Weaving a Web for the Next Generation of Science

Web Services and the Semantic Web: Motivations and Challenges.

Steve Bratt 
World Wide Web Consortium

Leading the Web to its Full Potential ...

http://www.w3.org/

Speaking at
16th IRIS Workshop

Tucson, Arizona, USA
12 June 2004

Slides: http://www.w3.org/2004/Talks/0612-sb-wsswapps/

Questions Addressed

• Why employ Web Services and Semantic Web?

• Who is making the leap? How? Where?

• How might your community and other scientific communities do this?

• How much might you benefit?

• How much might this cost?

• Builds on yesterday's talk:

  "Spinning Threads for the Next Generation of the Web"

  ... which focused on "What" are WS and SW

Re-Cap: Web Services and Semantic Web

Toward a Web of machine accessible resources Web Services: Web of applications* Standards for interactions between programs on the Web Easier to expose and use services Semantic Web: Web of data* Standards for data, relationships, descriptions on the Web Easier to Find, Share, Aggregate, Extend information

Today's Scientific Community (1)

Specialization

Increasingly specialized ... ... yet, important problems require multidisciplinary approach

Today's Scientific Community (2)

Data, Information, Knowledge

Massive quantities of data ... Growing in size and complexity

• Data in compartments not understandable outside one discipline
  • Same terms, inconsistent meanings
  • Different terms, same meaning
  • Different terms, related (not identical) meaning
  • ... info also in numerics, images, graphs, etc.

• Data not accessible outside one discipline
  • Specialized programs to read, write, exchange, convert

Today's Scientific Community (3)

The Web

Thanks to global standards and universality, the Web is crucial to:

Researching Publishing Data services

However, today:

• Most resources (applications, data) identified by text ...
• ... searched for using text strings+ ...
• Cannot find nor use resources on the Web that may be important.
• Resources that could be important are not on the Web in a way they can be used

Seismological Community

Recognition of challenges, and leadership to address them ...

Global data exchange and processing within community is a necessity "Standards" developed within subcommunities Exchange between subcommunities facilitated by docs and tools Inventing solutions to fill technology gaps Early adoption of commercial or open source solutions: Satellite comm and Internet for data exchange DBMS for metadata storage, data management Web for information exchange, data access Historic role of funding agencies in motivating new technologies

Problem: Finding

Current Situation:

• Search Web for information on "Love waves"

Issue:

• Some useful, some not-so-useful hits:
  • "Love Wave Research at Brown"
  • "Courtney Love Waves the Red Flag"
  • " 'Buy Faith, Hope Love: Waves and Reflections', at Amazon and SAVE"
  • etc. ...

Problem: Sharing

Current Situation:    

• Oceanography Web publication contains interesting data
• Want to add to my local data store

Issues: 

• Data store doesn't understand Web page
• Seismology does not understand oceanography jargon
• Must copy-convert-paste or develop specialized applications
• Neither flexible nor easily extensible

Issue: Aggregating

Current Situation:

• Web site A: fault information
• Web site B: electromagnetic observations
• Web site C: hydro and gas measurements
• Web sites D, E, F: earthquakes data from various networks
• Web sites G, H, I ... : map, source, propagation, acceleration, site response, etc. services
• Web sites X, Y, Z ... : gov't emergency authorities
• Want to aggregate data to (for R&D or realtime alerts):
  • compare network locations and compute new ones
  • examine correlations of geophysical observations
  • interface with authorities re: events and risks

Issues:

• Different databases, attributes, locations, services
• Can do now by building specialized systems
• Neither flexible nor easily extensible

Tomorrow's Scientific Community?

Web of machine usable applications and processable data:

Unique names for data, applications, descriptions, relationships (URIs) Web Service-enabled applications (SOAP, WSDL, etc.) Common data model for expressing information (RDF) Common vocabulary framework (OWL) Communities develop their own ontologies Shared ontologies for widely used vocabularies Core math/sci, people, calendar, geospatial ... Mapping ontologies between community ontologies, where needed

(Science and the Semantic Web, J. Hendler, Science, Jan 2003)

Possible Benefits to the Science and Society

... due to the use of global Web standards

• Computer assisted networking and understanding across communities

• Greater sharing of methods and tools
  • Research
  • Systems

• Enhanced opportunities for:
  • Communication
  • Collaboration
  • Creativity
  • Discovery
  • Efficiency and extensibility

How Might We Get There?

... perhaps in similar way to how Web got here:

1. the Web was born at CERN…
2. …was first picked up by high energy physicists…
3. …then by academia at large…
4. …then by small businesses and start-ups…
5. “big business” came later.

Network effects kicked in early ...

• Semantic Web: at #4, moving to #5.
• Web Services: at #5.

Questions for Geophysics

• Most pressing problems for geophysics and society?
  • EQs, volcanoes, resources, security, climate change, water, environment, magnetic field, near-earth objects, super storms, etc.
  • ... and the basic research necessary to address these.

• Data, knowledge, methods needed to solve them?
  • From within community?
  • From outside community?
  • How do you know?
  • Barriers to info flow, collaboration?

• Systems needed to support science?
  • Build new?
  • Maintain and extend existing?
  • Investment in systems vs. R&D

Neighbors

Greatest benefits will be from network effect of linking communities:

GEON

• http://www.geongrid.org/
• Building a prototype national, cross-disciplinary Geosciences Cyberinfrastructure Network
• Geology, geophysics, hydrology, petrology, geochemistry, structural geology, sedimentology,
• Applying XML and Semantic Web technologies

Semantic Web for Earth and Environmental Terminology

• http://sweet.jpl.nasa.gov/
• Broad range of earth, oceanic, atmospheric, space science terms

Other Web Service & Semantic Web Applications

Rapid growth of both ...

