View source for Open science, research, access, data, etc.

{{Library related}}
{{stub}}




===Open Science===

<!--
Open science is an idea and movement to make scientific research 
more accessible,
more transparent,
and hopefully its content a little more easily interpreted, to more of society.


It advocates for things like:
* Publishing '''[[open research]]'''
* Publishing '''[[open data]]'''
* Campaigning for '''[[open access]]'''
* Encouraging researchers to do '''open-notebook science'''
* ...and more
-->

===Open Access===
<!--
The idea that distribution of scientific output should be free of charge, and of other barriers.

This often implies an open license, copyright-wise.


This terms is seen a bunch around libraries, in part due to the way they are involved in finding and articles and their content.

https://en.wikipedia.org/wiki/Open_access
-->

<!--
'''For context'''

Historically, reading papers from other researchers required physical media - it needed someone to combine it, print it, move it around, etc. 

It made sense that this was centralized via journals/serials as in physical media. 
Publishing is a job, and if someone can do most of the work for you, that's useful for everyone.



Relevantly, researchers themselves put a bunch of stock into [https://en.wikipedia.org/wiki/Impact_factor Impact Factor], sort of the PageRank of academia, calculating relevance of a journal (and sometimes also of papers and of authors) by the amount they are quoted.


These factors and others create an uneasy balance. In the way implementation of that that we have now, publishers get a strong monopoly.

Consider Nature and Science, imposingly huge.
And staying that way because of [[network effects]] as there's not a scientist that would mind a paper in Nature or Science.

Such large journals are not open access. 
And aren't going to become it. Why would publishers give up a monopoly position?
Universities pay on the rough order of a million USD per year to get the ability to get the full text of a ''decent'' coverage of journals?


Which is bad enough for the 0.2% or so [http://chartsbin.com/view/1124] of the world population or so affiliated with a institution that ''probably'' has some sort of budget for that. 

The other 99.8% of the world, though pays $30 a piece. 
Which most researchers also wouldn't do, for roughly the same reason:
The abstract tells you fairly little, and you're thirty bucks down the line before you even ''know'' whether the article was truly useful to your interests. 


Which in pragmatic terms means science is fairly closed to the public, very expensive for the public to read, let alone verify, and fairly ivory-tower -- and easier to distrust for entirely avoidable reasons.


Scientists don't even want this, but we do play into this if we work the classical way of publishing via journals.


And this is mostly because of convenience.
Without journals, how do you find relevant articles?





Yes, everyone has to make a living, but when publishers sell papers, 
they sell other people's work, 
which has value because of things external to them,

They don't do any of the funding, but do reap the rewards.
They don't write the paper, 
they don't peer review the paper,
That's basically all done on taxpayer dime.

They sell the thing they got for free for you,
''may'' make you sign away your copyright claim to it in the process,
and then sell it for profit. 

They used to be focused a lot of physical copies, which had some cost and ris, but these digital days that's less of a thing.


This is basically taxpayer subsidized private profit.


And the companies made the thing as monopolistic as they could,
which is how each university pays on the (rough) order of a million dollar per year for access to a decent covering of getting actual articles.
and paywalls for everyone else, of course.


There are multiple distinct ways this can sit wrong with the principles of a lot of researchers -- and basically everyone else who thinks about it. 



If knowledge is public only to the rich, it is not public.


Open access can eventually help this, 
basically by taking away these monopolies.
-->


===Open Data===

<!--

In general, open data is data that is openly accessible, and usable for any purpose, and probably licensed under an [[open license]].


'''Open data in research'''

In research/academia, Open Data is a little more specific, 
namely the idea that data collected to support research should be freely available - to verify it, for transpanency, to use in further research beyond the one paper it was collected for (maybe a few more if it was useful data), etc.


Each field has their footnotes.

For example, most research data is expensive to collect. 

Since the practice of publishing papers was classically very slow (and not cheap to publish and distribute when that meant paper and books - which is part of why people were clasically protective of their data, an understandably so).

Even now, with papers being digital, we are only slowly moving away from this view, and with publishing first meaning a ''lot'' in academia, data is only made available after all publications are out, so that other labs can't profit off your 
expensively paid data collection. 



-->
https://en.wikipedia.org/wiki/Open_data

===Open Research===
{{stub}}

Research that is more transparent in more of its parts, often with the point of making reproducability and collaboration more likely.

This often amounts to sharing both the methods and data,
making a point to say more than "we did a thing, we didn't forget to pay attention to X and Y, and here are our conclusions".

https://en.wikipedia.org/wiki/Open_research


====Open-notebook science====

Open-notebook science - basically to make the process and developments on the way (lab notebook and/or personal notebook) public, and not just the end result.


It doesn't have to be ''everything'', nor does it have to be raw form, 
in fact both might be counterproductive because they might not be coherent enough.

The idea is more to 
: help science have public record of data that would otherwise not see the light of day because the result was less interesting to publish.
: help deeper understanding of the research (both laymen and scientific), by seeing intermediate results and development of theories, methods of analysis, etc.

https://en.wikipedia.org/wiki/Open-notebook_science

===Principles and ideas===

====FAIR====
{{stub}}

FAIR (Findable, Accessible, Interoperable and Re-usable) is a (minimal, community-agreed) set of principles and practices that make data not only open in the sense that it's out there, but also make it easy to find, and use.



It could be purely about metadata, or about data as well - this depends a bit on the type of repositories we're talking about, and the kind of organization behind it.
: Note that when it describes data, this overlaps with [[open access]] details.



