Lorcan Dempsey

Metadata sources

dempsey — Sun, 20 Sep 2009 18:36:37 -0700

A while ago, I suggested that it was interesting to think about four sources of metadata in our systems and services:

Professional. Produced by staff in support of particular business aims. Think of cataloging, or data produced within the book industry, or A&I data.
Crowdsourced. Produced by users of systems.Think of tags, reviews and ratings on consumer sites.
Programmatically promoted. Think of automatic extraction of metadata from digital files, automatic classifcation, entity identification, and so on.
Intentional. Data about choices and transactions which support analytics or business intelligence services. Think about ranking, relating, recommending in consumer sites (e.g people who like this also like this) based on collected transaction data.

We were discussing these types of data in a meeting at work the other day, and it occurred to me that what I call here crowdsourced, programmatically promoted and intentional data are all ways of managing abundance. Our model to date has been a 'professional' one, where metadata is manually created by trained staff. This model may not scale very well with large volumes of digital material. Nor does it necessarily anticipate the variety of ways in which resources might be related. The other sources will become increasingly important ....

Waking the unread

dempsey — Wed, 23 Sep 2009 18:38:11 -0700

While looking at Google Book Search the other day I was intrigued to discover that parts of my early oeuvre, such as it is, have been digitized from the University of Michigan collections.

I was particularly struck by the word clouds. Here is a word cloud from a 1991 work on libraries and networks with special reference to the now long forgotten OSI suite of protocols.

Here is the word cloud to a 1988 collection of conference papers on library automation, mostly with a UK focus.

In each case, I think the cloud does a reasonable job of reminding you of topics of interest at the time. Here are a couple of other volumes. Here is the cloud from Charles Hildreth's influential earlier work (1982) on online catalogs:

And here is the cloud from the first edition (1992) of Ed Krol's very influential The whole Internet: user's guide & catalog ..

They have a nice quality of 'glancability' ... not a definitive guide but useful hints and hooks.

Reputation enhancement

dempsey — Tue, 29 Sep 2009 13:30:41 -0700

Reputation management on the web - individual and institutional - has become a more conscious activity for many, as ranking, assessment and other reputational measures are increasingly influenced by network visibility. In particular, it raises for academic institutions an issue that has become a part of many service decisions: what is it appropriate to do locally? What should be sourced externally? And what should be left to others to do?

Think, for example, of faculty profiles: the managed disclosure of expertise and research activity. This has often been an informal personal or departmental activity. However, there is now a variety of institutional initiatives which may pull together data about expertise, experience, publications, grants, courses taught, and so on (see OSU Pro at OSU, or Vivo at Cornell, for example). Such initiatives may sit between between several organizational units on campus: Research Support, PR/Communications, IT, Library. They are also at the intersection of different systems: enterprise (Peoplesoft, for example), course lists, research/grants management, bibliographic. At the same time, researchers may have presences in emerging network level research social networks (Mendeley or Nature Network for example), in disciplinary resources (Repec, for example), and, of course, in general use services (Linkedin, for example). There are also commercial services which support such activity in different ways, Community of Science or Symplectic for example.

In this context, here is a note about several unrelated initiatives which I have come across in the last week or so. I don't try and create a single narrative around them, but together I think they point to this emerging sense of reputation management (or enhancement) as an important if not yet fully clear service category.

We are exploring such a service category in our Research Information Management theme. It looks at the intersection of library services and research administration on campus, and we are thinking about the variety of library services which might emerge (which include, in the context of this entry, bibliographic support, bibliometric advice, effective disclosure of expertise and research to the web, advice about SEO and copyright, and so on).

VIVO: research and expertise across Cornell. Some colleagues from Cornell visited last week (see details and video of Anne Kenney's presentation here) and VIVO came up in discussion.

VIVO (not an acronym) brings together in one site publicly available information on the people, departments, graduate fields, facilities, and other resources that collectively make up the research and scholarship environment in all disciplines at Cornell. [About VIVO]

Managed within the Library, it draws together a lot of data from various sources. Interestingly, it is based on Vitro, an 'Integrated Ontology Editor and Semantic Web Application'.

Bibliometrician. The University of Leicester Library advertised for a 'bibliometrician' whose role would be "to provide high-level expertise and advice to the University on the use of bibliometrics and related policies in the external and internal evaluation of the quality of the University's research"

Mendeley. My colleague John MacColl wrote a blog entry about Mendeley last week. Mendeley describes itself as being like "iTunes for research papers": "Organize, share, and discover research papers! Mendeley is a research management tool for desktop & web. You can also explore research trends and connect to other academics in your discipline." John contrasted Mendeley and institutional repository incentives and user experience for researchers. Mendeley is one of several social networking sites aimed at researchers.

Manchester escholar. The University of Manchester launched its repository service. The first line of its mission reads: "sustain and enhance the research reputations of individuals and organisations affiliated with The University of Manchester". It is also interesting to read the 'project business case and benefits' which have a strong reputation management focus. The first benefit for the research is "increase the visibility of your research findings, your work is easier to disseminate, easier to find and easier to read". The second emphasizes convenience: "make it easier to manage your list of publications on your personal website and your organisations website". For institutions, the first-listed benefit is "demonstrate to its employees, in particular the academic community, that individuals and their work are valued, by supporting mechanisms that reduce workload and maximise the benefits to them of their efforts" and the third is to "increase the visibility, reputation and prestige of the institution".

Ranking economists and Repec. Greg Mankiw is a Harvard economist, text-book writer, high profile blogger and sometime chair of President Bush's Council of Economic Advisors. He refers from time to time to a ranking of economists generated by Repec. He wrote a note about this year's Nobel prize last week, and pointed to the Repec ranking, noting that 6 out of the top 10 on the list had already won. The rankings are based on data about authors who have registered with the Repec Author Service, which aims "to link economists with their research output in the RePEc bibliographic database". Authors get a profile page and also receive statistics about downloads of their papers and citations to them. Many rankings are generated from the system.

Finally, I noticed the following tweet from Danah Boyd "It pains me when academics don't take care of their search engine presence. RateMyTeacher should never be an academic's top result".

Search engine presence is increasingly important to people and to institutions ... reputation management is emerging as a new service category which should be of interest to libraries.

Discoverability .. a report that's worth a look

dempsey — Wed, 07 Oct 2009 18:45:34 -0700

We are awash in assisted thinking, as I may have remarked. One document that is worth a look is Discoverability produced earlier this year by a team at the University of Minnesota.

In October 2008, the Web Services Steering Committee at the University of Minnesota Libraries created the Discoverability exploratory subgroup, charged to recommend ways to make relevant resources more visible and easier to find, particularly within the user's workflow. This report shares the findings of Phase 1, in which the primary activity was data‐gathering and analysis. Phase 2 of the group's work will take the discovery principles identified here and recommend specific strategies for the future. The report consists of four main sections. The first section is a brief description of the process and methodology. The second is a discussion of five key trends related to discovery that were identified in the literature, including a description of how each trend is reflected in current use of local systems. The third section contains a set of suggested principles to guide future decisions related to discovery. Finally, we have collected and analyzed usage data from many of our local systems. These reports are collected in our fourth section and are summarized in "A Month of Library Discovery". We have also included specific recommendations regarding future data‐gathering and analysis. Our appendices include a copy of the group's charge, a review of discovery principles at peer institutions, and a set of web statistics reports for the University Libraries' many websites. [Summary of Discoverability: phase 1. Final report. U of Minnesota.]

I particularly liked the way in which they identified trends through their environmental analysis and then looked for corroboration (or otherwise) in their own usage logs. This combination is quite effective.

Here are the trends they identified:

They tracked library systems usage during a 'month of library discovery' and present this summary of findings:

There are many interesting nuggets. In a discussion of the power of recommendation and 'collective intelligence' to drive traffic to resources, for example, they note that almost 15% of people who came to the Archives and Special Collections area of the website came there from Wikipedia. There is interesting discussion of how resolver requests break down between internal (e.g. A-Z lists) and external sources (e.g. Google and PubMed). The majority of library website pages are visited infrequently.

The report is available from the library's repository. I hope that the authors are considering putting some of their findings about patterns of usage into the literature.

Related entries:

Affinity strings, personalization and recommendation

Personal reference collections as digital libraries

Research behaviors and the library

Getting in the flow - beside the Mississippi ...

Government URIs: a write to reply

dempsey — Sat, 17 Oct 2009 19:54:38 -0700

Tony Hirst alerted me to an interesting document on the structure of URIs in UK government websites. There were two things of immediate interest.

The first was the emphasis a Government agency was putting on information architecture in a web environment. Other documents will follow. This is from the introduction:

8. URI sets will be an integral component of a UK Public Sector Information Architecture that supports many goals including the release of government data, reduced duplication, and increased information sharing towards transforming government services.

9. URI sets can be published by the UK public sector to provide comprehensive and reliable identifiers for 'Things' such as schools, roads, legislation, locations, projects, events and so on. Where the quality of these sets can be described consistently, other data owners will have the confidence to re-use them in their own data, leading to a web of data that can be linked, queried, and aggregated.

10. The existing UK public sector standards for metadata and 'findability' work well when applied to documents, but are not sufficient to support a 'Web of Data', where each individual statement can be queried and linked.

The second was that they were making it available through the Write to Reply service provided by Tony and Joss Winn. This provides paragraph level publishing, citing and commenting facilities. The purpose is to provide a venue for public review of documents. The site uses CommentPress.

