salgo60
commented
1 year ago Why cant they work together and produce ONE knowledgegraph and support citation graphs?
-
Kungliga biblioteket #25 LIBRISXL koppla "samma som" Riksdagens Öppna data / #17 LIBRISXL: citation graphs
-
Riksarkivet SBL #12 SBL leverera data som data - Things not strings
-
#4 Nationell dataverkstad skall jobba med FAIRDATA och persistenta identifierare
-
What I observe is a lack of an ecosystem - #datasilos.
While the project serves as a commendable example of GitHub utilization, it appears to overlook fundamental aspects such as semantic skills. Additionally, there seems to be little collaboration with Riksarkivet, SBL, museums, etc., suggesting they operate within another new data silo.
- Riksarkivet SBL - works 2023 like they did 1918 my question if they 2023 could create some structured data - It feels like achieving a world record in a non-learning organisation.
- Museums - dont have skills to do same as Wikidata, nothing has happened with the 2022 vision they had 2014 in this video - see "Museerna icke lärande organisationer #33"
- RAÄ - dont have a public backlog and create to much linkroot #36 and dont use tombstone pages -> see 2023 oct Maintenance hell
- Riksakivet - is an archive that dont support persistent identifiers #17 --> they havnt created a lifecycle management of information --> The Swedish "Domstolsverket" dont know where to find documents #1 and there is no vision of an echosystem and "Lack of information lifecycle thinking and handling 'persistent' identifiers that 'disappear' #63."
- Riksarkivet SBL how wrong they work with textstrings and not Things "#7 Riksarkivet SBL: Släktartiklar saknar persistenta identifierare för personerna i biografierna"
- Familia - a research project that has as an excuse that they are a research project and dont need to create linked data
- .....
My conclusion
The aforementioned project had machine learning professionals, and their use of GitHub was commendable. However, we require individuals with a digital foresight who can confidently communicate expectations to other organizations.
Those overseeing finances must acquire new competencies and possess a vision for building an ecosystem.
- Riksarkivet #25 Strukturerad kravinsamling
- Riksarkivet SBL #26 Strukturerad kravinsamling
- LIBRISXL #24 Strukturerad kravinsamling
- DIGG #99 Strukturerad kravinsamling
- Nationell dataverkstad #27 Strukturerad kravinsamling
- Riksdagens Öppna data #96 Strukturerad kravinsamling
- RAÄ #29 Strukturerad kravinsamling
See nfdi4culture.de/resources/user-stories
copy of
The big step I see with this project Riksdagens Corpus
What I lack from project Riksdagens corpus - 2023 oct
[ ] You have not explored Wikidata extensively to determine its potential for enhancing research, particularly in the realm of political research on an international scale, such as incorporating relationships between different countries' corpus
[ ] if the Wikidata approach could be used for better scaling political research and do it more international i.e. adding relation between different countries corpus. I compared the productivity with SBL text strings and SKBL structured data, and SKBL using structured data produced 100 times more female bios. My feeling and what I see with Wikipedia/Wikidata and moving to Linked data is the real game changer see Tim Berners Lee The Next web of open linked data and Google "Applied semantics: beyond the catalog". The representative from Google talks about the importance of good metadata, and asserts that if they do their job well, there should be no need for anyone else to look up information such as Angela Merkel's birthdate...
seeing all the cut and pasting and that SKBL conducts research on the same women that SBL has already researched is not 2023 and it cannot be justified that tax money is not used more efficiently, and everyone has to start from scratch sad that SKBL produced so bad data and didnt do Linked data....
I miss any semantic discussion like SKOS how 2 knowledge domains should connect or how to handle differences between sources (#222,#222) and how to describe a source with uncertainty so metadataroundtripping will work
my talk about metadata roundtripping and persistent identifiers / video that we already 1750 used identifiers for rune stones but often miss it in 2023 for projects like Riksdagens corpus. It works for ORCID and DOI but is needed everywhere and it should be available day one in a project like riksdagens corpus so you can track the development of good data and not just wgen publishing data. GIthub is great but we need to track every datapoint also have tombstone pages would be nice... we see this problem everywhere like even when there are available persistent identifiers people have excuses why not use them and owl.sameas. See issue #269 that should have be solved day one in the project... now wikidata cant reference the datapoints or track the differencies between the two domains... I also would like to see usage of SKOS and handle sources that we trust less or more see WD P1480 "sourcing circumstances"
good thoughts from Katherine McDonough is that digital humaniora needs to start work together and create curated data that works together... right now I feel Wikidata is an enabler but that is not serious....
[X] #269 Utilizing Persistent Identifiers (PIDs) is a practice aimed at ensuring the long-term accessibility and traceability of digital items. Wikidata, for instance, has employed PIDs to all uploaded images, enabling a more organized and searchable database and support of more than 300 languages. Moreover, they've introduced a feature allowing for parts of an image to be annotated to indicate what or who is depicted, enhancing the information retrieval process. An example of this is seen in the annotations of "The Coronation of Napoleon" image, where labels are provided in various languages including Chinese (zh), Swedish (sv), and English (en). By integrating such practices, you can significantly improve the management and sharing of digital resources within the European research data framework. This move towards a more structured and interlinked data environment can facilitate collaborative research efforts, and potentially unlock new insights through cross-referencing and analysis of a rich, multilingual data repository.
with the network iNaturalist we have
same as on
iNaturalist taxon-ID (P3151) = 831 000 taxonomi id:s same as e.g. Q25307 = 891696 Eurasioan magpie
iNaturalist place ID (P7471) = 52 200 places see map
[ ] It appears that there hasn't been a clear initiative to challenge organizations like Riksdagens Öppna data, Riksarkivet, Riksarkivet SBL, Kungliga biblioteket, and Digital museum regarding the quality of data they provide. Understanding the level of support or the lack thereof from these organizations is crucial as it may significantly impact research outcomes. The gaps in support might relate to various factors including easy to communicate like using GITHUB, data accuracy, completeness, accessibility, or interoperability which could hinder the progress and quality of research. By addressing these issues and advocating for better data practices, it could pave the way for more reliable and comprehensive research, fostering a conducive environment for scholarly endeavors. Furthermore, collaborating with these organizations to improve data quality and availability could potentially lead to more insightful findings and a richer knowledge base, thus advancing the broader research objectives.