11:27:54 >>KEVIN: Thank you.
11:28:00 Great, yes as curmudgeon my name is Kevin Worthington and I am the geospatial data manager at Colorado State University libraries.
11:28:13 Before I started my role here, I worked for a research group at UC Berkeley and my role there was really working on establishing a system for managing and sharing, streaming air quality data through developing a real-time calibration system.
11:28:22 And this was built around the Jango pipe on library, and that is actually the foundation of the software that I will be sharing here today.
11:28:37 And before that, as mentioned, I worked at the scholars portal and one of the platform's word was God's GeoPortal that Amber just demonstrated.
11:28:46 And today I will be presenting prototype application which I developed for use in harmonizing the disparate collections of GS geospatial data at DSU.
11:28:51 I believe this application could be used at other institutions as well, and it has been designed with that consideration.
11:29:11 The background image that you are saying is what this prototype application looks like, it was really modeled after scholars GeoPortal, and I will be walking for the journey of putting this prototype application together.
11:29:23 It is my hope to work with interested members of the academic geospatial data community in furthering the development of this open-source platform to really assist with sharing geospatial data for research and application.
11:29:29 So my position here really started with like a thorough investigation of geospatial data sharing platforms.
11:29:33 So that we might adopt one of them ourselves.
11:29:44 And basically there were 5 or or five, platforms in use, but the conversations with users, there have been an apparent shift of migration towards two main platforms.
11:29:50 So those would be ArcGIS hub as well as Geo blacklight.
11:30:01 And for academic libraries, Geo blacklight really stands out with many installations as indicated by these blue dots.
11:30:12 And truly a lot of migrations coming from open GeoPortal, which is really a cool platform but they themselves are really backing Geo blacklight as well.
11:30:30 But there are many reasons why Geo blacklight drilling has a large following, has a great set of features, but from our get scholars portal I have really come to rely upon many of the features that aren't part of Geo blacklight.
11:30:35 So I will just highlight some of the features that are a part of Geo blacklight platform.
11:30:41 Of course responsive design, people are accessing websites from the phones or their tablets as opposed to the desktop computers.
11:30:48 So that has to work really well.
11:30:48 Easy to update, customize and theme.
11:30:55 If you're looking for adoption, the ability to have others add their own sort of University colors or logos is critical.
11:31:03 Support for many different geospatial data formats.
11:31:10 There's a whole slew of them and Geo blacklight supports so many of those.
11:31:20 Also search engine optimization, that's the ability for people to find a layer in your map collection using the favorite search engine.
11:31:27 It is really important that those search engines can dig into your collection and index face.
11:31:28 And of course the special search piece.
11:31:43 The ability to find map layers without using like a conventional text search where you are reliance upon knowing the name of the place and in some cases like place names do change over time.
11:31:43 And so being able to do that special search is really helpful.
11:31:49 And all these features are really indispensable.
11:32:05 My approach was to build upon Geo blacklight substrates, adding in many of the features that I have come to expect with the scholars GeoPortal as well as include some of the features that identified my time while working at scholars portal.
11:32:09 So my approach was really this hybrid approach.
11:32:29 So taking the Geo blacklight search system, so a lot of time was spent in developing doubling core metadata my schema (away from mic) (sp?), and that is ingested into this Apaches solar searches system.
11:32:39 It's the same search engine that powers date of births and using this Geo blacklight schema as part of the platform that I looked to start with.
11:32:52 And so building upon that and really adding in the generation tool piece, because that is a super important piece, and I will go on to show what that looks like.
11:32:56 As well as adding into some of those enhanced to the front and that I mentioned.
11:33:07 Of course having multilingualism is really important, and there has been work done out of the University of British Columbia in adding multilingualism to Geo blacklight.
11:33:07 So that is cool.
11:33:19 I intended to have that and have built that from the ground up as well.
11:33:19 In the platform that I will be sharing.
11:33:39 Multilayer map viewer, just like scholars GeoPortal, to be able to see those different map layers together really helps tell that story, and I would argue that it helps at the context that you might need to determine if the data is in fact like sufficient for your needs and to show different data layers together makes that possible.
11:33:43 The ability to subset data.
11:33:52 Some data sets are huge, and to be able to download a smaller set of those is I think important to help with the research process.
11:34:11 The data set support pieces just talking about the fact that oftentimes in geospatial data layers are released, the releases kind of like a bundle, and so you have maybe like the parent would be the folder and then within that folder you have a whole bunch of different layers.
11:34:20 Maintaining that hierarchy is really important.
11:34:30 In a case like first features at risk, which is a lot of the data that comes up in the platforms that are work out, it allows you to see all the different species at risk in this like parents category.
11:34:45 Adding in map behavioral statistics, something that is really foundational to scholars GeoPortal and its ability to shows like what layers of people accessing and are they interacting with them prior to downloading them.
11:34:49 Really important stuff.
11:35:12 And even just to mention Leslie about layer size of sorting, the idea of sometimes map data layers cover the whole expanse of Canada or the US, and if you're just looking for something with like higher fidelity, it is going to be much smaller map and you can sort by map size if you are doing just a spatial research search for a particular area.
