Charleston Conference – Semantic Technology

Darrell Gunter, CEO of Gunter Media Group, Adam Marshall of Portland Press and Thane Kerner of Silver Chair presented on semantic technology and getting up to speed to better serve your user community.

Each of the speakers posed 5 Questions with 5 Answers

What prompted you to engage semantic technology into your products/applications?

  • Darrell Gunter (for Elsevier/Collexis) Needed to develop an expert profile database where the experts can connect with one another
  • Adam Marshall: So much data that we no longer know what we know and finding what we find is very difficult.  Most of the articles we keep are PDF, which are flat files with no interactivity, and they wanted to develop a new tool to provide interactivity b/t PDFs. Came up with Utopia Documents, which blends the best of the semantic web in PDFs.  It connects documents with online data, linking the flat PDF to online databases.
  • Thane Kerner: they have large volumes of very specialized content and needed better ways to search and connect content that wasn’t available in current search technology

what are some key advantages and how will it help to enhance your user’s experience?

  • Darrell Gunter:  take structured/unstructured data and using a knowedge base, create a digital fingerprint of dey relevant topics; ability to aggregate and visualize the data, displaying this information/connection to end users
  • Adam Marshall:  brings static PDFs to life, by seamlessly connecting the content of the PDF to a variety of online databases or other documents on similar/related topics
  • Thane Kerner:  efficiency, accuracy, anticipation of an information seeking path; able to retrieve based on passive/former user inputs and behaviors; integration of data and applications across organizational and platform boundaries

what were some of the problems solved or new enhancements created by the implementation of semantic technology?

  • Darrell Gunter:  helped connect the information better
  • Adam Marshall:  brought a static, flat document to new life by interconnecting it with other data.  Tables within the PDF document could then “come alive” in spreadsheets; the PDF is dead, semantic technology brings the PDF back to life
  • Thane Kerner:  created a product development infrastructure that enables a whole new kind of process; instead of big builds we needed rapid response; respond to changes in the user experience

what are some of the key lessons learned in getting started?

  • Darrell Gunter:  keep it practical, solve a problem (connecting researchers in/out network, etc.
  • Adam Marshall: don’t underestimate the type of problem that can come up when you implement new software; learn the language of software geeks; don’t assume researchers know what semantic technology is
  • Thane Kerner not a bolt-on, it is a foundation (ie not a new set of tires for an old car); hard sell; developing the infrastrucure is essential; taxonomies are or should be domain (try to find some that exist rather than create your own), not enterprise, at their core; work from the bottom up (content/application/product) and top down (domain knowledge map) simultaneously

what are the key next steps in further developing your semantic applications?

  • Darrell Gunter:  allowing other developers to use apps to develop the content further; trend analysis – show what concepts are trending up/down and where concepts have not co-occurred yet, they may have the opportunity to do so in the future
  • Adam Marshall:  PDF reader with enhanced linking out to a variety of STEMM databases; commenting/social networking either private or public; Utopia library – organize your personal collections of PDFs (like iTunes), sort and catalog
  • Thane Kerner:  move to more automated and scalable application of metadata; continu to develop ontological expressions (its like a sentence, noun/verb to put together); how can we align domain semantics with other cross-industry standards; moving towards a world of interoperability and this is the technology that’s going to help us get there

Why is this important for the library community?  This is a technology concept that will be present in all databases/products in the future

ROI seems to go beyond any one individual publishers ability, things need to be combined.  Can this be implemented within discovery tools?  Report, project that everyone should read (okay, read the executive summary, it’s free, the report is $3500).  We need everyone to participate to maximize the value, but you can achieve significant value w/in your own organization by using these technologies

Advice to a small publisher:  find some form of taxonomy that exists in your space, begin an experiment to develop a product and don’t expect to make money to start

Example of this in humanities/social sciences?

  • Alexander Street Press is doing some of this with their products
  • The state of North Carolina will soon develop a database of all faculty in the state, this will include social science/humanities folks
  • Institute for the advancement of new technologies in the humanities  at the Univ of VA

darrellwgunter@slideshare  for the materials