Semantic Library

If you are finding your way to the blog via the Talking with Talis podcast, hello! I wanted to expand on why I am interested in the Semantic Web as I only briefly touched on this in the podcast.

eResearch and Data

A couple of years ago, I attended a conference where the theme was eResearch. Librarians described how they have responded to the challenges of managing datasets, ever-increasing amounts of raw information and data, as well as grey literature, preprints, and other publications. Several scientists also gave perspectives on how they thought libraries could assist with their research. The scientists discussed the issues with being able to collect so much data, increased complexity in manipulating it, and how so much of their work has shifted online and in some fields, to Open Access. This led me to think about how librarians can work with researchers to assist them better, beyond what we do now. How can we assist with the way data is structured and shared, and perhaps even become part of research teams, assisting with the gigabytes upon gigabytes of data that teams create, use and share.

At the same time, there was a growing focus on research metrics, quality and impact in several countries. We know how limited ISI is, so what else can we build to do this better? How can we trace data through a published presentation back to where it was created? How can we connect ideas, people and projects online to find collaborators and like-minds in a field?

The conference didn’t mention the Semantic Web as a way to assist with these issues, but to me, as I read more about the concepts behind it, it seems a logical fit for Open Data, Open Access and issues of managing gigabytes of data.

Looking to the future

Other ideas, like Next Generation Catalogues are also really interesting. What is important is a focus on the structure and quality of the data we have in the catalog. There is no point to bells and whistles presentation like tag clouds and facets without having good data to work with. There’s a growing number of librarians who are focusing on this and taking a strong interest in RDA and other projects.

Beyond these issues, I’m interested in what’s next. I think libraries are a natural fit for the Semantic Web because of its emphasis on RDF, and data and metadata. In some ways it’s a return to what we do best - organising information, provenance, databases.

I do think that it should not be just from technical staff and cataloguers and research librarians - there really does need to be involvement from all types of librarians to ensure that we are really participating in projects that meet user needs, not just in libraries but on the Internet as a whole.

Read/Write Web has a great post on the Semantic Web which gives a brief primer on theory, technology, and applications. “Semantic Web Patterns: A Guide to Semantic Technologies” is a great introduction to what it’s all about with plenty of links to explore. The article also discusses the different meanings of the Semantic Web and how different technologies support different interests -

The Semantic Web means many things to different people, because there are a lot of pieces to it. To some, the Semantic Web is the web of data, where information is represented in RDF and OWL. Some people replace RDF with Microformats. Others think that the Semantic Web is about web services, while for many it is about artificial intelligence - computer programs solving complex optimization problems that are out of our reach. And business people always redefine the problem in terms of end user value, saying that whatever it is, it needs to have simple and tangible applications for consumers and enterprises.

A must read, but do also read the critique of this article over at Nodalities which debates some of Alex Iskold’s claims. I find this comment about Semantic search particularly interesting,

Success for the Semantic Web wouldn’t be improving search, but marginalizing it.

Read and enjoy them both! What both of these articles make clear is that what the Semantic Web “is” is far from clear, and given how much hyperbole there is around what is “2.0″ I wonder how far we are from that.

The BBC Internet Blog has a post about the accessibility of their new homepage, and the challenges in designing it. One of the things they did was use Semantic Web markup -

Using the results of our user-testing we made semantic mark-up of the page a priority, ensuring that each panel of the homepage design could be treated and navigated consistently. For example, the H3 heading always applies to the title of the content panel whether the content is in a table, a list or paragraph, and a logical structure of headings follow underneath.

A great example of a practical use for the Semantic Web. Presumably, with the consistency on the homepage that also lends the site to multichannel and transform content to different devices more easily.

Thanks to Rob Styles, I now have a Twine account and have been finding my way around this new Semantic-powered world.

There has been a great deal of discussion about Twine, from the supportive to the disappointed there’s a wide range of opinions about it. Bearing in mind all of this, I am trying it out with an open mind and a commitment to pass comments and feedback to the developers.

As I explore Twine I’m keeping in mind the things that interest me about the Semantic Web: the potential for trust, relationships and more targeted results and information. So far, I think it has a lot of potential.

Starting out: Social networking upside down

If you use many social software sites, like LinkedIn, Facebook and Twitter, you’ll be used to building a network of contacts based on people you know. Twine, at least for me, is the opposite approach. You begin with your interests and connect with people based on shared interests or Twine will recommend people to you. Starting with interests is perhaps a bit more like 43 Things. Note the wording - Connect not Friend. This makes me more likely to reach out and connect with people I don’t know, which is important especially on a new service with a (as yet) limited user base.

Recommendations

I’m not sure how it’s being powered, but once you have started to make connections and groups (called Twines) recommendations will appear for other people and Twines to connect with. It is too soon to tell how powerful this might become, but if it is really focused on it could become very valuable. For example, I’d like to find people who work in NGOs in Europe who are interested in web development. While a keyword search could pull up such a result, a Semantic approach might be able to give me a more accurate recommendation.

Starting a Twine

I started a twine on travel to try out linking and connecting interests. You can then invite others to join, add notes and post items to the Twine. This is a feature that is highly promoted. The idea is that you will be able to use Semantic-powered search to locate your item again and find other recommended items.

If you know del.icio.us, Connotea, CiteULike and another number of bookmarking services then you know how this works. Add a bookmarklet to your browser, find a page you like, and add it to your Twine, send it to a friend or save it for yourself. The popup window does need a bit of finetuning at this stage, as once you have more than a few connections or Twines it can be hard to see what you are adding an item to.

Interest Feed

All the actions you take in a Twine, or actions of others are added to your Interest Feed, much like the news feed in Facebook. You can control somewhat which notifications appear there, which is good because it quickly overflows with responses to discussions, new posted items etc. This can also be emailed to you daily but no RSS yet.

Definitely Beta

It’s clear that this really is a beta. There’s navigation inconsistency, search bugs, and incomplete features. But I am sending feedback through the feedback and beta Twines as I see things. It’s the first time I’ve really felt that I’ve had the opportunity to do that, and to do so easily. Most other ‘beta’ sites are of the perpetual beta sort and do not make giving feedback easy.

Do I see Twine taking over from my del.icio.us account? For the moment, I’m going to use both and evaluate them soon. For me, apart from the social bookmarking angle of Twine, what I am most interested in is recommendations and I am interested to see whether del.icio.us’ tag based approach to discovery or Twine’s Semantic push approach works better for me.

Twine has a lot of interesting features, and I like that they are really taking on board feedback. I received good comments about my first bit of feedback yesterday (request for OpenID in profiles and deduplication of posted items in Twines) and this does encourage me to send in more.  I’ll give a more detailed review soon!

Dean Guistini at UBC Academic Search explains why libraries have a role in the Semantic Web:

Without over-simplifying, this is what the semantic web is about - making connections between documents, their ideas (and intellectual content) and the people we serve in libraries. If the web is going to be the platform for ideas and lifelong learning well into the 21st century - perhaps something else will replace it? - then librarians need to take a good look at the emerging semantic web.

Take a look at the whole post - it’s a great argument about the expertise libraries bring to the Semantic Web. An article co-authored by Giustini and Allan Cho on web 3.0/Semantic Web is due very soon.