Paleomaps: SDI for paleoenvironment GIS data

right so is constructed and this is the final presentation for this session and so that will keep the door going haven't solved at president of bacterial maps STI for paleoenvironment GST thank you very much yes I'm from the University of Cologne directly years 50 not even 50 kilometres north of 1 and 2 yeah I would talk about the project developed out of a project the working in 4 yeah but what I do friends some years and this project of working and some users is called the collaborative research center at all 6 this is an interdisciplinary research project that of the disciplines of archeology the geosciences and cultural sciences that investigates and the question of how all of the number of anatomically modern humans Homo sapiens yeah dispersed or migrated the through the boat and in particular from this research scope into Central Europe at the end this project run since 2010 and to note since 2009 and we are funded onto somewhat 17 so that's then 8 years maybe we get an extension for 4 years and 4 more years that will be to to some 21 and if you're

curious about it you can get more information that this year about the whole research project so I'm In this

research project with about 80 scientists working there the responsibility for the data management side and the the

GS sky and also technical guy since 2010 I'm working there and I built this infrastructure called the collaborative research center 6 database therefore young managing the research outputs of this project and had to this with some of student assessment and to the are supplied to the small team but we have a nice left side you can visit today on the URIs can it has certain there are several different data catalogs and to this the idea the interface we build on you know technology burden the for the metadata of non geospatial data we have c can instances running then we have 5 back and lashed other factors that that I'm not going into to relational databases and the infrastructure that serves as a basis on type of free content management system and we develop several extensions for it to handle the different scopes of infrastructure endeavors yeah

all of this we we will be faced often the request for yeah providing GIS data for that geologists and the project to now basically but mapped to presented data in the palaeo environment to the con the context and to so we got a pile of a party environmental GS data we had and 3 just thought why don't we share it with a little wider community and yeah so we came up with this on a subproject of the state management project where we yeah so you might think we have lots of different sources for planning environmental information in the published record but very few faced sources for upon environmental information so all worked closely OK where detail in the next we have the

this Falmagne different sources of but this upon environmental information that would be as organized on along 2 axes the religious of the 1st axis as uh dada representations if it's unstructured structure to information or the spatial information if it's an intrinsic spatial information of the data or it's an annotated and quality and natural systems just came up for this project half a year ago or so to the the works so basically we have like published papers with text has written OK we have a tiered gracious and there were of the riverine to follow 200 thousand years ago when the this in this way just in plain text and we can somehow wallet GS dataset from this information then we have like structured text non years directing not not directed yes status in their like for archaeologists they often have big excel spreadsheets where's the tons of artifacts and did so this kind of information so uh and then we have like maps for for publishes books maps so we scanned them and you reference them and have them in this year and some few fuse other GSP status as a so there are like the song quite good poly environmental data modeling projects going on the using the same kind climate models like the IPCC you useful climate forecasting they applied to these models to part environmental conditions like what you can get from ice cores you can get the concentrations of several greenhouse gasses and yet the the orbital parameters of the percentage of solar radiation and just a physical equations that you can input in these models and then you can come up with quite a good models of the support environmental times and we can use it to so this solid

looks slices of the maps I spoke of Homer samples this like the textual information from some publications we put this different kinds of informations into would talk about this in a few seconds and then

we have this GIS datasets is different the that are available and the other

like relational databases spreadsheets chart of unstructured finds exam LCC so we we have further the collaborative database use internally only for storing this information sources and handling them and so querying them this is so yeah organisms in terms of musician from and from this we create these Jess datasets then publish the website showed before

and this is what tasty I cut it looks like we have did you know it for some of the larger datasets we have still a maps of running because we had to move so far used you know we still have it running for some datasets and you have met proxy for mainly caching external services which we also provides throughout the front end but also to because for example the the base layer because it just way faster that you saw in you at end of yeah so they haven't always within the university of local-area networks so it's not vulnerable to hackers and we either from the left front end which is then the proxy all proxy the that the WMS WFS etc. simple PHP reverse proxy to send the contents just just against attackers and because

