Transcript BL JISC UKOLN Workshop

Digital Libraries and the business
process: reflections on a theme
Dr Liz Lyon, Director
UKOLN, University of Bath, UK
BL/JISC/UKOLN Workshop, British Library
March 2006.
This work is licensed under a Creative Commons Licence
Attribution-ShareAlike 2.0
UKOLN is supported by:
www.ukoln.ac.uk
a centre of expertise in digital information management
www.bath.ac.uk
Overview
1. Mapping the business process: the
intricate mix of humans and machines
2. Some thoughts about workflow
3. Social networks and service development
4. Summary: take home message
British Library / JISC / UKOLN Workshop
2
What do we mean by
“business process”???
British Library / JISC / UKOLN Workshop
3
“A business process is a collection
of related structural activities that
produce something of value to the
organization, its stake holders or its
customers. It is, for example, the
process through which an
organization realizes its services to
its customers”.
British Library / JISC / UKOLN Workshop
4
“The linkage of business process
with value generation leads
some practitioners to view
business processes as the
workflows which realize an
organization's use cases”.
…..Workflows???
British Library / JISC / UKOLN Workshop
5
“Workflow at its simplest is the
movement of documents and/or tasks
through a work process.
More specifically, workflow is the
operational aspect of a work procedure:
how tasks are structured, who performs
them, what their relative order is, how
they are synchronized, how information
flows to support the tasks and how tasks
are being tracked”.
British Library / JISC / UKOLN Workshop
6
“Distinction can be made between
"scientific" and "business" workflow
paradigms.
While the former is mostly concerned
with throughput of data through various
algorithms, applications and services,
….the latter concentrates on scheduling
task executions, including dependencies
which are not necessarily data-driven
and may include human agents”.
British Library / JISC / UKOLN Workshop
7
eBusiness
eScience
Closed + secure systems
Extremely open (data)
Resources are finite +
known
Describe + discover
resources: rich
semantics + metadata
standards
High levels of trust
3rd party verification
Mission critical + liability
Peer review
3rd party repetition +
re-enactment
Static (mission critical)
Dynamic, agile, iterative,
flexible, rapid modificat’n
Small data volumes
Simple structures
Large data volumes
Highly complex
Transaction-centric
Not transaction-centric?
Customers + managers
Researchers are users
and managers
Comparing
workflow
Tom Oinn 2003
http://twiki.mygrid.org.uk/
British Library / JISC / UKOLN Workshop
8
eBusiness
eScience
eLibraries
Closed + secure systems
Extremely open (data)
Mixed model
OA+licensed content
Resources are finite +
known
Describe + discover
resources: rich
semantics + metadata
standards
Describe + discover: “core”
metadata schema, highlevel vocabularies, KOS
Community tagging
High levels of trust
3rd party verification
Mission critical + liability
Peer review
3rd party repetition +
re-enactment
Provenance, trusted
digital repositories,
trusted (reliable) services
Static (mission critical)
Dynamic, agile, iterative,
flexible, rapid modificat’n
Mixed model but trend to
be more agile
Small data volumes
Simple structures
Large data volumes
Highly complex
Mixed model: distributed,
federated, centralised
Transaction-centric
Not transaction-centric?
Mixed model: loans vs
preservation
Customers + managers
Researchers are users
and managers
British Library
/ JISC / UKOLN Workshop
Consumers and
producers?
9
OK - so in the context of our
institutions ……
(and digital libraries)…..
what exactly do we mean by
“business process” and
“workflow” ???
British Library / JISC / UKOLN Workshop
10
(Very simple) e-Research Cycle
(New) knowledge
extraction: data
mining, modelling,
analysis, synthesis
Data processing
Formulate hypothesis / ideas, test,
experiment, observe: data creation,
collection & capture
Data processing
Data processing
Data management
storage & validation:
description, deposit,
self-archiving,
preservation,
certification
e-Infrastructure
Adding value: Data
linking, annotation,
visualisation, simulation
Open access
Collaboration
Data processing
Data processing
Scholarly communications: data disclosure,
publication, citation, discovery, re-use
This work is licensed under a Creative Commons License
Attribution-ShareAlike 2.0
British Library / JISC / UKOLN Workshop
11
Gathering information about (e-)research
• Project StORe: Source-to-Output
Repositories (Edinburgh)
– primary data : research publications
– Survey questionnaire
• RepoMMan: Repository Metadata
and Management (Hull)
– Survey questionnaire and interviews
– Activity diagram
British Library / JISC / UKOLN Workshop
12
JISC Digital Repository Programme
DigiRep wiki
http://www.ukoln.ac.uk/repositories/digirep/index/JISC_
Digital_Repository_Wiki
British Library / JISC / UKOLN Workshop
13
Data capture
• R4L Repository for the Laboratory Project (JISC-funded)
automated data capture from instrumentation, deposit of
results (chemistry) at Univ. Southampton
• SMART TEA electronic Laboratory notebook +
annotations
• R4L deposit scenario
British Library / JISC / UKOLN Workshop
14
User scenario (…part of….)
1.
2.
3.
4.
5.
6.
7.
8.
Produce strategy for synthesis (=idea)
Submit plan to SmartTea system (incl. identifiers)
Retrieve and follow instructions (sub-workflow?)
Experimental synthesis metadata automatically
recorded on instruments (Smart Lab)
Create record for synthesised sample (+ proposed
chemical identifier) in R4L laboratory data
management system
Run spectral analyses on sample capturing further
analysis metadata (incl. time-stamp, analysis
software version, researcher details etc.)
