which is the simple knowledge organization system view, which is basically the broader, narrower and related concepts relationships. We can also match it to another concept in another reference set, such as one at the moment we've got there is a reference to mesh, which is the medical subject headings, which I'll tell more about that later on. On the right hand side, you can see some of the attributes and such as the preferred label and the alternative label here. And you can see we have GIT cancers as a synonym. You might also notice there's another little tab just up above there called clinical relationships. If I click on that, you get another screen here, which is basically this is our own custom schemas that we have developed ourselves, basically our own ontology. And the one, the concept I've selected on the screen is ME tryptoline, which is a medicine and active ingredient. But with the concept scheme, the custom schema we've created our own relationships. So we can make a relationship between this active ingredient and the conditions. And also we can do it to symptoms. It's used in surgical procedures. And also used in diagnostic procedures as well. So we can fill that out when making those relationships. We're not really using that at the moment. But if you look on the right hand side, we have the attributes based on this custom schema. And this is where we've mapped it to other data in other other data sets. And here you can see we've got the AMT ID. The AMT is Australian medicines terminology, which is a naming standard used for the pharmaceutical systems in Australia. We've also got a CAS number, which is a chemical number. We have a drug bank ID as well. And you'll also notice there's one there called the BEAS criteria. Now, that doesn't tell you if that ingredient is used in your fosters or your toolies new. What it refers to is the list of ingredients that are inappropriate to prescribe to older people. And so we've just got some Boolean logic there. And we also have down there the pregnancy category. You can see we have our pregnancy category C. And I think from memory of the C, it means that there is a risk to use this medicine if you are pregnant. And I'll show you how we're using those attributes a bit later on. Now, the thesaurus was originally developed by the Department of Health and Aging in about 2000. And it was based on MESH, which is maintained by the US Library of Medicine. They use it to classify their big medical databases such as PubMed and Medline Plus. And we took over the management of it about four or so years ago. And there's a number of different criteria we use to update and maintain the thesaurus. We've actually made quite a few changes since we've taken over that management of it. And we use analytics to help determine what are the new terms that we need to add to the thesaurus. Because we're consumer focus, we can analyse what are people searching on the website to see does this term exist in the thesaurus? If not, we should add it in as either a pref label or an alternative label. We also respond to current news developments. For example, when the Zika virus outbreak hit Australia a couple of years ago. It was the first time in Australia, so therefore we didn't actually have that concept in the thesaurus. So we added it to the thesaurus. And we also do big environmental research analysis for specific purposes. And when we launched my aged care website, we needed to go and analyse what are the we put in a huge corpus of aged care content and used our software to analyse the terms that we used. And therefore we can identify concepts that needed to be added to the aged care schemes. There are also a number of elements of other classification encoding systems here. These are mostly the clinical ones I was talking about before. We still need reference to them. So there's a reference here to ICD-10, which is an international classification of diseases. That's a World Health Organization classification system used for reporting purposes. We've got the DSM-5, which is for mental disorders. An example here is that the concept that we know of as autism, when DSM-5 was released a couple of years ago, the actual term now used by the profession is autism spectrum disorders. So that's another cinnamon we added to our thesaurus. We've got the anatomical therapeutic chemical classification scheme, which is about body systems and the medicines that work on those body systems. I've got MISH down there as well. And the big one is SNOMED, which is the global classification system used for health records. So we're actually going through a process of mapping to eat at the moment, and I'll give a bit more information about that a little bit later. So how do we use the Australian health thesaurus? We mainly use it for classification purposes. We have two types of main content. It's our local content, which we develop ourselves, we write ourselves, and we have an automatic classification process where we apply concepts to it. And we also have those 30,000 partner links as well. So we harvest the links in and the metadata from the source code is classified against the thesaurus. What that does then is allow us to use the classification in the ranking in the search results. So if you use a type in a search, if that term is one of our thesaurus terms, then it will get a number one ranking in the algorithm. If that term is then in the title, it gets the second ranking. If it's an N in the description, it gets further down the order there. So it's quite important to help with the relevance in the search results. It's also used for the auto-suggestion. So