'''Findable''' - ideally has rich metadata, [[persistent identifiers]], and is searchable by those

'''Accessible''' - fetchable at all, fetchable by identifier, allows auth if necessary (also, removing data doesn't mean removal of metadata)

'''Interoperable''' - uses metadata and knowledge representation that is accessible, broadly applicable and useful for analysis and processing {{verify}}
: somewhat vague. Roughly "is your metadata sensible for the job, and can I read it out at all"
: the last, 'are there qualified references' is vague. Does it mean links to related resources? Browsablility by subject headers? Using codes to disambiguate things confusable in text?

'''Re-usable''' - clearly disambiguated resources?, clear license, showing origin of data{{verify}}
: also "domain relevant community standards", which seems vague.



The basics can be considered in terms of compliance to clear use of PIDs, clear metadata and machine readability (also for indexing), clear licensing. 

Some further requirements are more open-ended, as pointed out in A Dunning (2020) "{{search|Are the FAIR Data Principles Fair?}}" evaluates to what degree repositories currently adhere to this 





See also:
* https://www.fairdatapoint.org/

* https://en.wikipedia.org/wiki/FAIR_data

* https://www.ands.org.au/working-with-data/fairdata

* "{{search|To be findable, accessible, interoperable and reusable: language data and technology infrastructure for supporting the FAIR data approach}} " [http://ceur-ws.org/Vol-1856/p04.pdf]

* https://www.force11.org/fairprinciples

===Relatedly===
<!--
'''Linked data'''

Not so much about the availability of data itself, 
but the ability to find similar or related data based on semantic queries.

inked data isn't necessarily open or free - but regularly is that too.

https://en.wikipedia.org/wiki/Linked_data




'''open knowledge''' - 
: knowledge that is free to use use, reuse, and redistribute 
: in part the umbrella for open data, open content, and open access
: https://en.wikipedia.org/wiki/Open_knowledge


'''open education''' 
: a loose term
:: refers to open educational resources (and so availability to more people)
:: and of data
: https://en.wikipedia.org/wiki/Open_education


'''Free content''', ''free information''' 
: applies to any work of art, functional work, or other creative content
: that has been made into a [https://en.wikipedia.org/wiki/Definition_of_Free_Cultural_Works free cultural work]
: https://en.wikipedia.org/wiki/Free_content


'''open knowledge'''


'''open specifications'''


'''open government'''
-->

===Open practices===

===Organisations an initiatives===
====COAR====
<!--
Confederatoion Of Open Access Repositories (COAR)



Next Generation Repositories (COAR-NGR) is an initiative, and a spec, 
and more broadly, the idea that repositories can be a central means of exchange, and additional services.

For this, repositories need to take a more active role (e.g OAI-PMH is pull-based and metadata-only)



[https://www.coar-repositories.org/files/M15-M6.2-Specification-of-Next-Generation-Repositories-M12-UMINHO.pdf Specifications of Next Generation Repositories]

ResourceSync - syncing metadata as well as content
* Metadata harvesting
* Content  harvesting

Signposting - basically making semantic relations more machine parseable [https://signposting.org/]
* Discovery through Navigation
* Enrichment with identifiers
* Enrichment with Metadata
* Enrichment with Usage Metrics
* Update
* Harvest
* Simple Web Service Concept




User stories from [https://www.coar-repositories.org/files/COAR-Next-Generation-Repositories-February-7-2017.pdf Next Generation Repositories (2017 – draft for public comment)]:
* Discovering metadata that describes a scholarly resource
:: as a person, you want to see the metadata to discern its relevance
:: as a tool, you want a standard way to grab it (often needs respository specific heuristics)

* Discovering the identifier of a scholarly resource
:: the identifier primarily as a means of getting a (persistent) URI for the actual resource, so that we can fetch it
:: we often refer to resources through their landing page
:: we would like both 
::: the persistence of the identifiers (so landing page may be preferable)
::: and to make it easy to fetch the actual resource
:::: e.g. via typed links that make it clear which links on the landing page
::::: are to the actual resource, and/or 
::::: contain identifiers are likely to resolve to a resource


* Discovering usage rights
:: a lot of scientific resources aren't fully open, and comes with licensing conditions
:: for people, e.g. logos can be clear
:: for machines, links to specific licenses can help (e.g. helps perservation miners)

* Recognizing the user
:: respositories recognize their users (meaning own university?)
:: social features like discussions
:: via: permanent IDs (scholars), social networks (other)

* Commenting, annotating, and peer-review
:: 

* Automated recommender systems for repositories
:: 

* Providing a social notification feed
:: 

* Resource syncing and notification
:: 

* Data mining
:: 

* Supporting researchers’ workflows
:: 

* Comparing usage
:: 

* Preservation
:: 




See also:

* https://www.coar-repositories.org/news-updates/what-we-do/next-generation-repositories/

* https://www.coar-repositories.org/files/COAR-Next-Generation-Repositories-February-7-2017.pdf



=====Pubfair=====


: https://www.coar-repositories.org/news-updates/pubfair-version-2-now-available/

=====Notify=====


: https://www.coar-repositories.org/notify/
-->

====IOI====

===DOAJ===

Directory of Open Access Journals

https://doaj.org/about/

===OJS===

PKP's Open Journal Systems is software that basically helps you set up an open access journal of your own.

https://www.tudelft.nl/library/library-voor-onderzoekers/library-voor-onderzoekers/publiceren-verspreiden/je-publicatiestrategie-creeren/creeer-een-open-access-tijdschrift

===Services and implementations===


====CORE====

Aggregator of open access research papers from repositories and journals. 


Aimed at stakeholders like researchers, the public, academic institutions, developers, funders, companies 



* APIs
: https://api.core.ac.uk/docs/v3

* Dataset download
[[Category:Library and research data]]