Discover
Home › Discover › Semantic Web › Semantic Web Overview

Semantic Web Overview

iPlant's Semantic Web Program delivers a semantic infrastructure to allow people to describe data and services (web sites), discover those services, engage them, and handle the results -- all based on the semantics (or meaning) of the data and/or services. iPlant Semantic Web technologies do this in a manner that allows computers to process this information in a high-throughput manner, for example, to aid you in finding new services based on what they do or the data they handle. We will be releasing these technologies throughout 2011 to build a semantically-aware capability into iPlant's cyberinfrastructure.

The iPlant Semantic Web Architecture uses the web itself as the palette for data and service integration. It is designed so that users get the value of a semantically-aware infrastructure, while never needing to know the underlying details. The technology is designed to allow iPlant or anyone to contribute and gain from semantic enhancement by allowing any web site to semantically enhance its offerings.

The platform allows any web site to:

  1. Describe its data and services using a standard, universal syntax and rich, computable semantic
  2. Be discovered by others seeking a type of service, type of input data, or type of output data
  3. Be invoked by others using a standard, universal syntax and shared, public semantic
  4. Return a response in a standard, universal syntax and shared, public semantic

This means that clients -- people like you and me using computer programs -- can:

  1. Discover services based on the type of service sought, or the type of data accepted, or the type of data returned
  2. Invoke discovered services using a standard, universal syntax and shared, public semantic
  3. Parse a service response using a standard, universal syntax and shared, public semantic

We achieve this by using a thin, lightweight protocol on top of industry-wide semantic web standards.  The protocol is SSWAP (Simple Semantic Web Architecture and Protocol; pronounced "swap") and the standards are the W3C standards of RDF, RDFS, XSD, OWL, and SPARQL. These standards are implemented in Java and JSON.

The good news is that much of this technology stack is hidden; we expose only the little bits that developers need to know to deliver
integrated, semantic value to the end user. This is done by providing a:

  • HTTP API (Application Programming Interface) for developers (no computer programming necessary)
  • Java API (Application Programming Interface) for developers (optimized for handling the minutiae, while allowing developers to concentrate on the job-at-hand)
  • SDK (Software Development Kit) for data and service providers to wrap their service offerings
  • Web site for end users to enable semantic searching.  We are currently using the research site http://sswap.info as we develop production support

iPlant Semantic Web Services are different from other platforms by embedding on-demand, transaction-time automated semantic reasoning in service negotiation. This means that there is no need for parties to pre-agree on domain vocabularies or be constrained by static data models. The iPlant Semantic Web Services platform employs reasoners to determine, at transaction time, if data or services embed operational meaning such as concept or property equivalency or concept or property subsumption. The platform handles the necessary conversions transparent to both the developer and end user. The only operation needed to be specialized by individual web sites is their own mapping of their idiosyncratic schemas into and out of a shared, public semantic.

Where to go from here: