The slides are up from the Infonortics Search Engine Meeting that was just held in Boston (4/28-4/29). Here are a number of presentations that caught my eye. Not all are related to federated search but breakthroughs in search should be of interest to all of us.
- Federated Search: True Enterprise Search. Abe Lederman, president and founder of Deep Web Technologies, and sponsor of this blog, shows how federated search overcomes limitations of enterprise search.
- Search as a Mode of Learning: Requirements for Next Generation Search Systems. While the presentation doesn’t specifically mention federated search, the vision should speak to all of us in the industry as it’s about finding one’s way through information to improve performance . Slide 6 takes a position on the future of search:
Search needs to be a core capability of applications that support learning and performance.
Search must integrate with content (searches become a type of content) and must support:
- Exploration and patten building
- Contextual delivery of information needed for performance
- Social (shared) searches
- The Next Big Thing in Search: Hybrid/Vertical Search. There are currently no slides available for this presentation. Abe says it was interesting, I’ve contacted the company presenting the talk, Exalead, and if I’m able to get slides I’ll post a link to them.
- Semantic Retrieval: Making the Computer do the Heavy lifting. This is of interest because federated search applications commonly find themselves performing search of scientific and technical documents, which especially benefit from ways of getting the most out of underspecified searches.
- Combining Semantics and Keyword Approaches to Enable Flexible Enterprise Search. While this presentation is targeted toward enterprise search, who wouldn’t be interested in the 98% precision and 99% recall claimed by the authors in a semantic capture of legacy documents?
- Search on Structured Data: The next step in the confluence of Search and BI. Microsoft subsidiary, Fast, which incorporates federated search in its technology, talks about the movement from data centric search to user centric search.
- Searching for the Good Lawyer: Emerging Best Practices In The Use of Search and Information Retrieval Methods in E-Discovery. This presentation busts some myths about search and recall. I particularly appreciated this hit of reality:
In a 40,000 document case (350,000 pages), lawyers estimated that a manual search would find 75% of relevant documents, when in fact the research showed only 20% or so had been found.
- Powering Search Results with Visualization. Federated search vendor, Groxis, discusses how search and visualization tools are converging.
Other presentations were also of interest. Check them all out.