[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

RE: Issues with XML and Semantic Web ?

• To: 'Chiusano Joseph' <chiusano_joseph@b...>, Jeff Rafter <lists@j...>, xml-dev@l...
• Subject: RE: Issues with XML and Semantic Web ?
• From: "Bullard, Claude L (Len)" <len.bullard@i...>
• Date: Wed, 10 Nov 2004 08:02:32 -0600

There are problems with that scenario.  

Cross-jurisdictional queries are often not 
enabled by treaty and law.  NCIC queries that reach into 
Canada, for example, can return hits but they cannot 
be acted on in most cases.   So again, there are names 
and then there are semantics and operations.   It is 
the context (the event types) that often determine the 
allowable operations.

It is necessary to be very clear about this.  Otherwise, 
consultancy leads to RFPs to which vendors cannot respond 
and agencies cannot procure.  The so called 'federated 
query' is a source of many issues some of which the 
Markle Foundation pointed out.   Even within the CONUS 
jurisdictions, the problems of crossing private/public 
boundaries with querying is always to be suspect and 
held to much higher auditing standards if the querying 
agency is a law enforcement agency.  Otherwise, it is 
impossible to protect civil liberties.

So it becomes a matter of associating the semantics 
of the ontology to the business rules based on event 
types (an ontology).  This is one approach to the 
problem of acquiring identity without opt-in as is 
the case for sensor webs capable of using biometric 
properties to acquire identity and issue queries.

len

From: Chiusano Joseph [mailto:chiusano_joseph@b...]
 
In addition to identifying opportunities for information sharing, these
artifacts can also identify opportunities for federated queries (perhaps
using Enterprise Information Integration - EII). Consider a hypothetical
situation in which 2 agencies have arrest information for an individual
- but one has it on a domestic basis, and another on an international
basis. A federated query between these two data sources - which can be
determined by comparing their ontologies and taxonomies - can yield an
arrest record for a given individual on an international basis.

I have found that in educating unfamiliar folks on these artifacts, it
works best to use examples within their own domain. Familiary with their
own data and concepts greatly eases the mental transition.

• Prev by Date: Re: Issues with XML and Semantic Web ?
• Next by Date: RDF necessary to SW? - was Issues with XML and Semantic Web ?
• Previous by thread: RE: Issues with XML and Semantic Web ?
• Next by thread: Announcement: xlinkit Rule Workbench release
• Index(es):
  • Date
  • Thread