|
|
|
David De Roure |
|
Nicholas Jennings |
|
Nigel Shadbolt |
|
|
|
|
|
David De Roure |
|
Distributed Systems |
|
Nigel Shadbolt |
|
Advanced Knowledge Technologies |
|
Nick Jennings |
|
Agent Based Computing |
|
|
|
Involved in three testbeds, two IRCs, TAG, ATF |
|
|
|
|
|
|
Motivation |
|
Service oriented architectures |
|
Distributed systems |
|
Semantic Grid |
|
Recommendations |
|
|
|
|
Vision of e-Science with high degree of easy to
use and seamless automation, with flexible collaborations and computations
on a global scale |
|
Gap between vision of e-Science and current
endeavours |
|
Three layer model compelling but much
hand-waving about knowledge layer |
|
Concern about scalability assumptions |
|
Lack of holistic approach – Grid starting at
socket on wall |
|
Need for universal architecture |
|
|
|
|
A Research Agenda aiming to move from the
current state-of-the-art in e-Science infrastructure to the future
infrastructure that is needed to support the full richness of the e-Science
vision. |
|
|
|
Caveat: This is not a blueprint for all
e-Science infrastructure research! The
absence of recommendations doesn’t mean there are no research issues. |
|
|
|
|
The Semantic Grid is to the Grid what the
Semantic Web is to the Web |
|
|
|
|
|
|
|
|
|
Which begs a question… |
|
|
|
|
|
|
“The Semantic Web is an extension of the
current Web in which information is given a well-defined meaning, better
enabling computers and people to work in cooperation. It is the idea of having data on the Web
defined and linked in a way that it can be used for more effective
discovery, automation, integration and reuse across various
applications. The Web can reach its
full potential if it becomes a place where data can be processed by
automated tools as well as people” |
|
- TBL |
|
|
|
|
Commissioned for Core e-Science Programme |
|
Not a (or the) roadmap J |
|
Aim to bridge communities |
|
Not a comprehensive survey, e.g. physical layer
and comms out of scope |
|
Draft distributed in July to TAG |
|
Samizdat publication was influential |
|
Completed in December after further comment
cycle |
|
Final version ready for formal distribution
December 2001 |
|
|
|
|
|
|
|
Uses 3 layer model as a narrative (and
sociological?) structure, not an architecture! |
|
|
|
Introduction |
|
Service-oriented architectures |
|
Data/Computation |
|
Information |
|
Knowledge |
|
Recommendations |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Abstract characterisation of some data or
processing capabilities |
|
Service of providing data on sea temperatures
for last 50 years |
|
Service of correlating sea temperature from data
source 1 with annual rainfall from data source 2 and making prediction for
coming year |
|
Service of finding all data on climate change
over past 20 years and being continually informed as new data becomes
available |
|
Service of finding all users who are interested
in climate change |
|
|
|
|
|
|
Services have owner(s) |
|
Owners set access conditions |
|
Universal vs. Restricted access |
|
Free vs. Paid for |
|
Services have consumers |
|
Service Level Agreement (contract) specifies
relationship between service owners and service consumers |
|
Services can be created and composed |
|
|
|
|
|
|
|
|
|
|
|
|
Service Owner |
|
Service creation |
|
Service advertisement |
|
Service level agreement creation |
|
Service delivery |
|
Service Consumer |
|
Service discovery |
|
Service location |
|
Service level agreement creation |
|
Service result receipt |
|
|
|
|
|
|
|
|
|
|
|
|
|
Web-based computing? |
|
Transactions aren’t large scale (or
long-running) |
|
Content is large scale but QoS is very poor |
|
Web is powerful because it is so simple |
|
Client-server constraint vs async message
passing |
|
Simplistic failure handling |
|
Crossing firewalls is a relevant issue |
|
|
|
|
|
Distributed object systems eg CORBA? |
|
Proven intranet solution |
|
Software engineering paradigm eg UML |
|
Not accepted as an internet computing solution |
|
Peer-to-peer? |
|
Self-organisation attractive |
|
Concerns over heterogeneity and security |
|
Resource discovery even more important |
|
Hybrid solutions desirable (as well as
inevitable) |
|
|
|
|
|
|
Broadly we need intranet features/services on an
internet scale; e.g. reliable secure messaging, information consistency,
directory services |
|
The document catalogues a number of example
technologies and deployments in order to establish requirements |
|
Identifies research issues in line with GGF |
|
Particular concern over need to cross
organisational boundaries |
|
|
|