Community is the new content

dempsey — Sun, 18 Oct 2009 17:50:54 -0700

We are now very used to interacting with resources in a social context. The application of community to content, in terms of discussion, recommendation, reviews, ratings and so on, is evident in many of the services we use, and in some form in most of the major network servies we use (Amazon, iTunes, Netflix, ...). Indeed, this is now so much a part of our experience that sites without this experience can seem bleached somehow, like black and white TV in a color world.

In a reductive view, here are three types of social experience, which may be present singly or in combination in these sites.

Conversation
Connection
Context

.
We explicitly talk about resources (conversation). And the traces that we leave intentionally or unintentionally can be mined to create connections between people and to add context to resources (relating, ranking, recommending), based on patterns of association between them. What I am calling 'context' here may not be explicitly social, but as it is often mined from aggregate behaviors it does not seem too much of a stretch to include it.

Conversation. Conversation about services is a natural part of our experience of them. I bought a Zune HD [1] recently; it was not clear to me how to turn it off. A search threw up an answer on one of the several Zune forums. It also showed that others had the same issue, so we can expect that this particular feature will change in future releases. In fact, the detailed instructions we might once have seen with a device like this seem to be a thing of the past; in this case, even the online documentation is not very full. Maybe their decision not be exhaustive is influenced by the knowledge that a rich documentation base will be collaboratively sourced across multiple conversational forums? Forums may have dominant contributors, and employees of the product provider may particpate. Such signed network presences are common: we are used to seeing 'signed' reviews, recommendations, comments and ongoing interactions on music, movie, book and general consumer and social sites. And online conversation clearly influences online behavior. I have been interested to receive letters from vendors of items I have acquired through Amazon asking that I give them the highest satisfaction rating, and urging that if I am unhappy in any way to contact them first so that they can rectify issues and preserve their rankings.

Connection. I like this quote from Hugh MacLeod:

14. The most important word on the internet is not "Search". The most important word on the internet is "Share". Sharing is the driver. Sharing is the DNA. We use Social Objects to share ourselves with other people. We're primates. we like to groom each other. It's in our nature. [gapingvoid: "cartoons drawn on the back of business cards": more thoughts on social objects]

We connect with others by sharing information about ourselves. Networks form around 'social objects', the focus of these shared interests. Think of social bookmarking, picture sharing and social bibliography sites, for example, where we connect around shared interests, in, respectively, interesting resources, pictures and collecting/reading interests. Facebook and Linkedin connect people - or their online signed identities - based on the networks of connections they have already made. Users of services - the Zune for example - who sign up are offered the opportunity to connect with users of like interests, and can prospect the interests of those who chose to disclose them.

Context. We leave traces everywhere. We click, buy, rate, follow pathways, add to playlists. We also create collections, lists, and playlists, which disclose our interests and can be compared to make connections or to generate recommendations, or to seed other lists. Services use this subterranean data not only to make connections with other users but to create context, to configure resources by patterns of relations created by shared user interests and choices, and to use these patterns to broaden the experience of their users. Google mobilized linking behaviors; Amazon made 'people who bought this, also bought this' types of association popular. Such context is now a central part of music, movie and other sites. Think of the rich recommendation apparatus of Netflix, or the generation of channels, playlists and recommendations in iTunes, Last.fm or other music sites. This approach has spread to more academic contexts. Mendeley, a research management and social networking site for researchers, is explicitly modeled after Last.fm (and some Last.fm veterans are among its investors). One of its aspirations is to generate impact rankings and relationships based on patterns of collecting and use of research literature.

Here are some random observations that occur in this context ... Thoughts about libraries may follow ...

Visitors and residents.
Managing scale and guided navigation.
Customer relationship management.
Real-time tracking of trends and analytics.
Social experiences around content.

Visitors and residents. This is a useful distinction introduced by Dave White. The resident lives a part of their lives online; their web selves have become an important projection of identity and they maintain online networks of friends and colleagues. "They are likely to see the web as a worthwhile place to put forward an opinion". The visitor uses the web to get their work done. "They are sceptical of services that offer them the ability to put their identity online as don't feel the need to express themselves by participating in online culture in the same manner as a Resident." In the schematic advanced above, the resident is interested in conversation and connection as an active participant - their traces are visible. Everybody leaves the subterranean traces which conribute to context.

Managing scale and guided navigation. The social provides a layer of interpretation, connection, context, direction, filtering which valuably orients us in large information resources. Services mobilize 'intentional data', data about usage and choices, and crowdsourced data to manage abundance where 'professional' approaches may not scale. Of course this can be managed. Nicholas Carr recently remarked of Netflix: ".. what I've noticed is that the company has deliberately geared its search, filtering, and recommendation tools to lead customers away from newly released hits".

Customer relationship management. Services use the data generated by the activities described above to develop customized engagement with customers, pesonalizing communications, providing recommendations, and so on. Such data-driven communication is often useful, sometimes intrusive, but is clearly a priority, where scale, again, make other approaches difficult.