11:35:23 And I wanted to emphasize the fact that I have added attention to the change management and data visualization piece of this.
11:35:36 So what I mean by change management here is the idea that the data that is being pulled into this collection are coming from elsewhere and often times one would want to augment that data by adding an additional information to help with the clarity of that.
11:35:44 The issue without being next time you do a it is going to blow your changes out of the water into be able to keep track without should be imported.
11:36:01 The data visualization ID is to maybe you want to change the style of the map layer, and this is something that is really our (term unknown) has put a lot of effort into and it is definitely not that complicated for a specific data set to make that possible.
11:36:15 So moving along, this is really the stack that was put in place to help make the prototype application that I will be presenting possible.
11:36:30 It is religious built upon open source software that is in fact readily available to all from my work at UC Berkeley, I really worked a lot with the Jango Laburnum as part of Python.
11:36:58 That is really because it is able to be so tightly knit to your database that you can establish the database columns and information that you stored in there without having to deal with the granularity that is often the case in working directly with the database and then trying to update your interface.
11:37:07 It allows flexibility, and the research that I worked with him and that was really important because sometimes you don't know what you know until you getting into it and you're like you needed to add this, and information comes abound over time.
11:37:07 And so having that flexibility is really important.
11:37:30 As well as this Apache searches system, to have an advanced search engine that allows for just a sort of like anywhere searches as well as more specific type pieces, the special search, fascinating, the idea that you can have indexes that cross different keywords or categories and how people search with that is really important.
11:37:42 Then on the front end, there's a whole slew of different plug-ins that have capitalized upon as part of this leaflet web map viewer, really popular map viewer.
11:37:48 And then, you know, jQuery has been around for a really long time.
11:37:51 There is some critical plug-ins that work as part of that.
11:37:52 A day.
11:37:59 Or, color.
11:38:00 Or, sidebar and the sort of things.
11:38:05 And then of course using bootstrap which is very common across the web as far as layout, styles, and icon designs.
11:38:16 We will start with kind of like birds eye view of what this platform looks like, it really begins with harvesting piece.
11:38:34 And so as you might imagine, each platform that exists that houses geospatial data or in the case of content DM, like my not but has historical maps for instance and a lot of institutions, having a specific harvester to pull in that data is important.
11:38:53 And so I've established a bunch of different routines that go through and they can be managed through a web interface to make it less of an arduous task.
11:38:53 Once you set it up, you are kind of good to go, and so I have a few of those that are in place.
11:39:07 What they do is they really pull the different like application protocol interfaces, which is just down they are managing their data and pulls it into a common stream.
11:39:21 And then ingest that into the database, which is really driven by this Jango application, and that is to rely in this other that you want to add into the record.
11:39:45 When you are content with what you have as part of your collection, you've done the cure ration aspect, you would move then into the ingestion phase, and you can decide what you want to ingest, depending upon you know, the collection or so forth and that you would want to push into your public facing site.
11:40:04 And then it would be curable from the front-end system and just like with those scholars GeoPortal do would want to have a bunch of different environments for this, like I'm testing mostly like a local platform I would push the announcement to like staging applications and so people can see that, and then finally into production for public view.
11:40:22 And then I wanted to also show that there is this a basic of geo-referencing piece of that ability and because sometimes content to light historical image doesn't actually have geospatial data, and was still want to provide that special search apiece.
11:40:31 So that could be done from an administrator or it could also be done from the community input, and that would be flanked within the record itself, depending upon who it was who is actually making that change.
11:40:49 In this at the bottom, just kind of speaks to the detection of changes as data is harvested as well is like the tracking of the generation efforts, and something that I have been thinking a lot about, I have some ideas on how to resolve it.
11:41:05 This on-the-fly error detection notification, resolution, a long way to say sometimes data layers that are changed between harvests, so generally speaking, a harvester might run every week.
11:41:16 So within the week, if a change was made and maybe that layer goes down or something, you will be good to know about that and resolve that so if someone else were to access it during the week, that would be resolved as well.
11:41:27 So I will now move on to what this interface looks like.
11:41:38 So very much like scholars GeoPortal, it starts kind of to be not super intimidating, you have a search form piece at the top.
11:41:38 This one is a bit simpler.
11:41:46 You don't defined a map bail.
11:41:46 It just uses the zoomed in areas for the map abounds.
11:41:55 Off the bat it is already assuming you might want to do a date search type piece from that.
11:42:03 And then it has your fascinating pieces as well, and this is an area just just portal will slide over to show the results and the details for the results that to come in.
11:42:20 And there are some items that you can use on the map, these little buttons would be used for geolocation defined maybe, yeah, what you are trying to look for.
11:42:20 You can do some measurements.
11:42:29 There's some annotation as well that you can add as part of that.
11:42:42 And just like scholars GeoPortal, what you start to add stuff in and really expose the capabilities of the system, it makes it possible through the web interfaces as areas expand to show them.
11:42:57 And so on the left you can see now in the map tab as opposed to the search tab that I initially started in, and you can see that there's a couple of layers that have been added, just like scholars GeoPortal, you can drive us to show one layer over the top of another.
