Integrated modeling with k.LAB... and QGIS
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 |
| |
| Serientitel | ||
| Anzahl der Teile | 266 | |
| Autor | ||
| 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 | 10.5446/66447 (DOI) | |
| Herausgeber | ||
| Erscheinungsjahr | ||
| Sprache |
Inhaltliche Metadaten
| Fachgebiet | ||
| Genre | ||
| Abstract |
|
FOSS4G Prizren Kosovo 2023152 / 266
10
17
23
44
45
46
47
48
49
50
52
53
80
84
85
91
110
116
129
148
164
167
169
173
174
181
182
183
186
187
199
202
204
206
209
215
241
248
265
00:00
Formale SemantikComputerspielSystemplattformMathematisches ModellSchnittmengeTermTypentheorieComputeranimationTafelbild
00:44
VerschlingungDienst <Informatik>Modul <Datentyp>Mathematisches ModellKomponente <Software>ProgrammierumgebungSystemplattformDokumentenserverMathematische ModellierungMathematikSichtenkonzeptRegistrierung <Bildverarbeitung>RechnernetzInstantiierungProdukt <Mathematik>Mathematisches ModellSchnittmengeMathematische ModellierungAnpassung <Mathematik>BestimmtheitsmaßSoftwareEinsLeistungsbewertungSystemplattformRechter WinkelProgrammierumgebungMereologieFlächeninhaltQuick-SortNatürliche SpracheGenerator <Informatik>Digitales ZertifikatCoxeter-GruppePhysikalisches SystemVirtuelle MaschineProjektive EbeneMAPDatenstrukturDienst <Informatik>EinfügungsdämpfungHilfesystemServerNeuronales NetzLuenberger-BeobachterBereichsschätzungMaschinenschreibenAuthentifikationForcingSoftwareentwicklerPunktAutorisierungComputerarchitekturIntegritätsbereichElektronischer ProgrammführerDatenfeldEigentliche AbbildungWasserdampftafelComputeranimation
04:50
ErhaltungssatzIkosaederVirtuelle MaschineMathematisches ModellAnalysisSchnittmengeModul <Datentyp>Desintegration <Mathematik>Mathematische ModellierungStellenringSampler <Musikinstrument>SystemplattformKontrollstrukturBitSchnittmengeKomponente <Software>Baum <Mathematik>Natürliche ZahlSystemplattformGamecontrollerStatistikProjektive EbeneSoftwareRechenschieberInformationDigitales ZertifikatProgrammierumgebungDivisionSchreib-Lese-KopfBetriebssystemFlächeninhaltBildverstehenFunktionalNeuroinformatikMereologieKreisflächeComputeranimation
06:46
AbfragePhysikalisches SystemNichtlinearer OperatorMathematisches ModellPhysikalisches SystemGenerator <Informatik>SchlussregelProjektive EbeneNichtlinearer OperatorYouTubeKartesische KoordinatenStrahlensätzeCASE <Informatik>DifferenteDatenflussLuenberger-BeobachterMAPTrennschärfe <Statistik>AbstandTypentheorieIdentitätsverwaltungPixelAbfrageComputeranimationVorlesung/Konferenz
08:07
Formale SemantikNichtlinearer OperatorLuenberger-BeobachterFormation <Mathematik>TopologiePunktDatenflussBitComputeranimation
08:34
Mathematisches ModellLokales MinimumInverser LimesNatürliche SpracheBildschirmmaskeComputerspielSpezifisches VolumenComputeranimation
09:02
Spannweite <Stochastik>Mathematisches ModellLeistungsbewertungÄhnlichkeitsgeometrieReelle ZahlRechter WinkelExogene VariableGenerator <Informatik>InformationsspeicherungInformationPhysikalisches SystemProzess <Informatik>Demoszene <Programmierung>Computeranimation
10:11