Real-time tracking of trends and analytics. Where there is a critical mass of participation, conversation and context can reveal emerging trends and behaviors. Twitter may be an obvious case, but consider, for example, what the realtime usage data coming from a service like Mendeley might tell us about academic interests. .

Social experience around content. I was struck by some remarks by Trip Hawkins, the CEO of Digital Chocolate and the founder of leading games publisher Electronic Arts, in an interview at the Web 2.0 Summit in 2007 [2]. He was talking about games in a mobile environment, where Digital Chocolate is active.
"In my opinion traditional content is dead ...", he said, and he went on to characterize traditional content as "about a playback and immersive experience and which involve a business model where you pay a fee for the privilege of escapism and checking out" (Hawkins, 2007). These traditional forms include reading and cinema experiences, and he suggests that participation in those media has leveled out. He contrasts this with a new type of content and associated experience, which is growing:
"... where the consumer is increasingly going to spend their money is on social value which is enabled by content where the content isn't for sale for its own sake - the content is there to enable improvements in your social life". (Hawkins, 2007) Depending on your point of view or cultural formation, this characterization might be plausible or startling ;-)

[1] Coda: A couple of eyebrows were raised at work when I mentioned that I had bought a Zune HD, the recently released third generation Zune. I just wanted a media device and this is very nice. It is growing on me. David Pogue notes that "the software design is fluid, beautiful and incredibly responsive".

[2] "Edge: Gaming moderated" by Morgan Webb with Trip Hawkins and Robert Kotick from Web 2.0 Summit, San Francisco, 18 October 2007, at http://blip.tv/file/441160.

Untangling the library systems environment

dempsey — Sun, 25 Oct 2009 19:37:17 -0700

NISO organized a meeting on library resource management a couple of weeks ago: I notice that the presentations are now available on the web. They make an interesting collection, and I return to them in a moment.

I have written about the library systems environment in these pages from time to time. A blog entry from Summer 2007 formed the basis of a Portal article [pdf] of the same name (Reconfiguring the library systems environment), which had these pictures:

Libraries manage a variety of materials workflows, each supported by a different systems apparatus. In each case there is a vertical arrangement, as materials are disclosed through discovery and delivery services which in turn relate to backend management services in closely coupled ways. Bought (often print) materials go through an ILS workflow and are disclosed through the catalog which is usually a part of the ILS itself. Systems to manage these materials in resource sharing arrangements may also be present. Licensed materials have an emerging apparatus of management systems support (ERM, knowledgebase, A-Z lists, and so on) and are disclosed in a variety of ways including metasearch and resolvers. A routine approach has not emerged for digital materials and they are managed in a variety of repository and other frameworks, and delivered to per-repository user interfaces. Different approaches may be taken with different categories - digitized special materials, web archives, institutional research materials, and so on - and metadata may be exposed for harvesting. Although not pictured here, one might add that there may also be additional workflows associated with archival materials or special collections. (Slide 4 of Mackenzie Smith's presentation [ppt] is an interesting depiction of the internal library systems environment showing the range of actual systems supported.)

The complex array of systems - at different stages of maturity and created in quite different technical environments - has encouraged a move towards some rationalization and integration. In particular, we can see a drive to integrate the management workflows across material types. Examples here are OCLC's Cooperative Webscale Management Systems initiative, Ex Libris's URM, and the open source OLE project. At the same time, we see two trends on the discovery side. The first is a drive towards deeper integration across types, both through greater use of metasearch and through actual consolidation of files. Althought it will not be comprehensive of all available materials, we will see much more of the latter approach for efficiency reasons. The second is a realization that library resources need to be disclosed more effectively to a variety of other environments, whether they are other institutional systems (e.g. course management) or external.

So, finally I note two major challenges. One is that on the management side libraries have to pull together a variety of systems and services whose legacy business and technical boundaries may no longer map very well onto user requirements. A second is that they have to project their resources into a variety of user environments and workflows over and above whatever integrated local library website environment they create. These include personal (people have a growing variety of ways of finding, saving and collecting information resources), institutional (think of course management systems, student portals, ...), and network-level services (search and discovery services; social networking, bookmarking and bibliography sites; ...). Related to this is the challenge of integrating community in library services: there is a gap between libraries as inevitably social organizations and the social component of the library experience on the web. A redrawing of this image today would have to include a better sense of user participation.

