NARA is in the very early phases of designing its next generation online catalog. Should we allow the public to tag catalog descriptions? Why or why not?
Our current online catalog, the Archival Research Catalog (ARC), does not allow for tagging. Staff archivists can add subject terms from our authority list of approved subject headings (based on the Library of Congress Subject Authority Headings), but in many cases there is not enough time to index the descriptions. People using the catalog would be able to add whatever tags or keywords to the catalog descriptions.
What do you think?
[Revised: August 14, 2009]
In response to your great suggestions about adding background info on tagging, here are a few resources we thought might be of interest:
A general definition from Wikipedia:
A mainly positive review from Newsweek:
- Levy, Steven. “In the New Game of Tag, All of Us Are It.” Newsweek, 18 April 2005/ http://www.newsweek.com/id/49636.
A more conservative assessment in the journal D-Lib Magazine:
- Guy, Marieke, and Emma Tonkin. “Folksonomies: Tidying up Tags?” D-Lib Magazine 12, no. 1 (2006) http://www.dlib.org/dlib/january06/guy/01guy.html.
As a researcher both at the NARA I in Washington, and a user of ARC, I definitely believe that allowing the use of tags would help to make this system much easier to use. The current ARC system can be a bit confusing to both new and experienced researchers, but this would be just one step that the NARA staff could take to make the search process a little more user-friendly. I also reviewed the ARC site in my column last week: http://bit.ly/lYok2
As someone who treasures libraries and has been associated with computer hardware and software since the late 1960’s, please consider that while legitimate users will add valuable tags, there will always be an element of the public who will use that opportunity to cause problems. Think of it as electronic roadside litter. Strongly suggest you speak with the folks at Google, Wikipedia, Digg and Delicious about how they minimize and manage improper labeling. That’s not my area of expertise, but there are folks who’ve had to deal with similar problems for years now. I’m in favor of anything that will help people find the information they seek, with the caveat that you must plan to deal with the attendant problems. Best wishes.
Although tagging is a very helpful tool, when opening anything to the public it can be abused too easily. As this is an official website and deserves respect, any tagging should be done by staff and not the public.
Tags should be encouraged, if they are also verified.
You should absolutely allow public tagging! Crowdsourcing is a powerful tool, but, as eunonymous said, you all have to be vigilant, esp. at the beginning. If it’s a free-for-all that no one monitors (see: comments section of any newspaper story), you’ll be swarmed with the vermin of the Internet. Help guide it to where you want to be, and it will be something useful.
Punny blog name is awesome, btw.
I’m not convinced that tagging is quite as useful as it’s cracked up to be. I’ve just never found it to be very accurate or helpful.
I agree with some of the other comments made, in that an archive of government records needs to maintain a certain degree of reliability, professional consistency, moderation etc. There is a bit of a challenge to do this and open up description etc to the general public. Not impossible by any means, but it needs to be well-considered.
I actually really like what The National Archives in the UK have done with ‘Your Archives.’
http://yourarchives.nationalarchives.gov.uk
I think that’s the better way to do this sort of thing.
Cheers,
Craig Tibbitts
Research Centre
Australian War Memorial
As an archivist, yes I do believe we should allow users to tag online collections and photographs. Everything in the archival literature has indicated that this has been a positive experience for both the archives and the users.
Great to see that NARA has a blog, good move!
I’m guessing that your aim is to find a broad readership among a wide range of stakeholders. Some of your potential readers may be younger, some older. Some may have a lot of archival and technical expertise, others less so. I’ve been following ongoing dialogue about tagging and crowd sourcing for some time. Given the potential broad readership of your blog, I wonder, however, whether it would be useful to link in your essay to a basic explanation of what tagging is. Something that takes into account casual readers who are older, who are interested in NARA, but who don’t follow web 2.0 issues closely. Can you find a good objective summary, one that is sympathetic to multiple viewpoints and questions but geared in a non-hostile, non-divisive way towards identifying and encouraging best practices? If so, it would be great if you could link to it as context for your question. Something low key, relaxed, and worded in language that clarifies the basic goals, objectives, and methodological issues. Your blog has that good tone itself (you raised the tagging question in a nice, unbiased way) but I think context for older, less technically inclined readers might be helpful. Not everyone follows these issues as closely as some of us do.
Thanks for considering my suggestion, I’ll be following the blog with interest.
Hi “Friend of NARA,”
Thanks for writing – adding background information on what tags are and how they’re being used is a great suggestion. We’ve added a few links in the original question that will hopefully help people understand the context a little better. Because pure neutrality is a bit difficult to find, we’ve decided to provide commentary from three different viewpoints: one mostly positive article from Newsweek, one slightly conservative (and fairly in-depth) paper published in D-Lib [Digital Libraries] Magazine, and a basic definition from Wikipedia. Let us know what you think- and if you know of any other helpful articles on the tagging/folksonomy question, feel free to post them here!
