Tuesday, March 23, 2010

ASIDIC Spring 2010: Smart Content Pulls Into View for Enterprise Publishers

It's always been fun to be a part of ASIDIC events, so I was very pleased to have been invited to moderate a Q&A period at this year's Spring ASIDIC get-together at the offices of Lyrasis in Philadelphia. It's a bit more low-key venue than for previous ASIDIC events, which reflects in some ways the challenges that many enterprise-oriented publishers have faced these days, but also the degree to which their business models are trying to catch up to the value points in publishing that revolve around metadata and search technologies. The good news is that the ASIDIC meeting pulled together some excellent case studies demonstrating how publishers are moving away from "pull up a document" styles of electronic publishing towards using sophisticated semantic processing to get their content ready for battle for use in contexts driven by metadata. Here are some links to the panel-by-panel posts that I recorded on Google Buzz (no login required to view, login required for comments):
  • IDC's Sue Feldman on the New Search Architecture
    Sue was in good form, I really enjoyed her insights. Key stats from IDC's 2010 enterprise user survey: 21 percent use colleagues as their first stop for information, 61 percent go to the Web first, only 1.8 percent to their subscription database services. My take: if you're not using the open Web and social media as marketing channels, you're missing more than 80 percent of your opportunities to be relevant in the "go-to" source for people who need your enterprise content.
  • Thane Kerner, Silverchair - A Primer on Semantic Technologies
    A good overview of today's semantic technologies and terminology. One of the nice things about this ASIDIC meeting is that it got pretty deep into the implementation of semantic technologies without lapsing into endless "geek speak."
  • Case Studies - IEEE and SciTech Strategies, Inc.
    This was a very interesting study of how the IEEE used domain mapping as a tool to reveal expertise appearing at the intersection of subject domains not usually associated with one another. By using taxonomies and domain mapping they revealed opportunities at the intersection of information technologies and medical science - the type of opportunities that innovation professionals are focusing on to build out new markets for products and services.
  • Case Studies - Enhancing the user's experience with semantic "smart linking."
    McGraw-Hill highlighted work that they are doing using metadata and XML-formatted content to build out new editorial content for their premium Aviation Week and Platt's enterprise services rapidly. These technologies are enabling them to generate "topic pages" rapidly that can be destinations for links embedded in their news coverage and archives. Metadata can also enable opportunities at the intersections of their publishing properties - for example, it would be interesting to see how information on commodities such as jet fuel prices from Platt's could be made useful in Aviation Week content.
  • Case Studies - Collexis and the American Association for Cancer Research
    This was an excellent example of how deep taxonomies and semantic technologies solved a very crucial problem for a scholarly publisher. Collexis enabled AACR to identify a much broader range of topic experts to be available for peer-reviewing scientific research articles and to filter out people who may have a conflict of interest. At a time when scholarly publishers are trying to position their assets more effectively against Open Access competitors, being able to demonstrate superior methodology for peer review via advanced technologies is a great idea.
  • Case Studies - Getting references right - how semantic technology helps linking, findability and analysis
    Interesting example of how the American Psychological Association went from a "square zero" in Smart Content to state-of-the art infrastructure to help it begin to build rich and powerful search experiences on Mark Logic's XML server. One of the real stories about semantic technologies today is that although it's not effortless to make the transition to Smart Content, today's technologies can enable publishers to make that transition much more rapidly and cost-effectively. Harder, though, is getting business models up to speed.
  • Closing keynote - Steve Sieck, SKS Advisors
    Steve always has powerful and thoughtful insights delivered with a good dose of understatement, a combination that makes him well worth listening to at events. Steve did a good job highlighting some of the key "what's next" themes for semantic content, including social media integration, "linked data" - enabling data to "talk to other data" on the Web in ways that enable semantic APIs - and the extension of semantics into marketing and branding.
All that and much more made the trip down to Philadelphia for the day well worth it. As I was discussing with an attendee afterwards, this is still the early days of semantic implementation for many publishers, with many high-value products and services only beginning to emerge for enterprise use. For example, what happens when you start applying semantics to newly released scientific research that puts previous research about a company's drugs or medical technologies in a negative light? All of a sudden technologies that were intended primarily as search interface tools then become powerful technologies for building real-time news and intelligence that can move securities markets rapidly. We're in the early days for these technologies, indeed, offering publishers opportunities to "leapfrog" their way into new value propositions.

Yet looming above all of these opportunities is the Web itself, that vast collection of human insight that most people still use as their primary reference so often. Precious little was said at ASIDIC about how to use Smart Content to built more Web-aware content. There was also an interesting interchange that I had at the end of the meeting with a long-time indexing expert who mused about how in many ways the metadata that was adding the most value in many of the examples discussed at the event were not necessarily those tried-and-true indexing tools that have been used for years. Yes, the truth about metadata is that much of what has been considered useful "information about information" is just the starting point for adding value to content today.

Here, also, the Web points the way. While Google is not thought of as a service that uses semantic tools in its presentation of content, in fact its content is rich with semantic inferences from Web page links, analysis of use statistics, evaluation of geo-tagged data and other content to derive useful information and experiences. These happen mostly "behind the scenes" in Google services, but they are there nevertheless, aiming towards the very "accuracy" that was discussed at the day's sessions. Ultimately Smart Content is the content that transforms what was previously thought of as just a publication or a search result into the input for sophisticated content-serving applications, whether they are presented as a publication or a problem-solving tool or a workflow service.

Thanks again to the ASIDIC team that put together a very interesting event with great attendees. Hopefully better times will enrich us with more events like this.
Post a Comment