Ontologies Come of Age: The Next Generation

OCAS October 24, 2011 Bonn, Germany

Deborah L. McGuinness

Tetherless World Senior Constellation ChairProfessor of Computer and Cognitive Science

Rensselaer Polytechnic InstituteTroy, NY, USA

Background

• Ontologies Come of Age (1)– Invited presentation at the Semantics for the Web Dagstuhl

meeting in March 2000.– Followed an Ontologies AAAI panel in 1999 where the first

ontology spectrum was generated– First Ontologies Come of Age paper described all points on the

ontology expressiveness spectrum and provided examples of value at each point (in Spinning the Sem Web, 2003)

– Built on current academic experiences: large knowledge base and ontology programs such as HPKB

– Built on current consulting experiences – building ontologies and ontology groups (then private ontologies for competitive advantage; now public ontologies for interoperability)

What is an Ontology?

Catalog/ID

GeneralLogical

constraints

Terms/glossary

Thesauri“narrower

term”relation

Formalis-a

Frames(properties)

Informalis-a

Formalinstance

Value Restrs.

Disjointness, Inverse, part-of…

Ontologies Come of Age McGuinness, 2001, and From AAAI Panel 99 – McGuinness, Welty, Uschold, Gruninger, LehmannPlus basis of Ontologies Come of Age – McGuinness, 2003

A Few Observations about Ontologies (March 2000 / 2011)

– Simple ontologies can be built by non-experts• Consider Verity’s Topic Editor, Collaborative Topic Builder, GFP interface, Chimaera, etc. More

tools, more ontologies, more expressiveness points– Ontologies can be semi-automatically generated (now more machine learning as well as from social

collaborative settings) • from crawls of site such as yahoo!, amazon, excite, etc.• Semi-structured sites can provide starting points

– Ontologies are exploding (business pull instead of technology push)• most e-commerce sites are using them - MySimon, Affinia, Amazon, Yahoo! Shopping,, etc.• Controlled vocabularies (for the web) abound - SIC codes, UMLS, UN/SPSC, Open Directory,

Rosetta Net, …• Business ontologies are including roles• DTDs are making more ontology information available • Businesses have ontology directors• “Real” ontologies are becoming more central to applications (and real ontologies arising from

massive data )• New models such as virtual observatories accelerating pull

But now….

Past emphasis was more on expressiveness and building (usually by trained experts)

Now the emphasis is more about the ecosystem in which ontology-enabled applications are embedded, maintained, understood, trusted, and used...

Semantic Agents

Semantically-enabled advisors utilize:

• Ontologies• Reasoning• Social• Mobile• Provenance • Context

Patton & McGuinness.et. altw.rpi.edu/web/project/Wineagent

Ontology Ecosystem Discussion & Directions

• Base ontology very simple– Wine, Winery, Grape, Flavor, Body, Color, Sugar– Stood the test of time: Living with Classic (1991),,

CLASSIC tutorials, Ontologies 101, OWL Guide, … – To scale however, needs to be compatible with WIDE

range of menus, wine lists, vocabularies. Not hard to obtain but significant enhancement required.

– Needs more ecosystem support – explanation, provenance, validation, inconsistency detection, prioritization scheme, UI considerations, additional social connections, citizen-oriented maintenance and evolution schemes, scale, partitioning, …

www.ksl.stanford.edu/people/dlm/papers/living-with-classic-abstract.html

SemantAqua / SemantEco

Aimed at helping people investigate local water quality Diverse datasets, regulations, datatypes Uses lightweight semantic technologies to produce mashups that make data accessible that would be otherwise difficult to view in perspective Maintains provenance about data and manipulations Potential to empower citizens with contextualized data and support citizen scientist questions and reporting

Tues Demo & Wed aft talk

Discussion and Directions

• Base ontology also very simple - Water, contaminant, threshold, test

• Simple use of recognition and easily extensible (e.g., recently with health impact data)

• To scale however, needs to be compatible with wide range of data source vocabularies, including a wide range of tests