– Kristen (Admin)
I say yes, being of the belief that controlled description and open tagging can peacefully coexist in a catalog. As you and others have mentioned, tagging enhances access and can offload some descriptive work onto users. I think there is little reason for fear of public taggers in this case since a research catalog environment yields a completely different set of users from that of a Wikipedia or any online newspaper. However, you might consider whether or not your users will actually make use of tagging functionality in any volume that will make the tags useful. If this seems unlikely, for whatever reason, that should be weighed against the cost of implementing tagging functionality.
Yes, I would support tagging. It’s an easy way to hopefully increase accessibility to desired records.
The UK Archives wiki referred to above is very nice, but I don’t see many people adding to catalog entries that way. It’s too time consuming.
Perhaps a few words of clarification on what tagging is and does (or hopes to accomplish) and how easy it will be would help.
Thanks! NARA is becoming ever more responsive to their public.
I think tagging is a good idea on two conditions: 1) it is seen as an enhancement of rather than a replacement for other types of “authoritative” description AND 2) the tagging data is used to eventually rolled into an algorithm that makes the catalog “smarter.”
Do those things, and I’m all for it.
My use of tagging has been mainly limited to blog entries and my delicious account. I’ve found that I modified my tags significantly with time, and that many earlier versions of tags I used were completely useless as I added more information. In other words, I ended up creating my own sort of authority file with my own sort of subject headings. That sort of refined tagging wouldn’t occur with public access. On the other hand, I like the idea of adding more access points to records. I worked briefly as a volunteer for an archivist, and I thought it was *insane* how many terms were applied to each item versus the number added to the record for a book. I suppose this means that I see more value in authority control than in adding limitless access points through tagging.
Public tagging can be useful, but as others have pointed out, could be abused by ill-intentioned people. I agree with Joshua Barton, however, that there probably are not too many of such ill-intentioned people hanging around the NARA website! With common-sense rules, public tagging could be quite useful.
You couldn’t resist the pun? Kristen, y’all are my kind of people!
Let the users TAG — and let the users REMOVE malicious TAGs.
I agree with Jason: I support tagging as long as “it is seen as an enhancement of rather than a replacement for other types of “authoritative” description”.
Tags are okay so long as someone monitors them occasionally. But please do not go to annotations as per the footnote website — yellow boxes by family genealogists are very distracting!
You need to change your link to Wikipedia from “http://en.wikipedia.org/wiki/Tag_(metadata” to “http://en.wikipedia.org/wiki/Tag_(metadata)”. Add the final parenthesis.
Peter,
Thanks for catching that. We fixed the link!
– Jill (Admin)
The Library of Congress has been putting its photo archives on Flickr for a while now, and letting registered users tag or add comments to them. From the viewpoint of a casual user, it looks like that has been a phenomenal success for them, as they have harnessed the power of all those eyeballs to help catalog its photo collection, which was somewhat un-indexed before. You should talk to the people who work there and see what sorts of problems, if any, they have run into with that sort of openness.
One thought: Flickr requires a user to have an account with them and to log in with that account before adding tags. Perhaps that lack of anonymity prevents a lot of the drive-by vandalism one might fear could occur with such an open system.
Also, no offense intended, but NARA may be so semi-obscure to the general public that it may not be seen an obvious or high-value target for mischief-makers.
I agree with Gene in that you should allow the public to add and remove tags. If need be staff should remove bad tags and also watch the comments for Spam.
Thanks for posting the links, good, quick response. I can’t think of a good “purely neutral” source on the tagging issue. Most writers whose pieces I’ve seen approach the issue as advocates more so than opinion brokers. What you linked to is fine and I appreciate your openness to the suggestion and willingness to revise the blog posting.
I’m a librarian of the old school, so take that into consideration as I comment. I’m not sure tagging by the public will be useful; I pine for the “good old days” of thesauri and “controlled language,” where input was based on knowledgeable experience. The public is always welcome to use Delicious to create their individual tagging system for personal sites of interest. As in all things, the pendulum today is swinging in the direction of social networking, of which this is another manifestation. I tend to prefer a pendulum a bit more centered.
Brooke,
We love the Library of Congress Flickr collection too! You’re right, the response has been really impressive. We spent some time talking to the people behind that project before recently launching our own photostream, http://www.flickr.com/photos/usnationalarchives/, which features selected documents from the NARA holdings. It was certainly helpful to get first-hand information on what works well and what doesn’t in terms of inviting users to add their own supplementary insights and historical knowledge to the existing (and frequently sparse) official catalog records.
And thanks to everyone for your participation so far- we’re excited to see such a great discussion building around our first collaborative description and access question!