of the you don't think it but there are some people who enjoy

too compromise so that a small academic projects for no reason I don't get it but you need this kind of measure from was so this is what

it looks like I can't even tried to show that online learning this slopes and in

so is just an example data sets this is a catalog you can search and then we

have like you know it from each G is that of so this kind of spatial data infrastructure but in an integrated from we we developed the customer type of 3 extension to integrate the geo note that here into all of its design June not has the rest of the API so we talk to those receiving and include here there's claimed you know functionality but in another design to in our front and so

this is what the data catalog looks like with a different sources if several food and yeah can search through it I'm inclined to have to see that

quite like that the search functionality we have like some looks like this it's so that

searches in different spaces and Jameson abbreviation for last glacial maximum and you see directly what kind of people are connected to what maps we have what data sets with people don't go and see some data

this is loading so this is what

the data set would look like there's several different resources on it like another man we can then load so OK here that the use so work yeah this is sort of what the infrastructure looks like and to yeah we

also the to to implement all these of the concepts of open science and open data and to you will chairing also stated that sometimes we we have some of the other new myself don't have these issues but there are some issues with as the copyright for this state underlying from it because when we get it from published sources and reuse of just the you know

we we can do because we just say it's a scientific work we build a new dataset from a tape of his publishers to it like a scientific resources in our database using a DOI Digital Object Identifier private cited like this thing a publication if you use it for example for as a basement for archaeological finds pretty nice can just type us and we get some the credits or impact from it and to solve this problem with this copyright issues so it's so basically in science and so the CC-BY you have to cite your sources of course so that we facilitate this

by having this sort of sanitized of documents around to like such a data product this dataset for the GMM we side to all the sources we use to describe how we built this data said in distance not

yet fully automated because 1 . 1 this we need to right this takes sources have of course in computer can do it or we are not capable to tell the computer to do this and yeah we we just can have this metadata be included in the slightest template and FIL however describing text in addition to it and have published it with the state and the and this

already it maybe also bit fast but yeah they give that that sustains has

lots of times for patterns of some of the here's a question of will work but you have to focus on 1 of his so thank you for the presentation in I'm working kind a similar set up it's in his talk chair science so my question is some not so much ask from the viewpoint of the SCI but more from the user feedback because welcome from you know we start out this project in on people come to us and say that OK have this lots of data and then they come and they give a like a become like a Microsoft Office document so obviously cannot to process that kind of data and also it's well the idea of putting under a permissive license is not that popular too state of or offensive which so my question is how do they become actual scientists react to that because I don't know how the hell you and community as and it would the set and kind of skill set that the 1st part and the other part will further attitudes towards permissive licenses that would be my question the yes yes they have exactly the same problems so we none of scientists really but not more on that very few are thinking about this at all and on yeah we are not really you know the yeah we we we can come mostly up with his own ideas to provide data to them so giving them to use because we get paid for it but then not often come to us for asking something so you need have but we collaborated at the end almost never they want to publish the data just for publishing it and sharing it it's the other way around their ready really frightened of follow their colleagues their congruence to that they ideas and publish it before they do it so what they would not share it before there really exploit all of it and that's never the case so they don't share it if they don't have to the founder of the German Research Foundation and the forces them to publish their research results that come after the project ends for at least 10 years after the project funding ended it must be available so that's why this platform was founded and they have to publish in the documents but they can't this is said this much of the Office Document table in a word document or pdf and say OK this side that my duty is my data and then they have you the option to say I just don't want to have a public in the they're just on request so users of interested people want to have access to data can press about this to request access and the data set all 9 and we get more information that there is no interest and the dataset on can has to argue which is locked in our system so he can answer why do you want it and then you can say 0 no not of interest in something like this but in this is not very often the case so and much of it also goes ground all infrastructure the more often just day-to-day contact them directly by e-mail and they talk about we have like maybe 1 rechristened from external entities tools on datasets so it not so much so we we have also the the visitor statistics are like 30 as it today so it's a really low traffic some said but so but yeah we we we need can this project and we also get some good feedback from this we just 70 % the other half a year this from for providing of pine environmental GS datasets and there are some 1st of science the this might be the increase the use of the infrastructure results of In this to hear discovered thanks thanks for the talk and many Amistad but I also able to query directly on the dotted have rhizomania platform to justify and this precedent as according of to story