GruppenkeimReelle ZahlDatensatzPunktSelbstrepräsentationOpen SourceMathematische ModellierungProgrammierumgebungGenerator <Informatik>TabelleGüte der AnpassungDigitalisierungComputeranimation
10:45
Mathematisches ModellAbstandÜberlagerung <Mathematik>Mathematisches ModellCASE <Informatik>SchnittmengeMultiplikationsoperatorKonfiguration <Informatik>Kartesische KoordinatenComputeranimation
11:32
Lie-GruppeÜberlagerung <Mathematik>BitCodierungAbstandOpen SourceMereologieCASE <Informatik>Anpassung <Mathematik>InstantiierungZweiPlastikkarteFormation <Mathematik>Notebook-ComputerLeistung <Physik>MAPGemeinsamer SpeicherComputeranimation
12:32
Diskrete-Elemente-MethodeMultiplikationsoperatorWald <Graphentheorie>Transformation <Mathematik>Mathematisches ModellÜberlagerung <Mathematik>Computeranimation
12:51
Mathematisches ModellWald <Graphentheorie>Diskrete-Elemente-MethodePhysikalisches SystemBitKonditionszahlDatenfeldComputeranimation
13:09
Wald <Graphentheorie>Diskrete-Elemente-MethodeDienst <Informatik>Computeranimation
13:26
TabelleKlasse <Mathematik>Prozess <Informatik>MathematikMathematische ModellierungLokales MinimumMathematisches ModellResultanteArithmetisches MittelMomentenproblemFormation <Mathematik>AggregatzustandMAPSelbstrepräsentationRandomisierungMereologieSpezifisches VolumenWort <Informatik>Komponente <Software>UnordnungFlächeninhaltQuick-SortProzess <Informatik>BenutzerbeteiligungComputeranimation
15:10
MathematikAggregatzustandEreignishorizontTemporale LogikFormation <Mathematik>Physikalisches SystemStellenringBenutzerbeteiligung
15:51
Natürliche SpracheOntologie <Wissensverarbeitung>IdentitätsverwaltungAbstraktionsebeneVolumenSpeicherabzugUnterraumFlächeninhaltVektorraumWeb SiteRepository <Informatik>VerschlingungLastBitmap-GraphikNotepad-ComputerSkriptspracheWeb-SeiteMeta-TagLemma <Logik>Installation <Informatik>ProgrammbibliothekAuflösung <Mathematik>Kontextbezogenes SystemGebäude <Mathematik>ClientInstantiierungLuenberger-BeobachterTexteditorResultanteKontextbezogenes SystemBitComputerspielWort <Informatik>Interface <Schaltung>MathematikFormation <Mathematik>MAPKomponente <Software>KonditionszahlSichtenkonzeptProgrammbibliothekVorzeichen <Mathematik>MereologieZahlenbereichProjektive EbeneMultiplikationsoperatorComputervirusNatürliche SpracheOrtsoperatorArithmetischer AusdruckInteraktives FernsehenAuflösung <Mathematik>Konfiguration <Informatik>Deskriptive StatistikSkriptspracheOffene MengeFolge <Mathematik>Ontologie <Wissensverarbeitung>TermVideokonferenzWurzel <Mathematik>Computeranimation
18:39
Kontextbezogenes SystemStreaming <Kommunikationstechnik>Kontextbezogenes SystemCoxeter-GruppeValiditätKomplex <Algebra>DatenverwaltungLuenberger-BeobachterArithmetischer AusdruckResultanteFormation <Mathematik>Computeranimation
19:31
Attributierte GrammatikLokales MinimumDatentypStatistikInterpolationInstantiierungService providerSchreib-Lese-KopfBitmap-GraphikBitmap-GraphikLastPasswortFlächeninhaltComputeranimation
Transkript: Englisch(automatisch erzeugt)
00:08
Okay, I'm going to talk about the way of modeling in K-Cup. I will just provide information. K-Cup is like a modeling platform solved for scientists, or people who want to just model and observe the world.
