Bestand wählen
Merken

Killed By A Thousand Paper Cuts? A Newcomer's Perspective On Possibilities And Gaps In Software Citation Workflows

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Beta
Erkannte Entitäten
Sprachtranskript
of it this now you myself as OK yeah thanks for coming back from the break after 3 talks of a yearlong experts now will have a little newcomers perspective on
citation workflows for software I want
in particular compare and contrast which options exist for software developers to expose the center the citation metadata for the project and to which options exist for the users to import these metadata and to actually applied somebody father had some introductions so as you probably know the TAB is like this Information Center for Science and Technology Inc. mathematics as a software engineering and so on for a few years now we've had a fair bit of research that Amendment expert management expertise so with my team leader and colleague of Tarjan and Craft and I recently joined to establish similar capabilities for software projects so software that library Carpentries workshop will be a part of this so very much looking forward to cooperating with the a sustainable software Institute and best of new also mentioned before yeah you can use these workshops train yourself and your team in best practices so come talk to me if interested and my background is from the natural sciences however I'm not an implementation of a mathematician but it drifted into IT in my PhD mostly because I had to do lots of method development or I choose to do was extremely interesting and dramatization automation and so on so I got to know all of these interesting programming all techniques so the IT and then I worked in industry for a little bits in laboratory information management systems in the pharmaceutical industry and then the ice image before I joined the TAB recently to support so for projects so this means this is newcomers perspective as a work-in-progress here in the introductory phase still of my work so please if if any corrections or suggestions from his come talk to me not assimilated the Developer Options and here I think I want to rank them by convenience because I would argue that convenience is a really really important factor of how people use software which functions they use and so now how the documents of for example and I think a really good to to expose the citation information is to use the adaptation of it because it's a defined snippet this is how I want to have the self was cited can be of course an article or book or not and software item 1 step further also extremely useful as new mention before you can get is in order DOI a back up and a landing page for yourself what source code if you've published it on GitHub this is extremely easy it's just 3 steps basically and then the node or works through an official integration it updates this 1 will automatically whenever you release in use of the worship using the Guidry's texts of feature and since Mrs. balanced off from last year DOI versioning has been implemented there as well so I think this is good news that things are moving forward in this area and then if you already it in a community of practice there's probably some kind of citation MIT and other format available so I know a bit about and packages always have this description file this is where you put in the information and that it can also be automatically parsed out again displayed on a web site converted into bit taste citations snippets and so on also of course a possible but now we're getting into the tasks that are really a bit annoying to do this if you uploaded for example something on the top was an old or something you can of course fill out the addition made at had former last before sorry forms manually so that helps people find the software but it doesn't really scale and lastly I want to mention the court format enters citation 5 format briefly out of this ranking because I don't want to say that they're particularly inconvenient or anything just saying there new in this list and therefore I think have to prove themselves and all the time so let's have a look at some of these details coordinator J. suggestion is an exchange format for software archives mostly so there's also still a discussion on going whether this should be a generated in a top-down approach more from the archive side of things or if it should be bottom-up approach where it's integrated into the bill processes and 1 example for this exists this accord may top are package which again works is packaging universe and the developers can take over the generation of this corporate adjacent fire themselves included in the software source code and early output should for example to this is how it looks it's Jason Linked Data actually so some of starting points for semantics is included and Jason is really something that should not be read by human eyes sort type by human hands this is for machines to communicate with machines the Citation 5 format as different also mentioned in the the discussion of after the last August kind of an attentive to implementing the missing the potential the plantation items and that of fuels and also they're all in communities that don't have their own mid and our standard already it could be a source format which is then possible to automatically convert into other formats like will be touches there's already partisans in Java and really at and a converted to dictate so I think from the start of this rather young project the tool kit is pretty good alright this is how it looks so it's Yemen-based of a friendly to both humans and machines so and from the 1st part of the slides will be of a lover was so sorry that's a bit faster than from the from this 1st part I got the impression that these up-and-coming alternatives unfortunately signal that this kind of low confidence dictation bit plantation is so the up to the job currently or will be in the future which I think is a shame because many many users are used to lot so on everybody in the natural sciences no was that it's probably something they should use even if they don't do it so establishing a different kind of paradigm used it for me that either format I think it is of course a quite a risk however definitely providing good imposes the for petitions on to develop further now let's have a look at the user's site and here I would argue that the really the most convenient way that the users can import the citation metadata is if they have fled the reference stalled many of them have like a browser plug-in with just 1 click they find find an interesting software and from some late landing page where that the top or S W map for Z be man or something else just import the citation MIT other with 1 click so that should the best what we should avoid I think is to have the users not so I have available the citation it out and then in the writing phase go back and I need to look up this metadata again because now I want to cite the software this is what I would think the users want to do and if in the moment where they find useful software they don't have the option to easily import the stuff many of them will not go back and when they rights and follow up with and citation information also OK if a bit less convenient maybe is copy-and-pasting it from either the test fire that they find doors but on the website or any of these newer formats assuming there would be import options into the reference managers another option also out of this ranking because it's quite newest site and it's only a database includes search engine on the Web that also include software and data sets and sets out to provide the citation information they're easy so you can check that out also software heritage as you probably know is suggesting an archiving all the public source code from several sources and this might make it also quite easy because any base the automatically have fled an archive page a landing page for certain software including all the versions all the files although the commit history from a acute project for example
however now I think you will got to see this on the 1 hand the developer options are quite a bit more advanced and for the users there are several inconveniences along the way is anybody from northeastern gently yeah you know where this is right on the other side broke down last week and this yes if you hockey but I want to start today I
