Semantic Search Beyond RDF – SemTech 2009 Video

Wen Ruan, TextWise

Ronald M. Kaplan, Powerset division of bing
Christian F. Hempelmann, RiverGlass, Inc.
Riza Berkan, hakia

Semantic Search technology in the Semantic Web community is often understood as retrieval of knowledge from tagged data such as RDF sources, which require substantial formatting and markup to realize. Understanding unstructured query and document text and conducting searches according to their meaning is another approach, exemplified by linguistically rooted semantic matching, ontological knowledge-based semantic interpretation, and statistically based semantic similarity search.

This panel will look at different ways to tackle semantic search as a problem of text understanding. Powerset division of bing’s natural-language processing engine does deep syntactical analysis to determine the meaning of a query or a sentence. Hakia relies on a language-independent ontology model and an ontology-based English lexicon to translate text into a representation of its meaning. RiverGlass has developed an ontological semantic approach to search and text analytics, emphasizing the in-context, linguistic meaning of textual content in order to return truly relevant results in response to information requests. TextWise’s Semantic Signature matching looks for similarities between a query and text at the topic level.

SPIN: An object-oriented framework for business rules using SPARQL – SemTech 2009 Video

Holger Knublauch
Vice President

The current generation of Semantic Web languages is well suited to link data and to define domain concepts and relationships. However, real-world applications that operate on those linked data models typically need higher expressivity than what is provided by OWL and RDF Schema alone. SPIN is an open-source framework that supports the use of SPARQL to define business rules and constraint checks on Semantic Web models with object-oriented modeling techniques. This simple yet powerful mechanism makes it possible to define self-describing domain models that can then be used by generic software components such as user interface renderers, schema mappers and workflow engines. Instead of hard-coding behavior in languages like Java, SPIN makes it possible to declaratively define complex business rules and processes. SPIN can also be used to define new higher-level modeling languages with built-in semantics.

This talk:

  • Sets the stage with a quick review of SPARQL (incl. CONSTRUCT keyword)
  • Introduces SPIN as a mechanism to attach SPARQL queries to class definitions
  • Shows how to define new SPARQL functions and reusable query templates with SPIN
  • Demonstrates the use of SPIN for tasks ranging from unit conversion to computer games
  • Shows how the ideas of SPIN give rise to a new software development paradigm around self-describing linked data models

Panel: Venture Capital Outlook – SemTech 2009 Video

Steve Bastasini, Cerebra

Eghosa Omoigui, Intel Capital
Peter Rip, Crosslink Capital
Michael S. Dunn, Hearst Interactive Media
Shawn Carolan, Menlo Ventures

After a period of caution about the viability of semantic technologies, investors seem more willing to fund semantic start-ups right now. And even with the economy in distress, semantics is managing to create excitement amongst the VCs. Semantic search has been hot for a couple of years – the possibility of finding the next Google being just too enticing – but the focus seems now to be shifting to enterprise and consumer apps where as Jim Hendler famously said "a little semantics goes a long way." Money is going into enterprise software, such as business intelligence tools, and innovative consumer apps based around social networks, smarter information filtering and productivity enhancement.

So what do the VCs want to see in the business plans for semantic start-ups now? Are there still plenty of good opportunities out there for entrepreneurs or have the best ideas already claimed their share of available capital?

Panel: Publishers – SemTech 2009 Video

Greg Stuart, gregstuart.com

Keith DeWeese, Tribune Company-Tribune Interactive
Evan Sandhaus, New York Times Company
Paul Berry, HuffingtonPost.com
Jim Stanley, CBS Interactive – Technology & News
Michael S. Dunn, Hearst Interactive Media


This discussion features representatives from major media companies who are seriously investigating or presently using semantic technologies in their sites. The conversation will focus on the business and operational benefits of using semantic technologies in publishing and media sites.

Webcasts , Videos, etc