00:23
And as such, I'm a really tiny bit, and I will be probably the more genius, the more creative type. Whereas this has, it's just the deep of mind, K-Cup is very well for a huge set that is built about semantics, and I will provide a couple of things in terms of whatever is behind all this.
00:45
So, what is K-Cup anyways? K-Cup stands for knowledge and laboratory. It's actually from a software point of view, a full stack for a integrated model. So it's forwards the production, the duration, the linking,
01:01
and I found it in the environment of psychographics. So the governments have 3D datasets, but that can be also modeled by scientists, because if you have an indie guide that is a dataset, but you can have an indie guide that is a model that can be made from other resources,
01:21
as such, you can have water models, you can have results models. So depending on what the country, the government, has available, K-Cup should be able to put together as a data field, and we'll see that later, and choose the right set. Let's start with what the pieces are.
01:42
The first thing is the K-Engine, and that's the part that does the handling. So it's a server instance that collects everything from K-Cup networks from this server instance, and puts together a data field that collects and, with machine reasoning,
02:03
puts together the proper resources and models to return the observation that was requested. Then you have notes. Notes are notes that actually represent the knowledge. You can put them on data sets, you can put them on models, and they will interact within the network with the engine
02:21
and with the other components, so that when the data field is actually generated, it will get whatever it can from the notes that will be the resources based on a waiting system and machine reasoning. Then there is another, because it is necessary
02:40
that there is some instance that handles the handling of the application, uses the narration certificates and coordinates all of your servers. So, as you can see, this is meant to be a distributed system. You can have various notes. You can even decide to use this,
03:03
untouched, have your own private network, have your own models, your own knowledge, but still untouched to the whole thing of the integrated network model, which I in a second explained. And then we have to interface this. The first one for scientists is the model.
03:21
It is really an integrated development environment, where you can model in a near natural language using concepts. And that is what I will show you today, the force of the presentation, but you can't really do this. And then there is the Explorer,
03:41
which is meant to be for just paint users, where you have a map, and you have a bar, and you can just insert requests to solutions. This is a software stack, but how do we come to a platform? And that is the AORUS project. It stands for artificial intelligence for a while,
04:02
hence its name really. And it is a platform that is made out of software stack, an environment that scientists can use. This means that the AORUS project, AORUS also refers to the chemoscientists, led by Fernando Villa, who was the initial author of all this.
04:22
So this AORUS team and project put together datasets, looked for, for example, planet-based datasets. They put them inside the adapters in the various chambers. They developed models and put them inside. And mainly, they created, from the ontologies,
04:43
the experimental architecture of the world, so that you have a language with which you express your modeling. A tiny bit of history has all started in 2007, so it is quite a bit in its realm. Here is the environment as a scientific project,
05:01
it began it, it attracted various fundings. In 2012, Professor Villa went to EC3, where the head of K.W.E.E.R. is resized as the last set of climate change. And the project itself had a big boom in the last years, as the UN decided to reuse its platform
05:25
for rapid and natural accounting on K.W.E.E.R.s. The statistical division of the UN provides a platform to all countries in the world to rapid and natural accounting.