In this context, here are some incomplete notes on some of the presentations at the NISO meeting:

General overviews were given by Oren Beit-Arie [ppt] and Marshall Breeding [pptx] in a wide-ranging opening keynote and conclusion respectively. Oren presents a picture not unlike the one suggested above, discussing the need to move from content-based verticals to service-based horizontals. The context for much of what he says is provided by a summary of responses from interviews with librarians. Here are the headline requirements (adapted slightly): a single interface for discovery and delivery of all library/institutional assets; consolidated workflows, uniting traditional library functions with those of the "digital library"; collaboration to increase productivity, leveraging "network effect"; re-use of metadata; SOA-based interoperability; software-as-a-service deployment option; user-provided data. He characterizes evolving services within a traditional/transitional/transformational framework, where traditional involves doing the same thing differently, transitional involves doing new things in support of the transitional, and transformational involves doing new things. Moving services to the cloud is given as an example of the traditional, as is evolution of the cataloging environment; exploiting network effects (e.g. collaborative collection development, tagging of library images in Flickr, ..) is given as an example of transitional; and the mobilization of usage data to transform services (new metrics of scholarly evaluation, recommender systems, analytics and ranking of results) is given as an example of transformational services.
System sourcing decisions and their implications figure large in the agenda given the recent focus on open source approaches. Marshall Breeding [pptx] cryptically notes 'many unannounced open source projects' which may change the current low defection rates from established ILS vendors. How will open source initiatives work with emerging software as a service trends?
Recommendation and other ways of mobilising usage (or intentional) data figure in several places, in line with general trends. I suggested recently that we will see services which don't use direct user input in the form of tagging or review, and indirect in the form of usage data to support ranking, relating and recommending, as bleached, rather like black and white TV in a color world. See Kevin Kidd's presentation [ppt], as well as Oren's, on this issue.
Scalar emphasis has become an important question for libraries. At what scale should things be done as institution-scale is increasingly the wrong level for many activities? Oren discusses the transitional effect of the network in broader collaborative settings, where the power of the network can be leveraged to improve services. Shared cataloging and resource sharing may be earlier instances of this. Consider now the potential for recommendations where circulation or other usage data is aggregated at a higher level. Consider incentives also in this context. Where are library users most likely to want to invest their effort? Kat Hagedorn [ppt] discusses a collaborative project of the HathiTrust, New York University, and the partners in the ReCAP shared print facility with the involvement of OCLC Research and CLIR. What policy and service apparatus needs to be in place to provide confidence of supply from HathiTrust and ReCAP sufficient to allow NYU relegate materials from its own collection? Such 'cloud library' provision will become more common as libraries seek to transfer resource away from 'infrastructure' and towards user engagement. Kyle Bannerjee describes [ppt] Orbis Cascade's work with OCLC on the integration of local, consortial and global discovery and delivery of resources. He suggests that operations should move to the highest appropriate level in the network, and speculates about what other services should move to the network level. Rachel Bruce [ppt] looks at library systems from the point of view of national-scale 'shared services'. What these and other presentations show is how decisions about level of operation - personal, local, consortial, national, global - are as important as particular discussions of functionality or sourcing. Libraries face interesting choices about sourcing - local, commercial, collaborative, public - as they look at how to achieve goals, and as shared approaches become more crucial as resources are stretched.
Academic library systems are part of an enterprise infrastructure, which is discussed here by Mackenzie Smith [ppt] of MIT and Diane Mirvis [pptx] of the University of Bridgeport. For me, these were the most interesting presentations here as they point to a set of influences that are not discussed very often. As more activity takes place on the network, as students, faculty and administrators create and use data from many sources, and as there is more pressure for new types of integration on both user and management side, it will be interesting to see how organizational and system boundaries change within institutions.

Note: Updated with minor edits, 26 Oct 09.

Research support services

dempsey — Sun, 25 Oct 2009 23:51:09 -0700

I am pleased to note a collaboration between OCLC Research and the Research Information Network in the UK to explore changing research support needs in universities.

We tend to focus on how technology changes library practices, but the impact of technology on libraries will be less important in the long term than the impact of technology on research practices, which in turn the library must be well-placed to support.

The project will investigate researchers' needs and desires in a small sample of UK and US universities to identify the significant patterns, intersections, gaps and issues from researchers' points of view, whatever the source of such services. The nature and scope of research support services will be documented, providing examples of good practice, recommending areas where new practice might emerge and identifying possible areas and scope for collaboration within and between institutions. Comparing transnational academic practices will provide evidence and encourage coordination to meet the needs of academic research internationally. [OCLC Research and The RIN announce joint project on research support services]

This work is being carried out in support of the RLG Partnership as part of the Research Information Management theme and we anticipate completing it in May 2010.

Community bibliography

dempsey — Sun, 01 Nov 2009 18:15:35 -0800

I prefer 'crowdsourced' to 'user contributed' but neither works very well for me. In particular 'user contributed' does not seem a good term at all for a variety of reasons. Anyway, I was looking at the new catalogue at Ottawa Public Library powered by Bibliocommons earlier (following a mention by Stephen Abram).

There is much to like about this. It is not organized around the bibliographic record, but places that data in a context more aligned with user interests. Indeed, the bibliographic data 'block' is hidden behind a tab. The most prominent features, alongside the picture, are user actions: it invites readers to do things. You can add data, add it to your own personal collection, add it to a list. I liked the Browse the shelf feature. This is one of several options which allows 'exploration' based on attributes shared with the displayed item.