• New processes create new vocabulary needs (e.g., protectingourwaters.wordpress.com/2011/06/16/black-water-and-brazenness-gas-drilling-disrupts-lives-endangers-health-in-bradford-county-pa/ )

• Needs more ecosystem support – explanation, provenance, validation, inconsistency detection, prioritization scheme, UI considerations, additional social connections, citizen-oriented maintenance and evolution schemes, scale (3 billion triples and counting), partitioning…

November 9, 2006 10

Interdisciplinary Virtual Observatory (VSTO)

• General: Find data subject to certain constraints and plot appropriately

• Specific: Plot the observed/measured Neutral Temperature as recorded by the Millstone Hill Fabry-Perot interferometer while looking in the vertical direction at any time of high geomagnetic activity in a way that makes sense for the data.

Semantic Web Methodology

Originally developed for VSTO, now in SSIII, SESDI, SESF, OOI …

McGuinness, Fox, West, Garcia, Cinquini, Benedict, Middleton The Virtual Solar-Terrestrial Observatory: A Deployed Semantic Web Application Case Study for Scientific

Research. Proc. 19 Conf. on Innovative Applications of Artificial Intelligence (IAAI-07),

http://www.vsto.org

Inference Web: Making Data Transparent and Actionable Using Semantic Technologies

• How and when does it make sense to use smart system results & how do we interact with them?

12

Knowledge Provenance in

Virtual Observatories

Hypothesis Investigation

/ Policy Advisors

(Mobile) Intelligent

Agents

Intelligence Analyst Tools

NSF Interops:SONETSSIII – Sea Ice

Ontology Ecosystem Discussion & Directions

• Base ontology relatively simple - Instrument, Observatory, Data Product, …

• Initially done for solar terrestrial physics but has been used in volcanology, plate tectonics, sea ice, water, … with relatively little rework (NSF: VSTO, SPCDIS, SESF, SSI, SONET … NASA: SESDI, …)

• Modularity has been key – both to reusing other ontologies (e.g., SWEET) and in expanding our reuse

• To scale and be maintainable however, need to be compatible with WIDE range of evolving vocabularies. (Unlike the wine agent and to some extent the water quality portal, this is not as uncomplicated,)

• Needs more ecosystem support – explanation, provenance, validation, inconsistency detection, prioritization scheme, UI considerations, additional social connections, citizen-oriented maintenance and evolution schemes, scaling, partitioning, …

What is different now (10+ years later)?

• Ontologies (at many points on an expressiveness spectrum) are in use in wide variety of settings and disciplines and are built by a broad(er) range of users

• Ontologies are becoming longer lived…. With that some best practices are emerging including – Modularity– Designing for reuse (minimizing tight constraints, naming, …)– Modules with more depth– Provenance considerations – provenance info included and service

connection example • Recommended Web Ontology Language (and business

consequences), Rules recommendation, Provenance on its way• Issues are much less about starting points for ontologies – they are

now about mapping, reusability, maintenance, and sustainability• Issues are not only technical – social issues of team acquisition and

maintenance are at least as important

Linked Data Cloud

What might we do?

• Guidelines for creating ontologies for reuse – modularity, limited conflict generators, ease of use considerations – one early one was explanation for debugging along the lines of McGuinness’ thesis

• Provenance - Representation (e.g., W3C working group), Watermarking , …• Semi-automatic tools for ontology creation and maintenance

– Checking– Expanding– Mapping

• Hybrid tools for working with learning / discovery tools AND humans • Exploit citizen xx and social• More testimonials in forms that serve as operational specifications• Directions for examples such as Watson-style work

• What do you need for ontologies to be practically and sustainably used in commodity computing? - forthcoming 4th paradigm blog post

Questions / Suggestions? dlm @ cs . rpi . edu

Questions?Come to Demo session on SemantAqua on Tuesday

Come to talk on Semantic Monitoring on Wednesday afternoon

dlm <at> cs <dot> rpi <dot> edu

Ontologies for the Real WorldDeborah L. McGuinness

Tetherless World Senior Constellation ChairProfessor of Computer and Cognitive ScienceRensselaer Polytechnic Institute

BACKUP SLIDES