Sandra Herber: Future Librarian

This week’s readings were thought-provoking.  Having tackled tagging and folksonomies through my group project and through the Crowdsourcing project, I admit that I still found myself struggling to see the benefits of folksonomies over traditional cataloguing and indexing techniques.  After this week’s readings, I gained a new appreciation for some of the more subtle benefits of folksonomies (aside from the obvious and most quoted one which is that folksonomies use the vocabulary that the user uses) and but I also found that my idea that there are applications where traditional cataloguing is better than folksonomies and applications where the reverse is true was reinforced.  Since I am hoping to work in an academic library I constantly apply the ideas of this class to that context and I am still left wondering if folksonomies have a place in the academic library.

Folksonomies – their benefits

Aside from some of the obvious benefits of folksonomies (one is mentioned above and others have been discussed in previous posts), some of the readings showed me new or more subtle benefits.  Liz Lawley in Social Consequences of Social Tagging explained how tagging is better for browsing (and serendipitous discoveries) than for finding: “browsing the system and its interlinked related tag sets is wonderful for finding things unexpectedly in a general area.”  This was interesting because I had been bemoaning the fact that folksonomies would not allow users to locate the best resources the way traditional cataloguing would.  But now I realize that that is like saying that a wrench doesn’t work as well for hammering in nails as a hammer does.  Folksonomies can help you find items, but they are really most appropriate for serendipitous browsing.

I thoroughly enjoyed What is Social Cataloging? the video of Tim Spalding speaking at LIANZA.  I had heard about LibraryThing but haven’t used it and he did a wonderful job of outlining some of the incredible benefits (from giving access to the library of Thomas Jefferson, to connecting with other people who share you reading interests) of his system.  Now, I hesitate to say that all this is due to folksonomies – more, I think should be attributed to Spalding’s vision and ability to see the new applications for social cataloguing.  I found his opinions about OCLC interesting – and agree that public libraries should be striving to be indexed in Google since that is where people are looking for information anyway.

Folksonomies – for the Web

I had mentioned in a previous post that when confronted with the huge, amorphous mass of the Internet, folksonomies are just about the only appropriate way of organizing information.  This idea was echoed in Adam Mathes’ Folksonomies – Cooperative Classification and Communication Through Shared Metadata and by Emanuele Quintarelli in Folksonomies: power to the people who describes the information produced by the Net as “an enormous, ever-changing, time-sensitive, not-clearly defined corpus of items”.  When put like that, which cataloguer would want to try to apply LCSH to that?  Quintarelli says that “Well-designed metadata is better than folksonomies on traditional axes of comparison.”  This is what I believe.  The popularity of folksonomies and tagging on the Net is not going to stop librarians from using LCSH.  He goes on, however, to make this emphatic point which I think may have been missing from my (and other’s people’s) analysis of which system is ‘better’: “it does not matter whether we ‘accept’ folksonomies, because we are not going to be given that choice. The mass amateurization of Web publishing makes the mass amateurization of cataloguing a forced move. Folksonomies are a trade-off between traditional structured centralized classification and no classification or metadata at all. And they are the best we actually have.”

Folksonomies – for academic libraries?

So, we have now established that folksonomies cannot be avoided on the Net and are actually quite appropriate for that environment.  I agree with this.  The more tricky question for me is whether they are appropriate (and in what circumstances) in academic libraries.  Carol Ou in White-Paperish Thing (about distributed classification) explored one area: cataloguing of journals and decided that the benefit of getting users to tag these sources was that “it shifts that burden precisely towards a group with the greatest stake in certain resource discovery, relying on the concept that users are most likely to be willing to participate in a system that enables them to enhance their own access to those resources they themselves find most relevant.”  So, we are appealing to the personal interests of the users (somewhat following the idea that personal interest comes before community interest in successful social bookmarking enterprises such as Delicious and LibraryThing), but is there enough incentive to do this when traditional methods work ‘well enough’?  I look at the Encore system at UWO and I must admit that I have not found one item that has been tagged.  Is that because there is not enough personal incentive? Or that users feel that the catalogue of an academic library is sacrosanct compared to the Internet?  I know that no one is suggesting that academic libraries stop using LC classification or LCSH, but I just wonder about what added benefit could come from tagging in an academic library and if there is one, how we could get our users to engage in this process.