What jumped out at me was how the block at the bottom of the screen was named: Community activity seemed like a very good name in this case.

The community features are interesting and feel familiar from other environments. A reader can 'connect' with others who have selected the item being viewed. The connections work across libraries in which Bibliocommons is deployed. Here is one of the people listed on this display.

The left hand column here carries through the 'narrow by facet' feature found on results pages.

Some details about Bibliocommons are provided in an article in the October 2009 issue of The Walrus. Here are some excerpts.

The system launched in Oakville in July 2008; Edmonton and Ottawa go live this fall; locations across Ontario and British Columbia, and as far away as California and Australia, will follow.

Instead of relying on librarians to act as gatekeepers for collections that are increasingly virtual and global in scope, BiblioCommons harnesses crowds of users to guide one another -- a familiar Internet tactic.

Getting people to participate in BiblioCommons will require seamless in-tegration with existing library catalogues, so that input from other users -- and the opportunity to contribute -- arise naturally from the experiences of browsing the catalogue and renewing books. The greater technical challenge lies in integrating the patchwork of different systems across the continent. "Librarians care a lot about the details of their catalogues," Jefferson says diplomatically. But libraries have one key advantage over commercial sites: "People really want to support their libraries," she says. "There's a goodwill that Amazon or Chapters simply don't inspire."

I enjoyed looking around this and other Bibliocommons sites. It will be interesting to see how Bibliocommons develops, and whether it achieves the scale of participation to succeed more widely. It is not about 'user contributed' content, but what my colleague Stu Weibel calls 'social bibliography', or maybe 'community bibliography' (OK, let's be generous abut the term 'bibliography ;-). Community, as I remarked the other day, is the new content.

Related entries:

On the discriminations of availability ...

dempsey — Sun, 01 Nov 2009 21:43:55 -0800

Seamus Heaney famously - and in poetry - complained about being included in an anthology of 'British' poetry. In the course of his poem he invokes Miroslav Holub's 'On the necessity of truth' where a man creates a disturbance in a cinema when he sees a beaver mistakenly called a muskrat on the screen. The man wants to set the record straight.

I don't have a copy of Heaney's work as I write this, but I can point to a discussion of the passage in Acting between the lines : the Field Day Theatre Company and Irish cultural politics, 1980-1984 courtesy of Google Books.

(And it would be nice to be able to easily reference the appropriate parts of both Heaney's and Holub's works on the web. I find that I am increasingly expecting to be able to find book text online - when looking for a quote, when helping with homework, etc.)

Now, a few weeks ago, Sergey Brin wrote an op-ed piece about Google Books in the New York Times.

He discusses the fate of books still potentially in copyright:

But the vast majority of books ever written are not accessible to anyone except the most tenacious researchers at premier academic libraries. Books written after 1923 quickly disappear into a literary black hole. With rare exceptions, one can buy them only for the small number of years they are in print. After that, they are found only in a vanishing number of libraries and used book stores. As the years pass, contracts get lost and forgotten, authors and publishers disappear, the rights holders become impossible to track down.

Inevitably, the few remaining copies of the books are left to deteriorate slowly or are lost to fires, floods and other disasters. While I was at Stanford in 1998, floods damaged or destroyed tens of thousands of books. Unfortunately, such events are not uncommon -- a similar flood happened at Stanford just 20 years prior. You could read about it in The Stanford-Lockheed Meyer Library Flood Report, published in 1980, but this book itself is no longer available.

[A library to last forever]

It was soon pointed out in the library community that The Stanford-Lockheed Meyer Library flood report, was still 'available' in as much as it was held by several libraries. Worldcat.org showed four libraries holding it, and there are probably more. There was some discussion on the Web4lib mailing list along these lines for example.

At the same time, on a closed mailing list in which I participate, one commenter argued that this level of availability meant that the volume was actually not available in any 'practical sense'.

As it is available in the 'library system', the report is available to library users across the country. However, they have to be affiliated with a library which offers an Inter Library Lending service, they have to know about it, they have to submit a request, and they have to wait for it to arrive.

Certainly, if the report were available through Google Books (or some other network level repository of digital books), its availability would be greatly amplified.

It is clear that there are grades of availability. As some level the transaction costs - or, as importantly, the price - of acquiring something may be considered too high for it to be considered available in a 'practical sense'. But your mileage may vary.

In this case, the fact that it is in the 'library system' means that it is potentially 'available' to library users anywhere through the inter library lending arrangements in which most North American libraries participate. The book is available in a very real way for somebody who wants to see it with a little persistence. And through the public availability of Worldcat and other resources, and the greater prominence and ease of use of end-user requesting, the transaction costs have gone down. And there is a link to Worldcat from the Google Books record for the report.