want to start here with some some good news because sometimes the the problems that hinder the users from importing the citation media that is really really easy so for example I'm is Terry was a it's a free and open-source reference manager the translators that used to translate and extract the citation information from websites are not publicly available you can look at them you can help improve them and I noticed that in this year and on network which is a comprehensive archiving network the transit just worked on the main site of the central site it turned out it just needed a new and broader wreck X so many people you know about regular expressions so this is gray about 1 1 third of the peoples of some of these problems are really really easy to solve even for people who didn't study this I can't puzzle this together in a few minutes you can probably see this immediately and another example this one's not so nice if you browse the sea around websites the normal way you can read all of these nice and rich description MIT had out with your eyes but there's the model is is the terror translator never triggers because this is hidden behind frame and only if you'd deep linking is used a few deep browse basically into this website structure it suddenly works it is recognized as a softer object and imports perfectly fine discussion of changing this is ongoing
1 more note about Synote all of them I would totally recommend it but again they cannot expose the software item type that they internally have because they use a citation style language which is the broad standard that is used also by many publishers and that expectation that lattice currently doesn't have software items included so again the translator detects that this is a self item type but when he actually imported it turns into report
ice is happening why is this happening kind of even the people want to do this they kind of their their workflow dies at the very basic
so to summarize I think it's not all a strategy of course so that was a joke about them is really really a lot of good groundwork that has been done there's lots of conventions and we can follow I mention some of them there's also more work in the Python identi in communities this is we I think we have really good footing here but 1 question it has anybody tried to add a reference managers for example do you think that it's better in any of the other because basically they also in many cases crowdsourced their translators and on I do not have the worth looking at so if you know anything about this this is no I think dictation Sotero a pretty good options for us to focus on here and to help improve what needs doing so there is without question a lot that needs to be done for example the get tough the nor integration has been done by them with the last signs that but more and more researchers institutions now also provide GitLab community editions sold with their discussions about this integration also ongoing so these participate in these discussions push this this of course you want to have the cold on our own decentralized with the back up maybe also on the top but maybe not exclusively on the service of the iron American company and the court may out was if you can please help test the world there's some available the same goes for the citation 5 format parsers and translators and so on and I think it's and interesting to talk about both of these tools in terms of should they be integrated into the development environments that people use into the bill systems or into the archives as image before should be a top down approach or a bottom-up approach and for the user options I would say there really a lot of hurdles and problems so we need to for example keep the Parzen translators up-to-date for software relevant sites I think the the bit test and the plantation topic that Mrs. bevor suggested before some interesting and we have to helped push this into the official sources so installing some some dictation fires additionally is OK for us and for our students maybe that's not from many more people so it really needs to be as easy as possible because then people will also do it the same goes for all the citations by languages the language the discussion 1st of what type of on going there as well I suggest to participate in the discussion and in both cases if that is done of course the all this diodes from the journals and so on also need to be upgraded to accommodate these new possibilities and you feel for software and so on last summary I think we have a really big problem here was of the funding schemes that we are currently living under promoting of many many projects but which are probably really small and all of them of course try to find the best solution and going for perfection but at the same time there is even 80 per cent solution that works for most people so we have to find solutions for this then I think it's really a good idea to promote the existing tools to help improve these because the core of sustainable software to reuse it's not reinvent the wheel all the time if it doesn't really have to be invented again so there's many instances where that is valid thing to do unimportant but in many cases maybe just liking or discussing of submitting a feature requests or something like this can go a long way already in the contributions to your own projects s new talked about before especially with version control systems like a tablet GitLab and this is really the work for the 21st century you enable other people who you don't even know you don't know what they can do but they might be able to help you even as I mentioned before they don't have to even understand your code to maybe fix a buck or just maybe right a tutorial anything is really helpful the major downtown the documentation and website thing and we've talked before and heard before about re collecting all of this information again to build these micro apps and Archive pages and so on but there's also many options available to put all of this into 1 basket so this for example generators tool to update your website from the GitHub repository directly so you don't even have to have different sources to collect to begin with and to archive because all of this can be pushed into it is ignored all into this versions in all the bundle automatically so this is I think the lesson we can draw from suffer development that more and more tools and more and more usefulness is gravitating into this especially the kids ecosystem and divide is a shared system after all so the scholarly comments as new a buzzword maybe that's been thrown around it's it's very diverse everybody can benefit from it we all should benefit from it but it also means we all should help take care of it for example very
pragmatic approach how about establishing an open-source Friday in your institution then if we hold a minute as of Friday afternoon a half of Friday afternoon where you improve your own projects make them is accessible to new contributors or just contribute back something to projects I useful to you again as I said even if it's just small problems to fix small things to help improve I know that I want to thank you
for your attention I want to thank my colleagues and land bus for reviewing the abstracts and to the developers below the sort of things for accepting it a
special thanks to the self recitation implementation working group members for lots of interesting discussions and advice and said to the TAB my employer that my lab colleagues for their support thank you let X
Sterbeziffer
Perspektive
Perspektive
Subtraktion
Gerichteter Graph
Punkt
Prozess <Physik>
Momentenproblem
Smith-Diagramm
Bilinearform
Zenonische Paradoxien
Gerichteter Graph
Deskriptive Statistik
Knotenmenge
Numerisches Modell
Bereichsschätzung
Exakter Test
Standardabweichung
Perspektive
Äußere Algebra eines Moduls
Biprodukt
Grundraum
Phasenumwandlung
Funktion <Mathematik>
Addition
Zentrische Streckung
Lineares Funktional
Grothendieck-Topologie
Mathematik
sinc-Funktion
Ähnlichkeitsgeometrie
Endlich erzeugte Gruppe
Physikalisches System
Grothendieck-Topologie
Nabel <Mathematik>
Teilbarkeit
Integral
Rechenschieber
Flächeninhalt
Menge
Sortierte Logik
Mathematikerin
Anpassung <Mathematik>
Mereologie
Ablöseblase
Projektive Ebene
Ordnung <Mathematik>
Koordinaten
Objekt <Kategorie>
Beobachtungsstudie
Deskriptive Statistik
Algebraische Struktur
Numerisches Modell
Grothendieck-Topologie
Gruppenoperation
Translation <Mathematik>
Numerisches Modell
Erwartungswert
Numerisches Modell
Verbandstheorie
Translation <Mathematik>
Zenonische Paradoxien
Explosion <Stochastik>
Gesetz <Physik>
Standardabweichung
Faserbündel
Offene Menge
Gerichteter Graph
Grothendieck-Topologie
Desintegration <Mathematik>
Güte der Anpassung
Abstimmung <Frequenz>
t-Test
Endlich erzeugte Gruppe
Physikalisches System
Zenonische Paradoxien
Term
Division
Gerichteter Graph
Integral
Numerisches Modell
Exakter Test
Kommutativgesetz
Translation <Mathematik>
Strategisches Spiel
Projektive Ebene
Faserbündel
Numerisches Modell
Sortierte Logik
Gruppenkeim