When I bookmarked my pages for this assignment (which asked us to bookmark at least 10 pages on Delicious with content applicable to this class) I was strangely thrilled when I found a page that had not been bookmarked before.  It felt as though I was introducing this useful piece of information to the Delicious community at large.  However, after I had bookmarked my pages I noticed that I was not really interested in other sites that had only been bookmarked by one person; I was interested in those that had been bookmarked by hundreds or even thousands of people.  I think my approach was individual (and my sites were part of the long tail) but I was drawn to the bookmarks that are the product of the wisdom of all Delicious members. Just as Andrea Mercado and Joshua Porter said, the original impetus to using these sorts of tools is individual, but the great benefit comes from harnessing these individual impulses to create the wisdom of crowds.  Does that mean that the odd and non-mainstream site does not get noticed, or does that mean that by bookmarking it in the first place in Delicious I am bringing it into a space where the ‘crowd’ has a chance to discover and publicize it?  I hope it is the latter.

I found the actual process of tagging difficult and I realized afterwards that it was because I was trying to do two things at the same time: bookmark for myself and bookmark for others.  These are not necessarily mutually exclusive, but tagging is much easier to do when you know why you are doing it.

The other issues I found with this process were ones discussed in my group project – the problem of trying to create tags for complex ideas or even for names that have more than one word (Delicious only allows single word tags) and the issues of using capital letters and plurals.  I also found that as I searched for pages and then tagged them, I (a single tagger) did not always use the same word for the same topic.  In traditional methods we worry about inter-indexer reliability, but I don’t even have interpersonal reliability!

So, there is no question that there are many issues with tagging and bookmarking on a site such as Delicious, but there is a great benefit (especially when using a common tag such as our class was doing).  As I write this post, there are 1,439 pages tagged with LIS9763.  That is an amazing number of pages and as I scrolled through them there were sites which jumped out at me and which I was curious to explore.  In determining which sites interested me I used a traditional method (title and notes) but I also looked for sites that had been bookmarked often.  This reinforced for me that though Delicious is a new and different way to find useful information it does not necessarily eclipse older methods of scanning and evaluating (and cataloguing).

The Goal

Our group wanted to create a subject guide on Black Canadian History that would be both easy to keep up-to-date and which would involve the larger community of a public library system.   We decided that a social bookmarking tool would satisfy both these criteria.  We chose to use Delicious as our underlying social bookmarking site for a number of reasons: it is well-known, so our patrons who have used social bookmarking before would probably already be comfortable with it, and it is generally easy to use.  We wanted our prototype to look like the webpage of a public library, so we embedded our Delicious page into a mock-up of a library webpage we created using NetVibes.  Interestingly, a number of libraries such as Central Medical Library at the University Center Goningen and Dublin City Public Library are actually using NetVibes to host their sites.

Our Library and Our Audience

Our library, which we called The Large Metropolitan Cosmopolitan Public Library (LMCPL), was modelled after the Toronto Public Library, as system which has a very large and diverse community.  The audience for our social bookmarking subject guide would be the whole community but we really wanted to get the Black community involved in contributing to a subject guide that would reflect their experiences and give them access to sites to learn about their history in Canada, and their role in the making of Canadian history.

Our Process

Because our group was spread out geographically (two people in Toronto, one in Waterloo and one in London) we ended up using some social software tools to bring this project together.  We met a number of times online in a Meebo chat room we created for our group and we edited this blog post as a group in Google Docs.

One of our group members did the initial creation of the subject guide in Delicious, much as a librarian at a public library would do.  Then each of the members of the group added other sites and other tags to simulate how the community could use this tool.