-Kristen
From my understanding of the concept of “Tags”, to allow anyone to Tag an article or data base means that the search engines can find them again using the “Tag” word.. So if I Tag something with “Stupid” (because that’s how I felt about it), then anyone searching for “Stupid” will pull up the article. In a perfect world, that would be great. But the world is not perfect, we don’t all think alike, use the same logic, and we don’t all have honest motives.
If my understanding is correct, then I would not like to see the Tagging” of official records allowed.
Thank you for allowing me to voice my opinion.
Jerry, your concern can be lessened by eliminating certain words from the tag lexicon. This is done all the time, particularly with words more offensive than “stupid,” lol.
Say Yes to tagging. There is a risk that vandalism occurs – Wikipedia shows how one can fight it. There was and is a risk that very bad catalog entries by incompetent archivists were made – so what?
Tagging has the potential to be a useful tool if NARA creates procedures in line with their desired goals. If one reviews some of the social tagging that has been added to the Flikr LOC photostream one can easily discern questionable or even erroneous tags.
Is this something NARA is comfortable with?
Does social tagging impact the archival concepts of reliability, authenticity, accuracy, completeness, integrity, consistency, and dependability? I do not know the answer, I merely feel it is a question worth considering when implementing a tagging system.
Here is one image that has some of the issues to consider:
http://www.flickr.com/photos/library_of_congress/2179041640/
Notice the comments regarding the captioning – technically not official tags, but the Flickr search mechanism does incorporate those captions. Also note the link to a “mini-essay” which seems to me to be one of the desired goals of posting digital assets to the web.
I say yes, but I also urge NARA to consider other forms of social computing (e.g., collaborative filtering, commenting — provides more description that tagging, etc. NARA necessarily has to describe record groups at such a high level that user contributions could really open up records.
There seem to be three solutions to ‘erroneous tag syndrome’, and you could probably employ all three. The first: as suggested above, allow for a wiki-like mode of user correction or flagging for seemingly erroneous tags. The second: allow users to exclude the user-made tags from their searches, if the purity of their searches would be tainted by the murk of ‘unofficial’ methods. The third: this help is trickier, but see if you can’t hide the tagging data for an item from search engines and thereby sidestep ‘google mistakes’ from bad tags. I hesitate to suggest a fourth solution–to only allow users to tag with the approved subject headings–simply because outsider headings might prove to be useful companions to the standard.
As stated in the question, there is not enough time for staff to index the description. The only way we will ever get a proper understanding of what is in the records will be to allow the people using the records to add tags. You would need a system to allow people to report problem tags and a registration system that verified an e-mail address as valid before they could post a tag. You might even be able to allow some trusted public volunteers to review and edit / delete problem tags instead if tying up staff hours. It sounds like a hassle but bottom line results should make it well worth the effort.
Thanks for your work to make records as accessible as possible.
Can you use both systems–public tagging alongside staff assignment of authoritative subject descriptions? And then can researchers choose to exclude one or the other during searches, in line with their personal research purposes?
I agree that a registration system would deter most inappropriate tags. For those who are not deterred, would a registration system allow NARA to identify and block repeat erroneous tagging offenders?
Overall, I think tagging is the way to go.
Christine,
We definitely could have it both ways and support both public tagging and topical subject indexing by staff. I believe it depends on the search engine/crawler/software, but most of them would probably allow you to exclude public tags from your search.
At this point, a registration system is on our wish list, but we don’t know for sure if it can be incorporated into the new catalog.
Thanks for your feedback!
I support user tags and believe that they can be leveraged to build out a more appropriate taxonomy for a current or eventual navigation structure. Beginning with LC subject headings is a start but they lag in currency and relevancy; they needed to be enhanced and cross-references are needed that reflect the ways in which the searching public looks for content.
Tagging by the public needs complementary activities by NARA: 1. harvest the tags 2. place them in a framework for a navigation scheme 3. also look at search logs and discover the terms being used to find content 4. build a taxonomy (with cross-references) that will be used to automatically categorize your content 5. ensure that the taxonomy is accessible to searchers who want to “navigate” the site.
In the short term, just tagging can be helpful but social tagging alone does not scale very well.
I agree with those who support allowing tagging. The more access points you can offer a researcher, the better.
The new catalog for our local library system (www.cabq.gov/library) allows tagging, but you have to sign in with your account number in order to add a tag. Perhaps if NARA is concerned about bad tags, it could require taggers to have a “customized user account” as mentioned in the most recent NARAtion post.
Can you set up a tagging system that is separate from the NARA archival staff’s tags – this will allow more wide usage – esp. among teachers and tech coordinators since the latter have come to expect it to organize everything they do. This option would also allow those not interested in that feature to ignore it.
Or, as the suggestions above say, ask users to verify identities.
Jerry, your concern can be lessened by eliminating certain words from the tag lexicon. This is done all the time, particularly with words more offensive than “stupid,” lol. It is something I have tested on my personal website and it works well.
Yes allow to public to tag in catalog. They give their best view on regards to that current topics.