Metadaten

Formale Metadaten

Titel Killed By A Thousand Paper Cuts? A Newcomer's Perspective On Possibilities And Gaps In Software Citation Workflows
Serientitel The Leibniz "Mathematical Modeling and Simulation" (MMS) Days 2018
Autor Leinweber, Katrin
Mitwirkende Leibniz-Institut für Oberflächenmodifizierung e.V. (IOP)
Leibniz-Institut für Troposphärenforschung (TROPOS)
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.
DOI 10.5446/35351
Herausgeber Weierstraß-Institut für Angewandte Analysis und Stochastik (WIAS), Technische Informationsbibliothek (TIB)
Erscheinungsjahr 2018
Sprache Englisch
Produktionsort Leipzig

Inhaltliche Metadaten

Fachgebiet Informatik, Mathematik
Abstract Citation workflows offer the opportunity to consider a wide range of aspects between low-level technical details (like metadata structures) up to user experience design (like the number of necessary clicks/commands). This talk summarises from a newcomer's perspective, how citing software equivalently to books, articles, etc. is already possible. To explore this, we'll be venturing through the realms of TeX code, XML configs and clicky-bunti web-apps. However, existing gaps in the software citation workflow are undeniable. We want to highlight some of those gaps specifically, discuss ideas of how to patch them, and derive general tactics and strategies to improve software citation workflows.

Zugehöriges Material

Ähnliche Filme

Loading...
Feedback