|
|
Web is designed for information distribution |
|
Useful as content infrastructure (see later!) |
|
But beware well known problems: |
|
Version control |
|
Provenance |
|
Quality |
|
Pull model |
|
Considerable activity in search, including
distributed search |
|
Links to databases community e.g. query
decomposition and routing |
|
|
|
|
|
Content and metacontent |
|
XML(S) |
|
RDF(S) |
|
Need shared vocabularies |
|
Semantic Web |
|
“Semantic Web is not AI” - TBL |
|
Some open research issues but this level
established |
|
Vision of metadata everywhere |
|
What will motivate the tools for metadata
creation? |
|
|
|
|
|
|
|
Instantiation of service-oriented model |
|
XML Protocol |
|
Web Services Description Language |
|
Universal Description Discovery and Integration |
|
Workflow description |
|
Web Services Flow Language |
|
XLang |
|
Other proposals emerging |
|
Many issues discussed earlier apply to Web
Services! |
|
|
|
|
|
|
|
i.e. not just publishing things at each other J |
|
Examples |
|
Live data and visualisation from experimental
kit |
|
Live video feeds |
|
Videoconferencing |
|
IRC, MUDs, chat rooms |
|
Collaborative Virtual Environments |
|
Access Grid |
|
Also need metadata |
|
How do these scale? |
|
|
|
|
|
Smart laboratory |
|
Mediated spaces |
|
Devices e.g. Electronic lab book |
|
Presence |
|
Of e-scientist and kit in virtual world |
|
Of virtual world in physical world (augmented
reality) |
|
|
|
|
|
Strategies for e-Science content-types |
|
Digital rights management in e-Science context |
|
Provenance, for |
|
Reuse |
|
Repeat |
|
Legal evidence |
|
Adaptation and personalisation |
|
Metadata in collaborative events |
|
Collaboration for large scale community |
|
Workflow description and enaction |
|
Access control in context of automation |
|
|
|
|
|
|
The aim of the knowledge layer is to act as an
infrastructure to support the management and application of scientific
knowledge to achieve particular types of goal and objective. |
|
Note sheer scale of information content – we
need abstracted and annotated content |
|
|
|
|
Acquire – make info usable (could be tacit) |
|
Model – bridges gap between acquisition and use |
|
Retrieve – finding knowledge, or a subset |
|
Reuse – rather than reaquire |
|
Publish – right knowledge to right person at
right time |
|
Maintain – updating (and discarding) |
|
|
|
|
|
Current work is extending RDF with concepts from
knowledge representation languages |
|
Shared vocabulary with rules/axioms for
inference |
|
DAML, DAML+OIL (UK strength) |
|
W3C Web Ontology working group has been created |
|
There are many kinds of ontologies |
|
Domain, Task, Quality, Value, Personalisation,
Argumentation |
|
Need to be developed and maintained |
|
Tools exist to work with emerging standards |
|
|
|
|
|
Services needed for Semantic Grid: |
|
Knowledge discovery (e.g. patterns in
information sets) |
|
Clustering and indexing |
|
Ontology mapping |
|
Dynamic annotation |
|
Summarisation |
|
Visualisation |
|
Monitoring, diagnosis, assessment |
|
|
|
|
|
|
|
|
Involve knowledge aspects |
|
Exploiting resources to solve particular
problems |
|
Approach problem in terms of application area
semantics |
|
Provide semi-automation |
|
Potentially supported by reasoning services |
|
Long history in knowledge-based systems |
|
Could incorporate collaboration e.g. Access Grid |
|
|
|
|
|
|
Languages and infrastructures for knowledge
services (e.g. DAML-S) |
|
Methods for large-scale ontologies |
|
Distributed annotation services |
|
Knowledge capture tools as plugins |
|
Dynamic linking, visualisation, navigation |
|
Retrieval based on annotations |
|
Routine natural language processing |
|
Internet-based reasoning services |
|
Promotion of collaboration, e.g. in Access Grid |
|
|
|
|
|
|
|
Technical and conceptual infrastructure |
|
Grid toolkits |
|
Smart Laboratories |
|
Service-oriented architectures |
|
Agent-based approaches |
|
Network philosophies |
|
Trust and provenance |
|
|
|
|
|
Content infrastructure |
|
Metadata and annotation |
|
Knowledge technologies |
|
Integrated media |
|
Content presentation |
|
|
|
|
|
Bootstrapping activities |
|
Starter kits |
|
Exemplar and reference sites |
|
Use cases |
|
|
|
|
|
Human Resource issues |
|
Community building |
|
System support |
|
Training |
|
|
|
|
|
e-Science intrinsics |
|
E-Science workflow and collaboration |
|
Pervasive e-Science |
|
|
|
|
|
|
Future proposed directions |
|
Core computer science research |
|
Extension to e-Anything |
|
e-Health |
|
e-Learning |
|
e-Society |
|
e-Business |
|
… |
|
Notes