05:42
And this is based completely on K.W.E.E.R.s. Ok, a quick slide about how you can do it if you want to try K.W.E.R. So we can just get to that link, and we can give you some information about your accreditation system,
06:01
and you are able to create a certificate in which you can access that code. The whole software side is a good resource, there is GitHub, you can find it on the UN's content, and you could install it yourself, but some components and all data you can exploit by accessing
06:23
would be a certificate to do the trick in the K.W.E.R.s. What you can download is a control center, a kind of software like this, from which you can install your own K.W.E.R. engine. So you would go through K.W.E.R.,
06:40
but attached to the whole network, you are able to remodel it. And from there, for example, you can do it with any tool, you just open the box, you see a bar, and from that particular bar, you can just enter what you want to observe, simply as an environmental or political,
07:01
it could be related to anything supported clearly, but you can also dismiss it. So if you start working on innovation, you can close it with the system, you can close it with the possible observation. And, okay, a simple one, you can just type select innovation, it will generate a lot of validation, which is the observation that comes from
07:22
the validation of this query maker. Okay, you can also add semantic operators, that semantic operators are those that transform concepts. So you could say distance to city, which is a subject, and once you've got this,
07:41
it will generate a data flow that observes in every pixel of the map the distance to the new city. And since the KEDA project embraces identity with fair concept, it really wants to reproduce with science,
08:02
and you can, there are a couple of tiny cups here, and if you click on one of the cups in the tree, that was a bit early, I will talk about this in a second, sorry. This is just one example, I will show you some of it. Anyways, when you insert an observation,
08:22
the data flow will put lots of resources, and any observation that is generated on the way will be shown below the main observation. At some point, if you're a scientist, you will get to the limit of what is explored, and there you really would like, probably,
08:42
to use the model. And here is where you see, actually, this language where it's saying, okay, model runoff observing circular inputs, which should be crop-efficient, minimal, maximum temperature, or something like that. So, this is a very complex form,
09:02
we should have it as some similar examples. This example here, if I could use the hard skills, they were like hanging, because this is a GIS example, I mean, you can always force a system, right? And here, actually, it's something where you model real concepts.
09:21
But in GIS, we usually do stuff where we just get the job done, and, for example, I want elevation between half and a thousand, that's not really a concept of the world, right, that you can model. But still, you can say, okay, modeling this concept, which I just gave a name,
09:41
observing geolocation, by giving this geolocation, and it will generate that. This is an experience, for example, and if you think of that now, on top, it will show you the data file, it will show you exactly what happens behind the scenes when you run this. This means, elevation in a certain range,
10:01
okay, you have, if you click on this here, you will see the generation that you entered. If you click on this, it will show you exactly where the data came from, because when you are the source of the node, you have to add more data, because at some point, for example,
10:21
for the U.N., this was extremely important, you are going to generate records automatically from the models, because everything is perfectly career-related and semantically-related, so you are able to generally take goods and use real and digital records. And, that's why if you think of the resource,
10:42
there will be a nice representation of the resource. Let's see another case. Here, we are trying to do something more reasonable, let's say we want to model what we say is a land cover for our purposes, so far from the city, and possibly not in the middle of the country.
11:02
And here, you can see, this model behavior, our drawing, of land cover observing the distance of the city in this, and the land cover time, so you have to define what you want to observe, and then you say, okay, it's unknown where the city is in the middle of the country,
11:23
and where the land cover impugnates another set of actions, and where the land cover is in the middle of the country. And this will result in something like this. If you then look at the data here, it's a bit more complex, but it will show you exactly what it does. For the cities, it creates a mystery map,
11:40
and then you can see the distance for the features, and you can click on every piece, and it will show you what's there. Land cover, for example, is a merged instance between ESA, land cover, and Korea land cover, because they share different parts of the world.
12:00
And extremely important, the fact to have a known scenario that can take the world from you is what makes it possible to part of our state plan. If you make, this would be playing on the land cover source, the land cover source in this case can be ESA land cover or Korea land cover at a different coding of the land cover itself.
