How to add new subject list in citation metadata.

瀏覽次數:130 次
跳到第一則未讀訊息

Shiva Shankar

未讀,
2017年1月29日 下午1:08:552017/1/29
收件者:Dataverse Users Community
The subject list in citation metadata is important,but the subjects like Biology or like civil engineering is not there. How to add any new subject there.




thanx.

Philip Durbin

未讀,
2017年1月30日 清晨7:50:552017/1/30
收件者:dataverse...@googlegroups.com
Hi Shiva,

There was just a meeting last week where we talked about a goal of increasing granularity[1] of subjects for datasets.

This related issue is currently slated for Dataverse 4.7: https://github.com/IQSS/dataverse/issues/434

There's also a Trello card about the subject metadata field and other features related to the home page: https://trello.com/c/x10i7aaB/2-harvard-dataverse-hmpg-4-7

We are very open to feedback on how these improvements are being designed!

Thanks,

Phil


--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/151990f9-dfb4-4bd3-91f6-a84dd50b5715%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.



--

Sherry Lake

未讀,
2017年1月30日 上午10:27:162017/1/30
收件者:Dataverse Users Community
Hi Shiva,

I've changed the subject list for our Dataverse at UVa (Libra Data: https://dataverse.lib.virginia.edu/). You can see our list in "Advanced Search" https://dataverse.lib.virginia.edu/dataverse/uva/search;jsessionid=ff66f773d89fd1eb6a905aaa8f94

It wasn't trivial since the original list is in the database table and the metadata info is imported via "tsv" files.

I can walk you through if you want to change subjects now, or wait until the changes Phil lists in his message.

Oh, Phil - is being able to (easily) edit metadata fields in any of the future releases???

--

Sherry Lake | Scholarly Repository Librarian | University of Virginia Library | shL...@virginia.edu | 434.924.6730 | @shLakeUVA | Alderman Library, 160 N. McCormick Road, Charlottesville, VA 22903 | Alderman 563 | LinkedIn Profile | “Keeper of the Dataverse" 

julian...@g.harvard.edu

未讀,
2017年1月30日 上午10:59:312017/1/30
收件者:Dataverse Users Community
Hi Shiva,

We're also discussing this in our community call tomorrow: http://dataverse.org/community-calls. You're very welcome to call in.



On Sunday, January 29, 2017 at 1:08:55 PM UTC-5, Shiva Shankar wrote:

Philip Durbin

未讀,
2017年1月30日 上午11:05:432017/1/30
收件者:dataverse...@googlegroups.com
Hi Sherry,

There's a goal called "expand data and metadata features for existing and new disciplines" at http://dataverse.org/goals-roadmap-and-releases but I don't think the changes you're making to the "citation" metadata block are necessarily encouraged. The Dataverse team has an assumption that the metadata blocks that are shipped with Dataverse are the same across installations. As Julian just mentioned on this thread, "Expanding Metadata Subjects" is on the agenda for tomorrow's community call ( http://dataverse.org/community-calls ) so that might be a good time to discuss the subject field, at least.

I hope this helps! :)

Phil

p.s. See my other reply for some links that capture some current thinking on the subject field: https://groups.google.com/d/msg/dataverse-community/TNRb7hoJd8Q/JJHIYIMhAQAJ

 

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse-community@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Shiva Shankar

未讀,
2017年1月30日 下午3:15:262017/1/30
收件者:Dataverse Users Community、philip...@harvard.edu
Eagerly waiting for outcomes of tomorrow's community meeting.

Thank you.Philip Durbin,Sherry Lake , Julian. :)


On Monday, January 30, 2017 at 9:35:43 PM UTC+5:30, Philip Durbin wrote:
Hi Sherry,

There's a goal called "expand data and metadata features for existing and new disciplines" at http://dataverse.org/goals-roadmap-and-releases but I don't think the changes you're making to the "citation" metadata block are necessarily encouraged. The Dataverse team has an assumption that the metadata blocks that are shipped with Dataverse are the same across installations. As Julian just mentioned on this thread, "Expanding Metadata Subjects" is on the agenda for tomorrow's community call ( http://dataverse.org/community-calls ) so that might be a good time to discuss the subject field, at least.

I hope this helps! :)

Phil

p.s. See my other reply for some links that capture some current thinking on the subject field: https://groups.google.com/d/msg/dataverse-community/TNRb7hoJd8Q/JJHIYIMhAQAJ

 
On Mon, Jan 30, 2017 at 10:27 AM, Sherry Lake <shla...@gmail.com> wrote:
Hi Shiva,

I've changed the subject list for our Dataverse at UVa (Libra Data: https://dataverse.lib.virginia.edu/). You can see our list in "Advanced Search" https://dataverse.lib.virginia.edu/dataverse/uva/search;jsessionid=ff66f773d89fd1eb6a905aaa8f94

It wasn't trivial since the original list is in the database table and the metadata info is imported via "tsv" files.