as and so we have you we we have make data according to each dataset and for the spatial data sets we query that and you know address up the and points but this is and you know what internet search I think it's a full text search on the method of items from and to on the show data sets side we have a sequence and point and then there's the Elastic and . 4 this 1 the data items and we have some features like for the we have the these so the 7 l and research has a similar projects in which you can then browse for their dataset items have an infrastructure or you can use the map for spatial rolling datasets so so you know you can give given extent or a location for 2 datasets for which you can browse then we have text you can further along I mean you can't query the dust directly like here for instance and file of a query which says give me all records from some plant species in your driest you can use there there let's say you have some time that not only method but also real that's agreeable in your DOS-based continue really query the dance itself and give me all the but upper level you're dryness plants yes when I started this project this was my aim and after some years for those who get have to say that I did not proceeded with his name because it is hard to say just too much for 1 person and 3 student assistance to get this done in any uh making sense where we have this internal MediaWiki with

which facilitates some of better than this infrastructure this is

the box of plant or if you do it so you need a big scissors stuff like hundreds of people to do it could some that's why we didn't implemented not enough resources careful and I have a great time enjoyed it and I love the idea of giving map data I I switch that been there when I was publishing data that which I had was if you go back 1 slide against media wiki page that 1

yes when he feature external literature do you have and extract the locations and suchlike 0 can you do that automatically yet no no it's that goes something like this from someone the seasons following from this publication this is so interesting information about this and that and then we take the bibliographical information and put it into our system and then we had some additional information we can grasp funded or the the the domain scientists and experts tells us us what it's variable from them and we just you take like which times less it's concerns which spatial extent which environmental variables are given no which architect and ontological cultures so are present and stuff like this it's really laborious taxes welcome and that even if you had of students working on it for free which we don't have that will not be enough it's not never enough so that so the cells yeah and thank you once again and I was wondering how you manage to get along with more flexible data models that actually change would have them and within the rest of each researcher because that's what I figured that's the kind of hard to represent especially in a relational database so perhaps you could describe the challenges in that kind of guess of the data modeling is really the top part which is not so seen from the outset there's what there's also this the most work which I have to think about and In this MediaWiki instance I have the data this is more or less my data model which has all the time extended and adapted so we have the flight for to cultural context class it's called and there you have different attributes and I always work on refining them and adapting them actually the solution this software is really flexible this so I just have to the change 1 point and it's adapts to the rest of the system this is in a relational database also this is 3 years like it has about 2 movies and you don't have to deal with the command line the SQL queries and and a to be honest and it's too easy to that destroy your as created over is if you don't back up then with some queries and so I'll stick to this and could not directly on the relational database for the data model and action because interest proved to be yeah more at the receiving end in some way to 2 arrows and it's just you can revert every year and change it's locked and and to yet but this is through the hard part to 2 modeling some my was when I started out to developing ontology and publish it and get its standard and other stuff to solve went down along the way it's too hot for wanted persons there there some standards from ontology community like the sea ducks here and this really uh well known and geospatial standards of course and that would this will be a combination of different vocabularies integrated into 1 domain ontology for this project I have this in some way implemented as a but it's not formalized I could do it maybe for some publication of sometime I try to formalize it so I just finished my phd you and me so I have to publish my dissertation stated that they did not went so far in this sort of pitch to formalize all of it so constant on several specific topics and you can think it's coffee time so thank you again Christians for the presentation