Getting INSPIRE'd with Hale Studio
This is a modal window.
Das Video konnte nicht geladen werden, da entweder ein Server- oder Netzwerkfehler auftrat oder das Format nicht unterstützt wird.
Formale Metadaten
| Titel |
| |
| Alternativer Titel |
| |
| Serientitel | ||
| Anzahl der Teile | 295 | |
| Autor | ||
| Mitwirkende | ||
| Lizenz | CC-Namensnennung 3.0 Deutschland: Sie dürfen das Werk bzw. den Inhalt zu jedem legalen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen. | |
| Identifikatoren | 10.5446/43525 (DOI) | |
| Herausgeber | ||
| Erscheinungsjahr | ||
| Sprache |
Inhaltliche Metadaten
| Fachgebiet | ||
| Genre | ||
| Abstract |
| |
| Schlagwörter |
FOSS4G Bucharest 201959 / 295
15
20
28
32
37
38
39
40
41
42
43
44
46
48
52
54
57
69
72
75
83
85
87
88
101
103
105
106
108
111
114
119
122
123
126
129
130
131
132
137
139
140
141
142
143
144
147
148
149
155
157
159
163
166
170
171
179
189
191
192
193
194
195
196
197
202
207
212
213
214
215
216
231
235
251
252
263
287
00:00
Offene MengeComputeranimationVorlesung/KonferenzBesprechung/Interview
00:20
Computeranimation
01:01
Transformation <Mathematik>Operations ResearchDienst <Informatik>Desintegration <Mathematik>Fraunhofer-Institut für Graphische DatenverarbeitungOpen SourceImplementierungStandardabweichungOffene MengeComputeranimation
01:31
StandardabweichungOffene MengeImplementierungSinguläres IntegralGruppenoperationDienst <Informatik>Open SourceFraunhofer-Institut für Graphische DatenverarbeitungOperations ResearchDesintegration <Mathematik>Einfach zusammenhängender RaumInteraktives FernsehenAbgeschlossene MengeSystemplattformMotion CapturingNichtlinearer OperatorAuswahlaxiomService providerMigration <Informatik>DruckverlaufDatenstrukturDatenmodellStellenringQuadratzahlPlastikkarteProzess <Informatik>Produkt <Mathematik>Endliche ModelltheorieDerivation <Algebra>InformationEntscheidungstheorieTransformation <Mathematik>BeschreibungskomplexitätTextur-MappingKomplex <Algebra>StreuungsmaßAggregatzustandElementargeometrieNotepad-ComputerProgrammierparadigmaSkalierbarkeitDateiformatDemo <Programm>ARM <Computerarchitektur>TexteditorBildschirmmaskeMapping <Computergraphik>RechenschieberSchlüsselverwaltungProgrammierparadigmaDeklarative ProgrammierspracheAutorisierungSelbst organisierendes SystemForcingMinimumImplementierungMAPTransformation <Mathematik>ValiditätEchtzeitsystemSoftwareentwicklerOpen SourceProjektive EbeneCASE <Informatik>FunktionalKeller <Informatik>DifferenteKomplex <Algebra>InterpretiererKartesische KoordinatenPunktEinfügungsdämpfungHeegaard-ZerlegungDateiformatDatenstrukturAggregatzustandVariableStandardabweichungInformationTypentheorieElementargeometrieMereologieHilfesystemMultiplikationsoperatorVorlesung/KonferenzBesprechung/InterviewComputeranimation
09:45
RechenschieberHypermediaAppletDatensichtgerätSichtenkonzeptMUDSummierbarkeitWeb SiteDickeZeichenketteMathematikRechnernetzWärmeausdehnungDatensatzOpen SourceDatentypInstantiierungTransformation <Mathematik>SchlussregelExplosion <Stochastik>Lemma <Logik>StellenringDatenstrukturCodeElementargeometrieMenütechnikDatenbankBildschirmsymbolRechenbuchInklusion <Mathematik>SALEM <Programm>Meta-TagLokales MinimumMakrobefehlZellularer AutomatNormierter RaumRegulärer Ausdruck <Textverarbeitung>WarpingElektronische PublikationElektronischer FingerabdruckKategorie <Mathematik>Hierarchische StrukturTermDemo <Programm>Interaktives FernsehenCLIDesintegration <Mathematik>Textur-MappingSinguläres IntegralÄhnlichkeitsgeometrieSondierungFunktion <Mathematik>Open SourceKonfiguration <Informatik>SichtenkonzeptPhysikalisches SystemCASE <Informatik>TermHilfesystemStandardabweichungKurvenanpassungKategorie <Mathematik>Autonomic ComputingTypentheorieCodeOffene MengeTransformation <Mathematik>Mapping <Computergraphik>FunktionalRechenschieberGüte der AnpassungIntegralExpertensystemLeistung <Physik>MultiplikationsoperatorDefaultSondierungÄhnlichkeitsgeometrieMengeCodierungRechter WinkelBildschirmmaskeWasserdampftafelDelisches ProblemComputeranimation
15:48
SondierungFunktion <Mathematik>InformationLesen <Datenverarbeitung>Open SourceDifferenteVorlesung/KonferenzBesprechung/InterviewComputeranimation
16:08
ExpertensystemOrdnung <Mathematik>Strategisches SpielSoftwareUmwandlungsenthalpieImplementierungSoftwareentwicklerCASE <Informatik>AppletBitVisualisierungVorlesung/KonferenzBesprechung/Interview
Transkript: Englisch(automatisch erzeugt)
00:22
Getting inspired.
