Judith Koren Global Collaborative Search: Watch This Space
Jinfo Blog

1st November 2007

By Judith Koren

Abstract

The Web's current evolutionary phase, Web 2.0, highlights user- generated content. These days, anyone can publish anything, and everyone shares everything. They do it partly to gain or enhance a reputation ('Here are the coolest/most popular videos on YouTube') and partly out of a real desire to make contact with and collaborate with other people. For the first time ever, the scope of collaboration and social networking is transcending geographical boundaries and aspiring to be truly global.

Item

Judith KorenThe Web's current evolutionary phase, Web 2.0, highlights user- generated content. These days, anyone can publish anything, and everyone shares everything. They do it partly to gain or enhance a reputation ('Here are the coolest/most popular videos on YouTube') and partly out of a real desire to make contact with and collaborate with other people. For the first time ever, the scope of collaboration and social networking is transcending geographical boundaries and aspiring to be truly global.

Since the Web is a major artery for information, the lifeblood of our profession, our tools and 'occupational culture' tend to evolve with it. So we're hearing a new buzzword: collaborative search.

Truly global collaborative search - both international and universal in reach - would be a way for all information professionals world- wide, whether employed or independent, to discuss and help out with the search needs of others. For our clients, end-users world-wide, it would be a way to have access to advice, help with searches, and, if they wanted, the paid services of any information professional. It would involve a place where end-users and info-people could meet and interact; and also where sub-groups could meet, such as info-people discussing search questions between themselves. How close are we to this concept of global collaborative search?

Not, it appears, as close as the buzz implies. As I tried to answer this question, I found that everything out there now could be split neatly into two types: on the one hand, services for librarians and information professionals, and on the other, services for the end- users. And never the twain shall meet? Let's take a look at what there is, and then discuss what there could be.

Services for the pros

1. Mailing lists

Long before the Web was born, reference librarians virtually lived in listservs such as BUSLIB-L, MEDLIB-L and STUMPERS-L. They were 'push' services - they landed in your email inbox, constantly reminding you that they were available even if nothing at the moment was of interest. And they were great collaborative search tools - a request for help would usually get an expert answer within a day or two. But since the 'Net back then only served the governmental and academic sectors, their reach, while international, was not universal.

The need for subject-focused discussion of reference questions still exists: when I describe mailing lists to younger reference librarians, they tend to look wistful and say, 'I could really do with something like that.'

2. Professional associations

Join SLA or AIIP and you find a community of peers eager to help with search problems (and anything else). They excel in collaboration, from advice to subcontracting. They include community listservs - 'push' search-help services. But professional associations can't reach everyone. That's because they:

  • Are aimed at specific subgroups - AIIP at independent information professionals, SLA at 'special libraries' and corporate information centres

  • Are, for the most part, US- or Western-centric despite attempts to reach out to a wider global community

  • Charge Western-level subscription fees that put them out of reach in many parts of the world.

3. Web communities

FreePint is perhaps the nearest we've come to date to a real global community of information professionals. Its membership is an order of magnitude larger than that of the professional associations. Like them, it's got a place to ask questions -a large percentage of which are requests for help with searches. It's got added services, like a job bulletin board. And the FreePint Newsletter is clearly collaborative - info pros updating other info pros.

But even though FreePint has a wide reach, it doesn't offer everything a search community needs. It doesn't really enable outsourcing - collaboration on projects between professionals, such as AIIP excels at. Like the listservs and associations, it doesn't give access to end-users or enable collaboration between them and professionals.

Services for end-users

1. Answer-type services

These are more about Web 2.0 than about search. Yahoo! Answers, for instance, promotes itself as 'a new way to find and share information. You can ask questions on any topic, get answers from real people, and share your insights.' This is fun, especially since any real person can reply, and what gives the service its warm'n'fuzzy Web 2.0 feel is that those who reply are from the same demographic as those who ask - they're not info pros. And the answers they supply are more opinion than fact.

So there's a lot of sharing going on, but there isn't much finding (in the info-pro sense of 'looking for') and if you're an information professional, you find yourself wondering whether what's being shared is really 'information'.

But these services indicate a felt need for help on the part of end- users, and a willingness to use collaborative tools to get that help.

2. Expert-type services

These all have one central feature in common - they're a for-fee service, in which the client can check out the experts' profiles but eventually has to choose one expert and pay a fee (usually per-minute) for a consultation. They provide a marketplace, which the collaborative communities lack. But the Experts themselves are competing, not collaborating; and it's sidelining the issue to call the sale of professional services 'collaboration', let alone collaborative search.