when a user starts typing into the search field, after a couple of letters are typed in, it'll then pull concepts from the thesaurus as suggestions for them to select. We can also display some more contextual items on the websites. For example, if you were on a page about asthma and that page has been classified with the concept asthma, then we can display the symptom checker flow for asthma because it's also classified with asthma as well. So based on that matching, we can contextually put pieces onto the website to provide more useful information for the user. We also have used the thesaurus to manage our medicines data. This has been a big piece of work. What we have done with that is we've developed our own terminology service where we use the thesaurus as the main control list for medicines. So we'll add in the active ingredients or medicinal products into the thesaurus. And as I showed you before, it has a reference to the BEAS criteria and also the TGA, the pregnancy database. We also pull in other data from other sources such as the therapeutic goods administration, the Australian medicines terminology and also data from the PBS. So the medicine is actually listed on the PBS. And as of a couple of weeks ago, we actually launched images of pills which we're pulling in from Guildlink which is the commercial of the pharmacy guild of Australia. So within the terminology service it's all in RDF format and we can just pull that data onto the website based on a user search. So here's an example of some of the information that's on a page that is presented to a user. You can see here on the top right hand, the top left hand corner here we've got the brand name and we tryptoline. You can see that there's a little box there that says active ingredient over 65 years of age. If you're over 65 years of age there may be specific risks and recommendations for use of this medicine. Please discuss with your friends. Now that flag there is only coming from that attribute within the thesaurus where we've got the bilion logic for true under the bees criteria. You can also see we're pulling other information there from the TGA about the pack, the type of pack and how you can store it. There's a link there to go to the PBS. I mentioned about the medicine itself, the appearance of it, its dosage form, its root of administration and of course right in the middle there there's an example of some of the images that we're pulling in. But if thesaurus is being used as the main control is for this, that is the manages all the rest of the other data. I mentioned before about the mapping to snow med. The Australian medicines terminology as of last year it's now become a subset of snow med. We've already done that mapping already. What we want to do now is map to the other conditions, the procedures, the techniques within snow med. That will give us access to a number of other classification coding systems that snow med is already mapped to such as loink which is used for pathology. ICPC is a primary care classification system and they've already mapped to the ICD 10 which is the diseases coding system as well. So we'll be using link data techniques to try to link to other data sets so we can pull in that data to add more value to our web pages. But there's another even more important one that we're doing first and that's an example here on the page of the mapping with snow med and we're using a tool developed by the CSIRO called snapper. You can see the first two columns here are the URI of our concepts then you've got one about the next column is our prep label. Then you've got the snow med ID in the third column and the snow med label in the fourth column. So we're going through a mapping process. We're happy with those mappings. We can then add the snow med ID as an attribute to the the source terms. The main use case we're starting off with though is that we're going to be integrating our app with my health record and you may have known that in the last budget the government announced when the every Australian will have a health record and allocated a certain amount of money to that as well. So we're just developing a prototype at the moment and you can see on the left hand screen here we've got a access to the view my health record. Once your health record comes up you can actually see what are the medications you've been prescribed with, what are the allergies your doctor has said you have, what are your recent Medicare benefits and we're trying to integrate as many of our services into this my health record. On the right hand side is a list there of some of those medicines that you've been prescribed with. The idea that what we're trying to do is if you are not sure what that medicine is you can click on that and it will display our own content pages which can provide more context to a user. By doing the mapping to snow med and the other conditions procedures the idea is that if you can if you see a procedural condition on your health record that you are not sure of what it is we want to be able to get the user to then click on that term and to expose and display our own content pages. Other things that we're doing is we're actually introducing natural language processing onto our websites now so the thesaurus is going to be used as an integral to that and also with the some chat bots and web chats as well so thesaurus will be powering those other services. So I think that's all my time that I've been allocated. I think it's my 15 minutes are up so if you've got any other questions later that I can't answer today please contact me on the email address on the screen at the moment.