• Selected compilations of Web Service apps:
  • http://www.xmethods.com/
  • http://directory.google.com/Top/Computers/Programming/Internet/Web_Services/

• One compilation of Semantic Web apps:
  • http://www.w3.org/2003/11/SWApplSurvey

• Both
  • http://wsindex.org/

... also, see additional case studies and links to development tools at end of this presentation.

Lessons Learned

Benefits of WS and SW technologies might depend upon:

• Complexity and dynamics of community

• Importance of planned vs. serendipitous discovery

• Services and products
  • Life cycle
  • Rate / volume of access

• Raw and metadata
  • Complexity
  • Volume
  • Rate of growth
  • Rate of schema change

• Potential benefits of linking with other communities
  • ... and characteristics of their services and data

High complexity, dynamics, usage, desire to interact suggest WS and SW could prove valuable.

Conclusions

Web Services and the Semantic Web seem to be worth considering:

• In conjunction with considering "Questions" in previous slide.

Ways to get started:

• Track W3C standards and best practices developments

• Explore tools

• Participate in current WS and SW efforts (e.g., GEON)

• Pick new candidate community apps and DBs for experimental adaptation

• Work with neighboring communities to build Web of science

W3C is happy to help in any way we can!

Resources Slides

Subsequent slides provide a survey of Web Services and Semantic Web tools (and applications built with them), and additional use cases.

Some Web Services Tools

• Web Services Project (Apache) apps
• SOAPy for Python (sourceFORGE)
• Java Web Services (Sun)
• .net (Microsoft) apps
• webLogic (BEA) apps
• Web Services Technology Center (Oracle)
• WebSphere (IBM) apps
• ... and many more (Iona, Systinet, ETC.)

* Most products not yet using new W3C standard versions of SOAP (1.2), WSDL (2.0), etc., but they will in the future.

Some Semantic Web Tools

RDF Environments (interpreters, etc)

• Jena (from HP; for Java, includes OWL reasoning)
• RDFLib (for Python)
• Redland (in C, with interfaces to Tcl, Java, PHP, Perl, …)
• Sesame (RDF Schema based repository and querying facility)
• cwm (general-purpose data processor for the semantic web)
• Protege (ontology editor)
• Access to Relational Databases (white paper)
• Annotea (annotation of Web documents)

SemWeb Graphical Editors

• IsaViz (Xerox Research/W3C),
• RDFAuthor (Univ. of Bristol)

Amazon Web Services

• http://www.amazon.com/gp/browse.html/104-6931018-3084717?node=3435361
• To integrate Amazon services into another vendors Web site

Haystack

• http://haystack.lcs.mit.edu
• MIT research effort on flexible, extensivble information management
• bioHaystack and Semantic Web browser application

Dublin Core

• http://www.dublincore.org
• vocabularies for distributed Digital Libraries
• one of the first metadata vocabularies in RDF

Web Content Syndication (RSS)

• http://purl.org/rss/
• Specify important content of a page
• Widely used in the weblog world
• Example: W3C home page syndicated

Data Integration with Organizations

• Achieve semantic integration of corporate resources or different databases
• RDF/OWL based vocabularies as an “interlingua” among system components
• Boeing example: http://www.cs.rutgers.edu/~shklar/www11/final_submissions/paper3.pdf
• Similar approaches: Artiste project, MuseoSuomi, …

Adobe XMP

• http://www.adobe.com/products/xmp/main.html
• Tool to add RDF-based metadata to all their file formats
• eg, Photoshop in Creative Suite - millions of people use RDF without knowing it…

Baby CareLink

• http://www.babycarelink.com/
• Information for the treatment of premature babies
• OWL service as a Web Service
• Combines disparate vocabularies like medical, insurance, etc
• Users can add new entries to ontologies; ask complex questions

Use Case: Drug Discovery

Revolution underway

• Huge focus on cross-disciplinary research:

  (e.g., Boston Globe 9 Jun 2004)

  • Biology, chemistry, computer science, math, business
  • Basic research, exploratory research, discovery, production

• Extremely dynamic, high-risk/reward environment

• Companies joining W3C to develop Web Services and Semantic Web technologies

Rising Cost Per Drug-to-Market

Pharmaceutical Process

(from http://akosgmbh.de/functions.htm)

Basic Drug Research

• Aventis combs massive quantities of information to optimize discovery process
  • “Embedded” semantics - key facts not machine-readable
  • Literature, images, databases with annotations, statistical analyses
  • Discipline-specific semantics / jargon
  • Scientists search text, cut-and-paste

Application of Web Services and Semantic Web

• Created few tens of Web Services to support data acquisition
  • Many simple services => mix and match
  • Focused on highest-usage functions
  • Used commercial tools to develop, manage, extend

• Longer-term goal is apply Semantic Web to provide:
  • Greater cross-discipline correlation
  • More flexible, extensible data management
  • Faster, more accurate ID of molecule-target pairs and new drugs

Drug Discovery Vision

Acknowledgements

Thanks to:

• Tim Berners-Lee
• David Booth
• Hugo Haas
• Ivan Herman
• Philippe Le Hégaret
• Eric Miller
• Eric Prud'hommeaux
• Ralph Swick
• John Wilbanks

END