The Prototype

You can see our finished project at the LMCPL Library website.  Our goal was to make a compelling and attractive portal to our social bookmarking-based subject guide, to allow our patrons to interact with the subject guide right from the library’s webpage and to keep the subject guide fresh and up-to-date.  We achieved only some of our goals.  We believe the site we created using NetVibes is attractive, and because it has a widget which displays the list of our Delicious bookmarks, it will remain fresh as long as people add new sites (the list that appears on the page shows the most recently added bookmarks). Using Netvibes also allowed us to embed relevant videos and artwork which further enriched and widened the scope of our subject guide. The ability to leave comments also adds interactivity to the subject guide through the NetVibes portal.

Our Issues

Interactivity. The major issue (and a big disappointment to our group) was that we were not able to set things up so that our patrons could add sites and tags right from our library page.  It is quite straightforward to embed a Delicious widget into NetVibes, but as it stands now, patrons would have to go to Delicious to add bookmarks or tags.  This isn’t a problem if patrons have their own Delicious account where they can tag items themselves, but for us, the goal here was to establish a social bookmarking tool that was interactive and gave the patron an option to add tags and links to our existing account to centralize the information. Since we were not about to make the user name and password of our delicious account public (for security reasons), we considered switching the project from Delicious to an open source software program like SemanticScuttle. SemanticScuttle would allow us to create a social bookmarking subject guide that would function much the way enhanced library OPACs work. Users would need to create an account and be logged on to contribute to our database of tags and bookmarks. As it stands, none of us were familiar enough with MySQL or PHP let alone had access to web space that would allow CGI scripts (UWO’s Panther web space does not allow CGIs) required to run the SemanticScuttle software.  Our solution (which was only partially successful in the minds of the group members) was to tag all the sites with the tag “LMCPLBlackCanadianHistory” and to tell our users add other tags those sites and bookmark other sites with that tag (which we could then add to our Delicious list and which would then appear on our NetVibes page).

Wonky Widgets. Some of problems were caused by NetVibes and their widgets.  The widgets which seemed to solve our interactivity problem did not work as promised. As well, when you click on any of the hyperlinks in a widget, it brings up the link inside the widget itself, instead of in a separate window.

Browser Compatibility. While our library website looks great in Internet Explorer, Safari and Firefox, the widgets for the Delicious bookmarks and the tag cloud do not appear when the page is accessed in Google Chrome.

Authority. We, as a group, created a list of sites that we thought were useful, authoritative and current but we realized that once we opened up the our social bookmarking account to our patrons, our subject guide would become a mixture and include sites that were not necessarily as authoritative as we, as librarians, might like.  But this, of course, is the nature of social bookmarking and any social activity.  The benefits of this social bookmarking is that the addition of tags would personalize subjects and bookmarks for our patrons.

Folksonomies in Social Bookmarking sites. Another issue we had was in naming our tags.  On Delicious (like most social bookmarking sites) you can only use one-word tags.  So, when we were trying to express a complicated concept such as Black Canadian History we used the tag BlackCanadianHistory, but we quickly realized that our community might use tags such as BlackHistory, CanadianBlackHistory and any number of variations of our original tag.  This gets right at the heart of some of the issues of using folksonomies versus using controlled vocabulary.  One member of our group is doing cataloguing work in her co-op, so it seemed natural that she would have concerns with it, but we all found that, like proper librarians, we felt a little uncomfortable with the free-for-all nature of folksonomies. Although our tags might not necessarily harmonize with other tags that general Delicious users might have chosen for this subject matter, when tagging in Delicious, previous tag words show up as you type in the letters. This at least helps maintain some level of consistency in word choices, although it hardly solves the problem of the lack of controlled vocabulary.  Another issue is related to the order that the bookmarks appear on the Delicious account. The bookmarks appear according to the order that they were bookmarked. Although the reader can click a particular tag from the list on the right side of the screen in Delicious and those particular tagged bookmarks will come up, there does not appear to be a way to organize the bookmarks once they are input other than by date of bookmarking or by popularity.  Assuming that the patron wants to peruse the bookmarks, and we now have 41, they do not appear in any logical or pre-arranged order.  We understand that we could have used tag bundles to group the tags on our page, but they will still appear randomly (last tagged) in the widget in NetVibes, not in the order they would usually appear in a subject guide (general to specific).