I can walk you through if you want to change subjects now, or wait until the changes Phil lists in his message.

Oh, Phil - is being able to (easily) edit metadata fields in any of the future releases???

--

Sherry Lake | Scholarly Repository Librarian | University of Virginia Library | shL...@virginia.edu | 434.924.6730 | @shLakeUVA | Alderman Library, 160 N. McCormick Road, Charlottesville, VA 22903 | Alderman 563 | LinkedIn Profile | “Keeper of the Dataverse" 



On Sunday, January 29, 2017 at 1:08:55 PM UTC-5, Shiva Shankar wrote:
The subject list in citation metadata is important,but the subjects like Biology or like civil engineering is not there. How to add any new subject there.




thanx.

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.

Sherry Lake

未讀,
2017年1月30日 下午4:13:352017/1/30
收件者:dataverse...@googlegroups.com
I'll miss the call tomorrow. Take good notes.

--
Sherry

julian...@g.harvard.edu

未讀,
2017年2月6日 上午9:31:552017/2/6
收件者:Dataverse Users Community、philip...@harvard.edu
Hi Shiva and all,

This is a follow up to your question about adding new terms or lists to the subject field. We mentioned during last week's community call that we're evaluating a large vocabulary called FAST. It's important that we get the community's thoughts on using it for the subject field. I included some info. about the vocabulary in the meeting notes. We'd also like to learn what vocabularies other Dataverse installations are considering and why.

Supporting easily adding custom metadata is something we discuss often, but we also want to make sure it's done in a way that improves harvesting and discoverability in general. This is part of what makes FAST, which has upward compatibility with LCSH terms, attractive.

Looking forward to hearing your thoughts.

Julian

Steven McEachern

未讀,
2017年2月7日 凌晨4:52:192017/2/7
收件者:Dataverse Users Community
One question I'd have is the extent to which the FAST subjects are suitable for use outside the US? (Being based on the Library of Congress headings).

Regards
Steve
Australian Data Archive

julian...@g.harvard.edu

未讀,
2017年2月7日 上午11:31:002017/2/7
收件者:Dataverse Users Community
Hi Steve. That's a great question, and I think another reason to make it easier for installations to choose their own vocabularies. In principle, because FAST is based on LCSH, we could exploit the many links that seem to exist between LCSH and different vocabularies in other areas of the world (some already using the same ontology FAST uses), which would help with interoperability while giving installations the option to improve discoverability by using a more appropriate representation of knowledge.

Measuring the extent to which the FAST subjects are suitable for use outside the US sounds like a great research project, although the vocabulary is developed more for describing bibliographic resources.

Steve, do you have an idea of what vocabularies you would consider?
訊息已遭刪除

Steven McEachern

未讀,
2017年2月8日 清晨6:45:392017/2/8
收件者:Dataverse Users Community
Hi Julian

My last look into this several years back wasn't very successful (LCSH and FAST came up then too).

One vocabulary in development we could consider might be ELSST:
https://elsst.ukdataservice.ac.uk

Comes from the SERISS project in Europe.

Regards
Steve

Gautier, Julian

未讀,
2017年2月8日 中午12:24:242017/2/8
收件者:dataverse...@googlegroups.com
Thanks Steven! Looks like ELSST is multilingual and uses SKOS, which might help with mapping.

When you say your last look, you meant looking into larger controlled vocabs for Dataverse, right? Was your thinking then that LCSH and FAST wouldn't be suitable for use outside of the US?

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--
Julian Gautier
Product Research Specialist, IQSS

Steven McEachern

未讀,
2017年2月9日 凌晨1:14:312017/2/9
收件者:Dataverse Users Community
Hi Julian,

I was looking at LCSH/FAST specifically for use here at ADA, so I found it wasn't immediately useful for the Australian context.
That said, I didn't find an Australian vocabulary either (the APAIS thesaurus run by the National Library of Australia, which we had used previously, was discontinued in 2009).

It's not to say that FAST couldn't be adapted - just that I didn't have the capacity to do that adaptation at the time.

Cheers,
Steve


On Thursday, February 9, 2017 at 4:24:24 AM UTC+11, Gautier, Julian wrote:
Thanks Steven! Looks like ELSST is multilingual and uses SKOS, which might help with mapping.

When you say your last look, you meant looking into larger controlled vocabs for Dataverse, right? Was your thinking then that LCSH and FAST wouldn't be suitable for use outside of the US?
On Wed, Feb 8, 2017 at 6:45 AM, Steven McEachern <stev...@gmail.com> wrote:
Hi Julian

My last look into this several years back wasn't very successful (LCSH and FAST came up then too).

One vocabulary in development we could consider might be ELSST:
https://elsst.ukdataservice.ac.uk

Comes from the SERISS project in Europe.

Regards
Steve

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-community+unsub...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.
回覆所有人
回覆作者
轉寄
0 則新訊息