05:57
Yeah. All right, great. So a common problem is data transformation.
06:01
This is faced by over 70% of Inspire implementers, which is mostly because of transformation complexity. When you're moving from, let's say, a shapefile to a complex GML file, which is one of the most common cases, you're essentially moving from a Tableau form to a very complex nested form. So the schema mapping has to be correct. That takes a lot of time.
06:20
The transformation in itself has to be executed correctly. And once it's done, you have to validate the transformed data too. So a brief history on HAIL Studio. This is our tool that's specifically here to deal with the Inspire implementation. It was developed in 2006, 2011, with the help of Humboldt. That's also a part of the acronym,
06:42
where actually the full form is the Humboldt Alignment Editor. We have over 1,000 active users and 5,000 downloads per year, and we've had major deployments in the city of Hamburg, with the armed forces in Germany, with Prisma Solutions in Austria, and many others. This is possible thanks to the contributions of Fraunhofer GeoSolutions in Italy
07:01
and Epsilon Italia and many other organizations. But what exactly is HAIL Studio? So it's an ETL tool. There we go. So it's an ETL tool that's used to explore complex models and data.
07:20
You can author efficient transformation mappings with this declarative mapping paradigm, get real-time preview and validation publishing as shown in the map at the bottom left corner. It provides interactive documentation, which actually means that you can see how a project was done, what kind of transformations were done, and how you can even play around with the mappings if you feel free to, if you want to do so.
07:40
It's an open-source platform, of course, so you can really reap the full benefits of it with the custom developments. So the declarative mapping paradigm is something, okay, this is really not cooperating with me. So the declarative mapping paradigm that you see here,
08:02
it allows you to see what your data looks like before the transformation, when the alignment is actually done, how it's done, and what your data is gonna look like once you move on to it. Now this paradigm is completely independent of the data complexity. What this essentially means is that no matter how complex your data is, you can use the same alignment functions.
08:21
So it's easier to author and reuse with documentation. It's fast and scalable, and you can apply it to any structure and data format since it doesn't depend on the data complexity. Now we're gonna move on to a use case, which is the AAA to Inspire use case. So the AAA standard is a German conceptual application schema created in UNML.
08:42
It's very complicated, even more so than Inspire, and there was a lot of variability across the 16 states that wanted to implement it. This is due to their different interpretations of the standards, and they had a different tech stack and legal requirements, which meant that their baseline with this AAA standard, their starting point was very different, and they all wanted it the same. They all wanted to move on to the same end goal,
09:03
which was Inspire compliance. Now there were also geometry type mismatches to deal with. There were many split and merge operations, and when you're doing that with a lot of complex data, this leads to a loss of information. We used Hale Studio, and the key here was the interactive documentation
09:21
that was reused by various German organizations, so they could actually share the solution that they had, and as you can see on the slide over here, they could access the kind of mappings that were used for certain functions. The declarative mapping paradigm also dealt with the data loss that was caused with the split and merge functions, but while it's one thing to talk about it,
09:42
it's another one to actually see it happen. So I'm going to do a little demonstration for you. And let's get to Hale Studio first. Yeah. So over here, we can see the default view of Hale Studio. Now this represents the source data. This here is the target data,
10:01
and that's everything that's going on. So this is one transformation that's already been done. Sadly, I can't actually do that, because A, I don't know how to, and B, it takes a lot of time. It requires a lot of expert knowledge, but what I will do for you is show you how to transform one property to another according to the Inspire standard.
10:20
Now we can get a good idea of what's actually happening within the data set by expanding this. So now we can see very nicely what's going on in the source data over here. Now to upload your source data, you have to first upload the data schema, and then you upload the data within itself,
10:42
and this is done by uploading the schema from the same data set that you actually want to upload. This reduces any chance of upload mistakes, and the target data schema, well, we provide presets for that. It's built from the ground up to support the Inspire open standard. And here we can see that we have a code.