12:22
So when we have the resources, we give the scenario adaptation of each code, and they merge to the second meaning, because it's not connected. This is another example, just a quick one, so if you want to model for this elevation,
12:40
so you go to a certain elevation in the land cover type, and you can just say, okay, show just where the land cover, show the elevation just where the land cover is, so on and so on. And the same here, you see where everything comes from. To make it even a bit more complex, this is really more jealous doing,
13:03
so it's more so that the system can also have, where the scope is able to do more conditions than the data field that represents everything. It's just to show you what you can do. Usually this is used more to observe existing concepts, not really to find great stuff
13:21
that usually engineers or technicians need. One thing that I found is really cool, you can also generate, modify, look at things. So you say, model the level of elevation, that sort of elevation, really. When I look up, and you just throw down here a look up here,
13:41
you'll see a lot of people that are here and under, and everything that it is right now, that's the integrated model, is part of the world view, which is all the same that goes down to ontology. And if we generate a map, where the elevation is classified by hydrology,
14:03
in a very, very simple way. K-Lab is also changing time, so we started adding a couple of hydrological models, one of which is in this, if you are into hydrology and modeling, and this means,
14:22
and we, for the hydrological models, we needed writing data, and Copernicus was beautiful for that, so we created an adapter, and this means you can model the Copernicus, the Copernicus resource,
14:40
as a web process. You don't see it very well, but you can define steps in time, and it will be quick over time, and you can model where to raise the volume, or a minimal atmospheric temperature, maximum atmospheric temperature, or temperature itself, which is the average, because you have the same thing.
15:02
This means the world level. What does that mean? That if you do this for a year, and you want the result for every month, you will then be able to browse it. But what happens if I do it? Because I mean, the Copernicus data, or maybe you have local web data,
15:21
and then you can get the data, and you can decide if you want the data out of that quality. And if you do that, you will need to uncreate it. So, again, since it's a molecular data, it knows what to do with it right away. It will accumulate over the month, and the temperature, it says,
15:41
it will upwork, and it will come together. So both in the spatial, in the temporal state, things are harmonized in the world. Last word about the whole view, which was kind of like this. So think of it, I've never been able to do something like this, but you have to define,
16:02
really to the very, very bottom, where you have to define what is a puncture. And this goes down, down to the whole final phonologies, which is why in the syntax, really, you have to follow coherently. You can't say, give me elevation in the roots,
16:23
because it will go down, and ontology, because elevation against ontology will fail. So it will tell you, it will give you a syntax error, and you have to kind of express elevation in the roots. So it's extremely powerful. And what we did recently,
16:41
since we wanted for a project, to have interactive ontology.js way, we made a kind of fighting, which is a major video on Py. And this time, this is a number from a small library, and we created,
17:01
I think some of the QGIS would seem to be the best option, where you can just put some saying into the script, in the middle of the QGIS. So, just to give you a couple of steps, you start with a bit, and then you have a couple of, get in what to do,
17:20
and then you're ready for the fun stuff. So, think of this as that interface, where you have a map, which is the bar, okay? So, say you are a QGIS, you have your script in there, you're open, and you have your map, and it's showing an open script map, of some kind. And then,
17:40
this is a bit of PyGIS printing, that I will show you, in terms of the part of the kingdom library. So, you extract from the current map region, the polygonal extent, you create a region, then create a kingdom by themselves,
18:02
and you can start to generate observables, as a grid with a certain resolution, and a position in time, and so you need to define, since these kind of observations could also take long, because you get the result automatically,
18:20
it's a sequence concept, so you submit it, you have a ticket name that will wait and hope for the result every now and then. But when it comes, this is a script one, you will get the context. Context is what the map was, it was a spatial context, it has also a camera component, but here we see more of what we're looking at now,
18:43
as a spatial context. So that will be, you won't see anything down here, but I hope the presentation is going to get online soon. So, it's what we saw before, and once you have this spatial context, you can start finding observations.
19:02
Let's say, observe a create, like a chart with validation, mentioned at zero, you submit it to the concept, to the context now, and you will get a ticket name, it will do the observation, create a generated engine, which I've already laid out,
19:20
and extrude it, and confirm the results, and then you have a way to export it. You know that validation will be a raster, so you can export to file like this, and you can then just load automatically as a tutorial, as a raster area, into the methods. I wanted to show also,
19:41
some of you in the packages, workshops, so you know, somewhere, but we didn't see the raster part, so, you can, these are steps where you can just get out of the raster area.