Aggregation tools

A brief overview of the ILRT MedCERTAIN work on aggregation tools, software and applications.

Our MedCERTAIN developments are divided into two broad categories: metadata production and metdata consumption. The production tools focus on creation of rating and evaluation metadata; the consumer tools focus on the interchange, querying, storage and use of this data by a diverse family of applications.

Principles

Work so far

We can show...

A worked example: endorsements

An RDF Vocabulary for endorsements. Some sample data (which could be derrived from a MedCERTAIN database). [med2.xml]

this uses two attributes: med:assuredby med:ratedbadby

Aggregation: Querying the data

Using generic technology for cross-domain metadata query. Compare a medcertain example with a non-medical RDF query example query (from an RSS portal application):

"find pages that advertise jobs, and the salary of the job, where the salary is greater than 55000". 


eg.1:
	SELECT ?z, ?a 
	FROM http://ilrt.org/discovery/2000/11/rss-query/jobs-rss.rdf, http://ilrt.org/discovery/2000/11/rss-query/jobs.rss 
	WHERE 
	(job::advertises ?x ?y) 
	(job::salary ?y ?z) 
	(job::title ?y ?a) 
	AND ?z > 55000 
	USING job for http://ilrt.org/discovery/2000/11/rss-query/jobvocab.rdf#

Using MedCERTAIN metadata with RDF Query

A simple medically themed RDF query example (same software; different query and data...):

"Find pages (and their Dublin Core titles)and services (plus Dublin Core titles) where the service endorses the page." 

eg.2:
	SELECT ?page, ?service, ?title, ?who
	FROM http://ratings.example.com/some-data.rdf
	WHERE 
	(medx::endorses ?service ?page) 
	(dc::title ?page ?title) 
	(dc::title ?service ?who) 
	USING medx for http://medcertain.org/vocab/medx#
	dc for http://purl.org/dc/elements/1.1/

Practical MedCERTAIN RDF Query: retrieving info about a site

We can use RDF queries against RDF data loaded from the Web or other sources (optionally checking digital signature information). The query language allows for matching against data structures defined as rdf vocabularies, and is designed to look like a simplified version of SQL. It supports functionality comparable to the PICS-Rules specification and hence can serve to characterise diverse policies defined over MedCERTAIN data.


eg.3:
       "find the content policy, privacy policy, and chief quality officer
	for providers that are HealthPortals; and get the Dublin Core description
	of the content policy."

	SELECT ?provider, ?contentpolicy, ?privacypolicy, ?desc, ?cqoname
	FROM http://ratings.example.com/some-data.rdf

	WHERE 
	(hidl::content_policy ?provider ?contentpolicy) 
	(dc::description ?contentpolicy ?desc)
	(hidl::privacy_policy ?provider ?privacypolicy)
	(hidl::privacy_policy ?provider ?cqoname)
	(rdf:type ?provider hic::HealthPortal)

	USING 
	      hidl for http://medcertain.org/vocab/hidl#
	      hic for http://medcertain.org/vocab/hic#
	      rdf for http://www.w3.org/1999/02/22-rdf-syntax-ns#
              dc for http://purl.org/dc/elements/1.1/
	

results:
[ provider | contentpolicy | privacypolicy | desc | cqoname ]

[ healthco.com    bar.html        priv1.html      "...."  john smith ]
[ myhealth.com    content.html        policy.html      "...."  sam johns ]
[ etc... ]

What this shows: