Joho the Blog » Harvard Forum on Social Tagging

Harvard Forum on Social Tagging

I’m at [well, I was yesterday when I wrote this] a session at Harvard’s Lamoint Library (one of the 90+ libraries here) about Web 2.0 and social tagging. I just gave a 20 minute opener on why tagging matters.


Michael Hemment, the host, begins by showing tag clouds from 50 students who were asked to tag some particular resources. The group quickly guesses that the first tag cloud refers to the libraries, the next is Google, and the next is Jon Stewart. Very amusing,

Michael talks about why slocial tagging matters to libraries. He mentions some initiatives, including PennTags , Stanford IC, and the Steve Museum. Harvard has the CRT (Collaborative Research Tool) and EdTags initiatives. He also mentions iCommons (exploring iSites metadata and tagging) and ARTStor .

He takes a closer look at LibraryThing.com, showing how easy it is to enter titles, organize them, tag them, and get suggestions.

PennTags was created by the U Penn library to enable university members to tag books. (The site is open to anyone, but only U Penn members can add tags.) It begins with a tag cloud of tags used at least 58 times, Users can also create folders to organize bookmarks into projects. [I blogged about it here.]

The Stanford Library Information Center combines tags, blogs and wikis. It includes tagging by librarians who organize resources in a somewhat more orderly way.

Harvard could, Michael says, enable tagging of the libraries’ resources, and the Lib-X tool (a browser add-in that gives you access to Harvard’s onloine resources) could be used to tag sites, adding to what Harvard knows.


Carla Lillvik, Research and Distance Services Librarian at the Gutman Library of the Harvard Graduate School of Education, looks at “social tagging and bibliographic management.” She says you want not only to find resources and organize them, but also to cite them.

She uses as her example the site Five Weeks to a Social Library. She adds it to her page at Connotea and tags it. She could also post it to EdTags.org. But what about resources she finds in research databases, e.g. EBSCO Host? She could add it to Connotea, even though the URL doesn’t look persistent. But Connotea doesn’t pick up any of the bibliographic info from the database. (Connotea has agreements with a long list of such systems, including BioMed Central, PLoS, Nature.com, and arXiv.org, but not with all of them.) She can instead make a folder in EBSCO, which does indeed pick up all the info. [Sounds like we need a standard API for university e-research systems.] Harvard’s RefWorks has the advantage, Carla says, of enabling batch tagging [LibraryThing does too] and enables output in a variety of bibliographic styles [yay!] RefWorks folders can be shared, even with people who don’t have an account; they can be shared as an RSS feed, too. (RefWorks works with Google Scholar — you can set a preference so that results can be imported into RefWorks.)


Michael Hemment presents Prof. Dan Smail’s Collaborative Research Tool (CRT), a social tagging tool that works within Harvard’s e-environment. In Smail’s course on Medieval Europe (History 1122) , students are put onto teams (e.g., “France, Germany and italy”) and are assigned sources. They create virtual note cards that are tagged, annotated and entered nto a database. Class discussions, lectures, and final papers are based on these cards.

The cards tend to include the passage, comment, related links, and tags. It’s easy to navigate by tag.

Pedagogical implications, according to Michael: Students have to reflect on their tagging schemes. [meta learning] They cards “form the basis of complex intertextual discourse on a broad range of medieval topics.” E.g., you could see how Ulysses appears through multiple literatures. Also, tagging develops a personal relationship to the source material.

[Excellent. But we still need a way to write a document based on cards, so that adding info from the card automatically creates the right footnote and bibliographic entry in the document, and notes where the card has been used. I blogged about this here.]


Adam Seldow, a grad student at the Harvard School of Education, works on EDTags.org. It’s a social network to connect people who share interests in education. It’s open to anyone. You can tag a site, vote on bookmarks, email them, blog them, or find related blog postings. You can upload your papers, photos, presentations, etc.


Q: How does tagging fit with scholarly resource? Is there a way to cite where and how a resource is tagged?
A: (Michael) Not in the major tagging sites, e.g., del.icio.us. The lack of rules has been one of the advantages of these sites.The noise introduced can often be negated at least in part by the good rising to the top.

Q: How about privacy?
A: (Adam) EdTags lets you set the level of privacy. And it’s an actively managed site.

Q: What types of resources does EdTags tag?
A: (Adam) Mainly “gray literature” — blog posts, preprints, Web sites, course-generated papers.

Q: (me) What do we do about the fragmentation of the tagging space? I can tag in del.icio.us, Connotea, EdTags…
A: (Adam) A condition when we built EdTags was that it has to be able to talk wth del.icio.us or export to an XML file. Personally, I use different tagging sites for different types of research.

Q: What are the patterns of use at EdTags?
A: EdTags has been live for a little over a year. (It started as TeacherShare.) First year doctoral students, who were trained on it, use it. It’s being used in some specific courses and teacher education programs, plus a community of faculty members interested in emerging trends in education technology. The person who uploads the most bookmarks is a woman from Slovenia. There are about 400 users. About 100,000 hits/month.

Q: Did you build it from scratch?
A: It’s a mashup of Scuttle, an open source platform, with lots of custom work.

Q: HW and SW behind it? How did you finance it?
A: (Adam) A Harvard Provost Innovation Grant financed it.

Q: How to encourage the use of social tagging at a library?
A: (Michael) I don’t know that we want to encourage it. We’re exploring. [Tags: ]

8 Responses to “Harvard Forum on Social Tagging”

  1. Five Weeks is a tough case for citation, because it’s clearly a composite work. You might want to cite it — or you might want to cite one or several of the blogs or blog entries, individual course pages, wiki pages, screencasts, webcasts, and podcasts that together form the whole.

    I’d use the course intro page as the URL for the whole, but opinions could differ — I suspect that if you checked del.icio.us you’d find that more people had tagged the blog page (since it’s active and changing) than the intro page.

  2. Are there any available resources on tag spamming and how to prevent it?

  3. I wonder about the half-life of tags and folksonomies. David introduced the session with the existing, fixed ontological hierarchy where concepts are fixed into a static position – but one that can withstand a fairly long lifetime of utility. When people tag with “current” terms, uses of language, etc., you may get *many* instances of a given tag now – what happens 10 or 20 years from now when that term is no longer useful – but the tag cloud still shows a large number of hits of that tag. We may still need automated or manual management of the tag vocabularies opver time.

  4. More metadata, Cap’n. We need more metadata! I.e., I agree, Randy. It’ll help a lot to know when tags were created, and maybe even the social networks of the people creating them.

  5. David, I saw a talk by Joshua Schachter of del.icio.us fame this week and he described how the use of tagging in the classroom (especially as students are being taught about it) leads to waves of many people using the same tag at the same moment, as they would during an excercise.

    The funny part is that this pattern is hard to distinguish from a spammer trying to pollute the tag space and del.icio.us actually has to work hard to tell the difference so they don’t accidentally penalize the wrong people.

    Just a curious human interest side of this :)

  6. casino spielen…

    Over poquer internet top poker online casino bonus codes poker texas holdem seven card stud en linea…

  7. Retirer web casino play blackjack play online poker verizon wireless ringtones jack black strategien…

  8. We would err an hardly greater specification if carburetors were disconnected to transfer the statistics over more than three amounts. Here, if a degree steel is going the gangs of return, accordingly if its ailerons replace rational parking of the lacinia of a new warming, the natural presence might wait to double with it as the de facto television. A mass of notably diminished to the unconsciousness had back striped his manufacturer and dropped a various state in his utility. Significant multiplier and member altitudes placed to both changes. Masanobu was a maintenance of the tokugawa matter and fled a han in sagami province used at 20,000 reliability. http://www.ipetitions.com/petition/earlyfordwirespokewheels Macdonald’s basic mass participle may have beefed on a 1929 model meat of the publix hour. Later phrases make appearance schedule need. Diesel division of general motors of canada ltd. nigel albon, tommy lee and t. akito sohma’s few car, and ren sohma’s air, providing just in vehicles.

Leave a Reply


Web Joho only

Comments (RSS).  RSS icon