11:03
Now we want to see what's actually happening within this, so we want to see the occurring values in this. Just going to click refresh. So we see that there are four codes. Now these four codes represent the type of water body it actually is. So is it a river, a channel, or a canal?
11:22
And we want to map this onto somewhere in the target data set into the origin. Now this requires implicit knowledge of the Inspire data, which you can find in documentation, and as for the specific knowledge about these codes, that's usually present within the documentation of the data set in itself. So this is OS Meridian 2 data,
11:42
and you can find everything relevant to that in the documentation. Now through that, I know that 6232 is a canal. Everything which is a man-made water body, of course, and everything else here is natural in the form of rivers or channels and so on. And here we have the origin property,
12:00
which is either man-made or natural. So we want to use this data to move onto origin over here. So for that, we already have code selected here, and we just click origin over here. Now what we want to do is actually set up the mapping in itself, which is done by this neat little button. And here we have the classification.
12:23
So we'll do that. Now it shows the type, what it's going to, the source and the target is pretty clear. So to populate the source values, there's just a button over here, which is going to populate it with the four codes. So there we go. Now the target values are not present because we want to map that right now.
12:42
And for that, we just double click on it, and enter a target value. So like I said, we know 6232 is a canal, which is man-made, and by definition, everything else is natural. So either I can double click on it, which I won't because I'm lazy,
13:01
and just assign a fixed value, which is going to be natural. Okay, so now we have this transformation set up, and we just finish it.
13:21
And now we can see what's actually happened here. So we moved onto code from origin, and just by hovering over it, we can see what exactly it did. So we can also see that it's automatically documented for each of the transformation mappings, which gives you a very good idea, and this is what actually helps with effective reuse. Now the next step is,
13:41
well, where are we actually moving onto with Hail Studio? Where do we want to develop it? What's the future looking like? So, now we have two things planned out. Well, let me see here. Who actually uses Hail Studio? Okay, that's good.
14:01
I'm happy to see that. Now we also need, well, we want to crowdfund ideas to what's the next functionality that we develop out here. So who here wants to actually help figure out what kind of functionalities to implement? I'm not asking you to code anything, just what you would prefer in terms of functionality. Can I see a quick show of hands for that as well?
14:25
Yep, so we want to see, well, if you want to help us develop a certain functionality for Hail Studio, if you want to provide any input for that. So just a quick show of hands in that case. Okay, thanks. So we have two things planned out so far. Well, there's the Hail 2.0,
14:41
which offers a Live Transformer API, and this is a significant improvement over our standard, already powerful CLI API that enables integration into your infrastructure. So this is very useful for systems integrators, and it's highly interactive and provides you with the option of live data manipulations. There, you can see an example of that in a slide there.
15:01
That's what it's going to look like. And the next one is autonomous mapping suggestions. These demonstrate how others create similar transformation projects and apply best practices. So when you're working through a project, Hail Studio will give you an idea of what kind of mapping you can actually use. So this is useful for, well, the actual users of the software,
15:20
and, well, for beginners especially, because it really decreases the learning curve. With the help of this, we estimate that you can save up to 90% of effort. Now, thank you very much for listening, everyone. I'll be passing out a little survey later on, so it'd be very helpful if you fill that out, and we can decide which functionalities to move on with.
15:41
And I'd also like to open the floor for questions. Q, I have some question.
16:02
I do have one. As it's an ETL, you have a driver to read different data sources. Which technology do you use? Is it based on GeoTools, as it's a Java component? Honestly, I don't think I can tell you that. Like I said, I'm not a technical expert in the development of this, but I can definitely get back to you on that. To my knowledge, we're using a photon engine,
16:21
but the specifics of it I can't really offer you. I'd rather not give you blank answers. Thank you. Yeah. Thanks a lot for the nice overview of Hail. I think it's a very important piece of software for those who implement Inspire in Europe.
16:42
My question is regarding alternative encodings, because I know that currently it is possible to export GeoJSON inside. I don't think there is support for GeoPackage. What would be the strategy we transform and the community around to have in order to address those? So right now, GeoJSON is all the craze
17:00
because it's very good with the data visualization aspect, which happens to be a common problem of the Inspire GMO data in and of itself. So to actually move on to a GeoPackage, while we have to look into concrete use cases, what's existing so far, and how we can improve on certain capabilities, and what the Inspire implementation falls short of.
17:22
So if we can find something that GeoPackage is, well we talked a little bit about it in your presentation. If we're talking about a use case where GeoPackage concretely addresses a detriment of the Inspire implementation, that would be of strategic value to us, and room for development over there.
17:41
Other questions? Then thank you very much. Yeah, thanks a lot.