Pushing our non-tech savvy patrons. We couldn’t make it easy for our patrons to add and tag bookmarks from our library webpage, so patrons would have to go to the Delicious site to do any of these things.  We were worried that some of our less tech savvy patrons would just give up on such a seemingly complicated process.  As well, only the first 12 bookmarks (those we added first) appear in the widget on the library webpage.  Would our less savvy patrons assume that there were only 12 bookmarks in our list? Another problem uncovered here was that the overall layout of the Delicious widget makes for a cramped display. To clarify, both the website link as well as the tags are displayed in one line with only a “/” to separate website link from tag links. Will a non-tech savvy patron understand the difference between the tag and the associated link by looking at this page?  It is possible for our users to access an RSS feed our Delicious bookmarks (from Delicious) so they can keep up to date with changes, but that would only be for the more tech-savvy ones.  We hope that by embedding the bookmarks in our library’s NetVibes site, we’re making it as easy as possible to follow any changes (certainly that is one of the benefits of the fact that the widget in NetVibes lists the most recently added sites first).

Longevity. As noted in the Corrado and Fredrick reading, it isn’t very likely that Delicious will disappear overnight but because it is a free, third-party service it is, however, still necessary for any library using social bookmarking to find a means of backing up this information. Our group members have shared our links via email and some of us have saved them in Word documents – while these are not necessarily the best solutions it does still ensure that we have access to the majority of our information.

I’m in the group responsible for this week’s project and I’ll be posting our prototype and site soon, but since I’ve engaged with these ideas (both through the readings and our project), I thought I would do a blog post this week as well.

In Hammond’s article Social Bookmarking Tools (I): A General Review it is pointed out that “to anyone familiar with top-down classification schemes, this approach [social bookmarking] could look like a fearful muddle.”  The “anyone familiar” are creators and users of traditional hierarchical classification systems – that is, us (librarians and librarians in training) among others.  Though I do see the benefits of social bookmarking, I am very aware of its drawbacks and I thoroughly believe that the hierarchical systems that we have been trained in are superior for giving access to bibliographic material (that’s why I wonder about the benefits of encouraging students to tag material in the new UWO catalogue).  But tagging and folksonomies are not being used to organize bibliographic materials (or at least they haven’t taken over from traditional classification systems) they are being used to ‘catalogue’ and ‘organize’ that great amorphous mass – the Internet.  The Internet cannot be catalogued by traditional means (I guess Yahoo tried that and look what happened to them when Google came along?).  The Internet has been created ad hoc and should be classified ad hoc.  But as Hammond says, “a free tagging approach to classification is a jumbled, hit-and-miss affair, and any system that it may throw up must be discovered, or learned, after the event” and that is where we are now: trying to understand the system that has been created.  The greatest benefit, to me, to this type of system is that it may tag material with the terms that users would actually use to find it – but as our group discovered, there are many, many ways to name the same concept.

I thought Rainie made an interesting point in Tagging: “some worry that folksonomies can be a type of ‘tyranny of the majority’, in which the prevalent group’s way of thinking about the world overwhelms the local and the quirky.”  This was interesting to me, since in the same article Rainie points out that taggers are mostly under age 40 with high levels of education and income.  They are the people determining what is noteworthy on the Internet and that is both good and bad.

I enjoyed Joshua Porter’s The Del.icio.us Lesson where he made the very good point that tagging works not because people do it for altruistic reasons, but because it will benefit them directly.  The network value only comes in when we aggregate what people have created.  Andrea Mercado, in Tagging on Flickr & del.icio.us, makes a similar point: “Flickr follows a ‘desire lines’ philosophy, letting people create their own metadata, laying paths where people are walking instead of trying to lay out paths and assuming people will follow them (like, say, structured classification)”.  This, I think, should be the strategy that we take when using social software tools in libraries.  We need to make sure that we aren’t laying out the paths and hoping that people will want to walk along them, but rather letting people create the paths because they see the use in them.  After they have done that we can find a way to create a network benefit from all their work.