11:43:27 One feature that we talked a lot about but was a bit tricky to add in after the fact is this like split view, and so what that allows you to do is show kind of half of the map on one side and have it not shown on the other, and so you could slide this around to be able to show and reveal, to help with that changed over time.
11:43:31 And this would really just be as an alternative to like the slider or the transparency slider that you could use to show or hide the layer.
11:43:39 There's the table view button, so for vector data you can actually see all the information.
11:43:45 And there is also a table sort piece that you can determine how to sort.
11:44:04 One thing that the system doesn't yet have but it does not, it is not so difficult to add in and something that is really quite powerful and scholars of GeoPortal is the ability to query this information.
11:44:08 You can actually write your own application or it is like you can write some SQL that will allow you to search through this.
11:44:13 And I encourage you to look through how scholars GeoPortal does this.
11:44:19 It is really quite powerful in its ability to filter down into the data.
11:44:26 Similarly, the system allows you to limits by search downs, so you can only show those results based on what the map is looking at.
11:44:33 And yeah, let me just move along and then I will jump into the live application for this.
11:44:53 This is just, to the fact that there is a some map data layers that don't actually have geospatial information as part of them, and I've got it in a basic geo-reference her for this.
11:44:54 My mentioned this could be done through administrators and this could also be done by the general public.
11:45:10 And some of the benefits of this being like to engage with the community, improve of collections completeness, all of the knowledge transfer we worked with academic institutions and for people to know kind of what this process is like is a really powerful piece I believe.
11:45:18 So let me just show backend systems.
11:45:34 So what I pulled up on my screen right here, and this is not something that one would often see in the presentation, but you can imagine behind scholars GeoPortal and behind this prototype application is that administration piece.
11:45:49 And what that allows us to do is edit the particular data records and to actually add in our own changes without having them get overwritten on the next harvest.
11:46:03 And so this is all sort of prebuilt as a part of the J application.
11:46:03 This is why this library is just so powerful.
11:46:14 This is all pulling from a database, but yet it was driven by the Jango system for that.
11:46:14 So I'm just looking at resources.
11:46:14 There is a bunch of them.
11:46:21 You will see some duplicates, is because there is this parent to child relationship that is built into your, and I can show on the next tab here.
11:46:29 So this would be like a record that you would go in and if you wanted to add additional information to that, you could do.
11:46:36 So I keep the raw mated data.
11:46:49 I should say that this application could be used to two ways, as you do subsequent ingestion's, you can keep the old ones and it would have a date stamp appended to it.
11:47:04 Or what my intention is to do is to follow like the open geo-metadata standard where you rely upon it to keep track of your metadata changes into the tracking of that there.
11:47:13 And so you only have to keep 1 record of it, which is nice, and GT will handle the change pieces of it.
11:47:16 I did keep the mated data as part of the record itself.
11:47:21 And then you can see in this case, this would be how a child record shows up within the system.
11:47:34 And it is really like that data set level of I guess like management that I feel is quite important as part of that.
11:47:54 So going into the public interface, what I wanted to show here is the fact that it shows the different facets for the areas that you have.
11:48:07 If I can afford using like a basic text of search, I can look for in this case I was looking for a drivers trail out, this is sort of what that parent child relationship looks like where I can go and see the children of that parent record.
11:48:12 As part of it.
11:48:22 And then I can also, if I wanted to, for a record that you might notice this one doesn't have geospatial information, it just shows up here on the left.
11:48:37 But if I go into the details of it, and this is what the community sort of peace I mentioned about adding the geospatial referencing, you could geo-reference this image right away, and it would pop up a window that looks like this.
11:48:37 It has instructions on how you would do that.
11:48:46 And it's a bit formatted idly right now.
11:48:46 But here's what the map would look like.
11:49:01 And then if you wanted to add, let's if you wanted to say this is exactly where it should be, you could go and say this is kind of like this.
11:49:06 You could then save that as just a community member and that would then show up in the administration system and a moderator would approve that and that it would show up in the main search system.
11:49:25 I wanted to show this is kind of what it looks like when there's been any layers added, in this case just the 2, plus the base map piece that we edit that split view feature that omission.
11:49:25 I can go forth, I can split one image on the left.
11:49:26 I can split the other on the right.
11:49:30 As part of that.
11:49:37 And this will change color for a specific vector, files.
11:49:42 And really, I think kind of a cool feature but of course there is also that transparency slider you could use as an alternative to that system as well.
11:49:48 I will jump back to my presentation right now.
11:50:17 Oh, just before that I will say I push the code to GT yesterday so it is available for others that are interested in kicking the tires of this application, and without I will shout out to so many people who have just been amazing in giving me their time and expertise as I develop the system and a sort of drill them with questions on
11:50:23 how to move forward and and in different steps.
11:50:23 I will end off with sort of some next steps.
11:50:27 There's short, medium and long.
11:50:31 Type in ambitions as part of this.
11:50:41 Here's the Purell for this repository as well as if you're interested in being involved, please do contact me.
11:50:49 It would be great to hear your thoughts and how you might see this application evolving.
11:50:50 So with that, I will stop and thank you all for your time.