However, it would seem that the transaction costs are still too high for many. Libraries do not yet appear as a 'system' on the web, in the sense of being able to support well-seamed easy to use discovery, request and delivery across the system. And, of course, instant digital availability sets a different expectation than such a system currently provides.

That said, it seemed to me (as it did to the librarians on the web4lib discussion list) that saying that this volume was no longer available was a stronger statement than the situation warranted. I could go with 'not easily available', but 'no longer available' was too much ...

And as I sat there looking at something being called a muskrat, I wanted to say, no, it is a beaver .... ;-)

Libraries and the long tail: some thoughts about libraries in a network age

Libraries and the long tail: intro

dempsey — Sun, 01 Nov 2009 23:44:04 -0800

Discussing grades of availability in my last post, I mention an article I wrote a few years ago on libraries and the long tail. Here is how it starts:

Discussions of the long tail that I have seen or heard in the library community strike me as somewhat partial. Much of that discussion is about how libraries contain deep and rich collections, and about how their system-wide aggregation represents a very long tail of scholarly and cultural materials (a system may be at the level of a consortium, or a state, or a country). However, I am not sure that we have absorbed the real relevance of the long tail argument, which is about how well supply and demand are matched in a network environment. It is not enough for materials to be present within the system: they have to be readily accessible ('every reader his or her book', in Ranganathan's terms), potentially interested readers have to be aware of them ('every book its reader'), and the system for matching supply and demand has to be efficient ('save the time of the user'). [Libraries and the long tail: some thoughts about libraries in a network age]

Incidentally, I think Ranganathan's 5 laws [Wikipedia entry] remain relevent in lots of ways to current discussions, as above.

QOTD: protocol-based time travel for the web

dempsey — Sat, 07 Nov 2009 19:40:55 -0800

We are pleased that Herbert Van de Sompel will be talking about Memento, a joint project of Los Alamos National Laboratory and Old Dominion University, at OCLC later this month. We will make a webcast available; see the details here. If you are in Central Ohio, come by ....

Here is a recent paper describing the work:

The Web is ephemeral. Many resources have representa- tions that change over time, and many of those represen- tations are lost forever. A lucky few manage to reappear as archived resources that carry their own URIs. For ex- ample, some content management systems maintain version pages that reflect a frozen prior state of their changing re- sources. Archives recurrently crawl the web to obtain the actual representation of resources, and subsequently make those available via special-purpose archived resources. In both cases, the archival copies have URIs that are protocol- wise disconnected from the URI of the resource of which they represent a prior state. Indeed, the lack of temporal capabilities in the most common Web protocol, HTTP, pre- vents getting to an archived resource on the basis of the URI of its original. This turns accessing archived resources into a signicant discovery challenge for both human and software agents, which typically involves following a mul- titude of links from the original to the archival resource, or of searching archives for the original URI. This paper proposes the protocol-based Memento solution to address this problem, and describes a proof-of-concept experiment that includes major servers of archival content, including Wikipedia and the Internet Archive. The Memento solution is based on existing HTTP capabilities applied in a novel way to add the temporal dimension. The result is a frame- work in which archived resources can seamlessly be reached via the URI of their original: protocol-based time travel for the Web. [Memento]

Reputation enhancement redux

dempsey — Thu, 12 Nov 2009 09:34:58 -0800

I wrote recently about the growing interesting in reputation management on the web.

Reputation management on the web - individual and institutional - has become a more conscious activity for many, as ranking, assessment and other reputational measures are increasingly influenced by network visibility. In particular, it raises for academic institutions an issue that has become a part of many service decisions: what is it appropriate to do locally? What should be sourced externally? And what should be left to others to do? [Reputation enhancement]

This is a wide-ranging issue, pulling together in various ways overlapping issues such as individual and institutional disclosure of research and other outputs; emerging academic social networking practices; formal expertise and research output management; search engine optimization strategies; practices for improving citation, ranking and reputation measures; social reference/bibliography; and so on. I think that we will see some of this activity become more routine in organizational and operational terms over the next few years.

In this context, I was interested to see a presentation on research support by Rachel Cowan and Alex Hardman from the University of Manchester. They focus on reputation and network identity as important parts of overall research management.

Technology Enabled Research

View more presentations from Alex Hardman.

The presentation has three strands: developing reputation through a digital identity, keeping on top of the literature, and extending research connections. Of these, the first and third relate broadly to reputation enhancement or management in a web environment.

They ask the audience if a personal Google search does a good job of showcasing their identity and research. (This reminds me of Tony Hirst's comment that our 'home page' in now the first page of Google results.) Then they talk through some of the ways in which people develop digital identities (blogs, twitter, ...). They also review some social networking and other tools of interest in an academic context.