Save spectrum in native and common formats
Invoke R4L data capture service and deposit files +
metadata in laboratory repository….
British Library / JISC / UKOLN Workshop
15
Services for simple & rapid deposit
Data manipulation toolbox
Associated Metadata
Value added
Format
conversion
British Library / JISC / UKOLN Workshop
16
Crystallography workflow
RAW DATA
DERIVED DATA
RESULTS DATA
• Initialisation: mount new sample set up data collection
• Collection: collect data
• Processing: process and correct images
• Solution: solve structures
• Refinement: refine structure
• CIF: produce CIF (Crystallographic Information File)
• Validation: chemical & crystallographic checks
• Report: generate Crystal Structure Report
British Library / JISC / UKOLN Workshop
17
A data repository entry
ecrystals.chem.soton.ac.uk
British Library / JISC / UKOLN Workshop
18
Access to the underlying data
British Library / JISC / UKOLN Workshop
19
Laboratory Repositories R4L
Slide: Simon Coles,
Univ. Southampton
British Library / JISC / UKOLN Workshop
20
eBank UK Project
http://www.ukoln.ac.uk/projects/ebank-uk/
• Aggregator service harvests metadata from institutional repository
(e-crystals archive)
• eBank service embedded in PSIgate portal for 3rd party search
• Service linking from data to derived research publication
• Embedding eBank service in learning workflows
UKOLN (lead), University of Southampton, University of Manchester
British Library / JISC / UKOLN Workshop
21
But….
….how should we be
“formalising” workflows?
British Library / JISC / UKOLN Workshop
22
Workflow systems & standards
•
•
•
•
•
•
•
•
•
•
YAWL
METEOR-S
BPEL
OpenWFE
RADRunner
BPSS (ebXML)
PSL
Geo-Opera
JDF
XLANG
•
•
•
•
•
•
•
•
•
•
Taverna
Kepler
Pegasus
Triana
SPA
ICENI
BioOpera
Wildfire
BPML
WS-CDL
Is “workflow standard” an oxymoron?
British Library / JISC / UKOLN Workshop
23
Kepler Project
http://kepler-project.org/Wiki.jsp?page=KeplerProject
British Library / JISC / UKOLN Workshop
24
http://taverna.sourceforge.net/
British Library / JISC / UKOLN Workshop
25
Slide: Carole Goble
British Library / JISC / UKOLN Workshop
26
DL workflows : a complex picture
• Workflows for data capture, deposit,
preservation, citation, discovery, mining &&….
• Multiple workflows interacting together
• Workflows may call on each other, in a
defined order
• Multiple workflows may use “common”
services e.g. Assign (identifier)
• Require sequential or parallel execution, have
dependencies, be time-limited, repetitive
• Have an owner (control)
• Include essential human interventions
• ???
British Library / JISC / UKOLN Workshop
27
Workflow…the answers to Who?
What? When? in a business
process.
A workflow is only as good as the
business process beneath it.
Margie Virdell, IBM developerWorks
British Library / JISC / UKOLN Workshop
28
Some observations….
1. We don’t know enough about institutional business
process:
– Learning & teaching, research, admin, enterprise
2. How to analyse, express and model processes
3. What types of models?
– At what levels of granularity: strategic (for a manager) vs
detailed mathematical specifications (for a developer)
4. Which workflow tools & standards should we use?
5. Learn from e-Science projects
6. Which processes are best driven by machines and
which by humans?
7. How do human-directed processes interact with
machine-driven ones?
8. What are the digital library “touch points” in these
processes?
British Library / JISC / UKOLN Workshop
29
Service-oriented architectures
for Digital Libraries
•
•
•
•
•
•
•
•
•
Produce process models (DLF?)
Experience of VRE projects
Integrative Biology user scenarios
Service typology (e-Framework?)
Identify services: service definitions
Service interactions: service patterns
Orchestration of Web services
Choreography of Web services
Workflow interoperability….
(another oxymoron?)
British Library / JISC / UKOLN Workshop
30
“Orchestrating the knitting”
“Integration trumps re-invention”
“We work in
a services ecosystem”
“new social models for DLs”
“Polygamous recombination”
British Library / JISC / UKOLN Workshop
31
Discovering data:
• Domain identifier:
International
Chemical Identifier
(INChI) code
• Google molecule
using INChI
Slide from Simon Coles
Coles, S.J., Day, N.E., Murray-Rust, P., Rzepa, H.S., Zhang, Y., Org.
Biomol. Chem., 2005, (10),1832-1834. DOI: 10.1039/b502828k
British Library / JISC / UKOLN Workshop
32
Avian flu outbreaks mashup - Nature January 2006
British Library / JISC / UKOLN Workshop
33
New
prototype
services
British Library / JISC / UKOLN Workshop
34
British Library / JISC / UKOLN Workshop
35
Take home messages
•
•
•
•
Need to understand more about institutional
business process: cultural heritage, learning &
teaching, research, admin, enterprise…
Assessment of the value of workflow studies
Evaluation of workflow systems, tools & standards
How best to analyse, express and model processes
–
–
•
•
Types of models
At what levels of granularity
Interactions between human-directed processes &
machine-driven ones: implications for services
Social development of Digital Library services:
creation, interaction, recombination and integration
….an intricate mix of humans &
machines
British Library / JISC / UKOLN Workshop
36
Thank you.
More information: UKOLN http://www.ukoln.ac.uk/
UKOLN receives core funding from the Joint Information Systems
Committee (JISC) and the Museums, Libraries & Archives Council
(MLA) and is based at the University of Bath, UK.