Information professionals are conspicuously absent from both Answers and Experts sites. Go into Guru.com or Elance.com or pretty well any other 'experts' service, and you find at best an 'Internet Search' (!) category where the vast majority of the Experts are programmers/geeks. Why? It's not because 'we don't do that type of information' - we certainly do. Is it because we don't 'do' online help charged by the minute? But reference librarians certainly 'do' online help, not charged extra for at all. Would we feel more inclined to collaborate if the service were vertical - a collaborative-search community, rather than a general experts site?

3. Community information-sharing tools

Well, we've got wikis. They're good for collaborative content- building; but they're not so suitable a platform for collaborative search.

And we've got social bookmarking sites. You can search them and find collections of bookmarks made by others. You can post your own and find people like you. But, as in answers-type services, what's going on here is sharing and social networking, not collaboration.

Finally, there are a few new services - perhaps in them we'll find our Holy Grail of collaborative search?

  • Mahalo <http://www.mahalo.com> is a compendium of subject resources built by Mahalo's 'Guides' or contributors. It's collaborative in the sense that anyone can suggest a link - but then so you can in the ODP or Yahoo!'s directory. Isn't this simply an updated version of a Web directory?

  • Trexy <http://www.trexy.com> offers a downloadable toolbar which records the URLs of pages you've visited after using a search engine. This list of URLs is automatically created and uploaded to Trexy's site, where it becomes a 'trail' for the keyword you searched. Other people can see it; and the site co-locates the most popular sites from all the 'trails' suggested for the same keyword, into one list.

This is a cute idea, especially since it does give end-users something that only info pros have had till now: an easy way to share good resources you've found while searching. But collaboration, again, is limited, because you can't share your knowledge of how to search, only the URLs of sites selected from a search-results list. You can't add comments or advice about those sites.

Where does all this leave us?

I emerged from this quick survey with a few conclusions:

  • Info pros feel the need to collaborate - to get ideas for how to approach a project; to get suggestions for good sources; and sometimes to outsource to others

  • End users feel a need for direct access to people who can answer their questions. They love free community-type services, asking each other, but they'll also turn to professional experts, and at least some of them are prepared to pay for the service

  • All the 'collaboration' tools I found are variants of known remedies such as resource lists, social bookmarking and discussion groups. They tend to address either the community of info pros, or the community of end users. There isn't a place where endusers can interact with search specialists (as opposed to whizkids who think they're search experts because they dream in Java)

  • End-user tools for finding information don't often involve searching for it; and when they do, they're geared to simple searches - the sort that don't help when you're researching a complex question

  • There are no current resources for helping end-users perform complex comprehensive searches - the kind that go beyond brief factual replies to a defined specific question

  • Services that try to be global are nonetheless Western-centric: the non-US, non-European market is not well represented in them.

So we need GCS - Global Collaborative Search

A Web-based GCS community would support:

  • Searchers collaborating with other searchers to build lists, or 'trails', of 'best resources' for a specific query, lists that are more content-rich than what's now available - not just links but also metadata about the resources and tips on how to use them; AND

  • End-users looking for such lists - these are the ones who want to learn how to fish (without having to choose an expert and pay in order to ask the question); AND

  • End-users looking for research services - these are the ones who want their fish filleted on a plate with the right sauce added; AND

  • End-users and info pros who want to comment on and add to the resource lists already built and the questions already asked; AND

  • A way of comparing 'trails' for the same or related questions, real- time; AND

  • A way of supplying all this as a 'push' service - sending all the trails for a question to anyone, anywhere, or putting it in a blog or Web page and having it continue to update itself, without requiring you to return to the original GCS website, as more collaborators add suggestions, comments, other queries and other related trails of sources.

How can we do this? Do librarians and info people want to do this, considering our absence from the experts and answers sites? What's stopping us from reaching out to the wider community of information- seekers? That's what we'll be discussing at the Round Table on Collaborative Search at the upcoming London '07 Online Conference.

I believe that the main things holding us back are lack of incentive and lack of suitable tools. So I and my colleagues have made a start with TrailMap, a Flash search-trail visualisation that solves the last problem in the list above. It's a packaged list of subject resources (Trail) and associated metadata that you can send to anyone or post online, where it continues to update itself real-time. And we've got some ideas about all the other points too. But I've already overrun my space in FreePint, so join us at the Online Conference for the next instalment.


Related FreePint links:

« Blog