Here is their overview of activities mapped onto services (click to see in situ with ability to enlarge):

Libraries and e-science

dempsey — Fri, 13 Nov 2009 13:51:04 -0800

Emerging data-intensive e-science presents many support challenges for institutions, disciplines and national bodies to work through. The role of the academic library in this multiscale world is also an open question. Two recent reports discuss e-science (or 'cyberinfrastructure' or 'e-research') in general terms and repay reading.

Liz Lyon, the Director of UKOLN, and also a principal in the Digital Curation Centre, has focused on this area for several years now and has produced an interesting synthesising report for the JISC: Open science at web-scale: optimising participation and predictive potential: consultative report [Summary; Full report PDF]. An important theme of the report is 'data informatics', defined in this way: "library and information science methodologies which have been applied to research data".

The report is organized around six 'consultation challenges'. The first is 'scale, complexity and predictive potential'. Here is the summary:

Data-intensive science powered by contemporary computational hardware, software and research techniques, enables scientists to perform experiments and calculations at different orders of magnitude of scale and volume: research that was completed in a year can now be repeated in a weekend. Sustained growth in data modelling, complex simulations and visualisations, facilitate interpretation and analysis by humans and machines, leading to the development of predictive science scenarios in a wider range of disciplines. Examples of data intensive science at these extremes of scale, which enable forecasting and predictive assertions, have been described.

Assessments of the accuracy and robustness of predictions are linked to uncertainty quantification, the accuracy of the underlying model, and the integrity of the data. Key questions address community awareness and understanding of the potential implications and impact of (open) data-intensive science at new extremes of scale and complexity, and the service requirements for associated data curation and preservation. [Open science at web-scale: Optimising participation and predictive potential - summary]

To give some flavor of concerns, here are the other challenges: Continuum of openness; citizen science; credentials, incentives and rewards; institutional readiness and response; data informatics capacity and capability. A brief chapter is devoted to each.

The author is positive about the role of libraries and librarians, particularly in the data informatics section. That said, given the absence of routine service and organizational responses the library role is still expressed in very general terms. What it might mean in practice is naturally less well developed.

The other publication is a collection of essays assembled in honor of Jim Gray:

In The Fourth Paradigm: Data-Intensive Scientific Discovery, the collection of essays expands on the vision of pioneering computer scientist Jim Gray for a new, fourth paradigm of discovery based on data-intensive science and offers insights into how it can be fully realized. [The fourth paradigm]

For Gray the first three paradigms are experimental, theoretical, and computational.

We said, "Look, computational science is a third leg." Originally, there was just experimental science, and then there was theoretical science, with Kepler's Laws, Newton's Laws of Motion, Maxwell's equations, and so on. Then, for many problems, the theoretical models grew too complicated to solve analytically, and people had to start simulating. These simulations have carried us through much of the last half of the last millennium. At this point, these simulations are generating a whole lot of data, along with a huge increase in data from the experimental sciences. People now do not actually look through telescopes. Instead, they are "looking" through large-scale, complex instruments which relay data to datacenters, and only then do they look at the information on their computers.

The world of science has changed, and there is no question about this. The new model is for the data to be captured by instruments or generated by simulations before being processed by software and for the resulting information or knowledge to be stored in computers. Scientists only get to look at their data fairly late in this pipeline. The techniques and technologies for such data-intensive science are so different that it is worth distinguishing data-intensive science from computational science as a new, fourth paradigm for scientific exploration [1]. [Jim Gray on escience - PDF.]

The collection of essays is divided into these sections: Earth and environment; Health and wellbeing; Scientific infrastructure; Scholarly communications. And there are opening and concluding sections. The contributions are readable and in the form of short essays rather than research papers. There is a contribution by Cliff Lynch on the changing scholarly record, by Timo Hannay on the impact of the network on the structure of science, and by Herbert Van de Sompel and Carl Lagoze on the enhancement of the scholarly record with actionable structure.

There is no specific contribution on libraries, and it is interesting to note that the directions of much of the occasional mention of libraries is towards network level digital libraries.

It is important for libraries to understand these changes. The reshaping impact of the network on learning and research behaviors is a more important factor for libraries than the direct impact of the network on library processes themselves.

Social tools and science

dempsey — Sun, 15 Nov 2009 18:57:48 -0800

In her report on Open science at webscale, I was interested to see Liz Lyon give the following list of tools used to share their work by researchers.

Currently, researchers are using open science tools such as:

Connotea for reference management

Mendeley (which applies LastFM principles associated with music selections to journal

articles)

Friendfeed (for threaded discussion and aggregation)

Scivee and YouTube (for sharing experimental methodologies and protocols)

SciLink and Nature Networks (for social networking)

myExperiment (for sharing workflows)

eyeLIMS (an open source Laboratory Information Management System)

LabLit.com (about science/laboratory culture in the literature and media)

ConceptWeb (from WikiProfessional and includes WikiPeople and WikiProteins)
[Open science at webscale - PDF]