Transcript Microsoft PowerPoint - NCRM EPrints Repository

What is e-Research?
Rob Procter
Manchester eResearch Centre
University of Manchester
Research Methods Festival 2010
Outline
■ Overview of e-Research


What is e-Research?
e-Research drivers
■ e-Research in the social sciences




Data collection
Analysis
Visualisation
Collaboration
■ What might e-Research mean for you?
■ Where to find out more
■ Questions
What is e-Research?
■ Application of advanced digital methods
and tools in all parts of research lifecycle:
 Locate and access research resources.
 Discover, access, integrate and analyse
digital data on a hitherto unrealisable scale.
 Facilitate sharing and collaboration.
e-Research: enhancing research practice
Data curation
Publication
Literature search
Visualisation
Analysis
Research
Lifecycle
Data fusion
Data preparation
Literature
review
Data discovery
Data collection /
re-use
e-Research drivers
■ Research challenges become more
complex:
 Larger in scale, multi-disciplinary
■ The ‘data deluge’:
 Volume of digital research data is
increasing at an exponential rate.
Looking for the ‘God particle’
Large Hadron Collider
The data deluge
ユビキタス
セキュリティ
1ZB
(2010)
161EB
(2006)
ではない 情報系アンブレラ
GRID/ペタコン
ITS
Slide: Satoshi Matsuoka
Social Science research challenges
1. Global Economic Performance,
Policy and Management
2. Health and Wellbeing
3. Understanding Individual
Behaviour
4. New Technology, Innovation
and Skills
5. Environment, Energy and
Resilience
6. Security, Conflict and Justice
7. Social Diversity and Population
Dynamics
The data deluge in social sciences
■ ‘Born digital’ data is generated on increasing
scale as by product of everyday activities:
 Patterns of consumption:
- Public and private goods and services
 Patterns of communication:
- Email, bulletin boards, weblogs, chat rooms, news feeds, mobile phones,
SMS
 Patterns of movement of people and goods:
- CCTV, speed cameras, traffic monitoring, GPRS, embedded devices
■ Move from survey-based methods to using
administrative data
Ian Diamond
http://www.understandingsociety.org.uk/
Open data
Easier access
The social Web
Statistical analysis
Multilevel
modelling through
MLwiN and e-Stat
Geographically Weighted
Regression
http://www.cmm.bristol.ac.uk/research/NCESS-EStat/
Social simulation
National e-Infrastructure for Social
Simulation:
– Introduce social scientists to new ways
of thinking about social problems
– Enable researchers to to run
simulations, visualise and analyse
results, publish for future discovery,
sharing
– Facilitate development and sharing of
social simulation resources
http://www.geog.leeds.ac.uk/projects/neiss/about.php
2031
2031
2001
Transport
2015
Traffic Intensity *
0
0.6
0.1
0.7
0.2
0.8
0.3
0.9
0.4
1.0
0.5
* Traffic Intensity=Traffic load/Road capacity
Digital ethnography
http://www.mrl.nott.ac.uk/research/projects/dress/
Text mining: document analysis
 Identification of
conceptually similar
documents using most
commonly occurring terms
and words in the source
document
 Highlighting selected
semantic information
within the document
 Selecting terms
according to importance
and using them to browse
documents
www.nactem.ac.uk/assist/
Text mining: sentiment analysis
Subjective Sentiment
Automatic estimation of the
opinion of the writer
regarding a fact or an event
 Negative opinion
 Neutral opinion
 Positive opinion
www.nactem.ac.uk/assist/
Web mining
Using
website
links to
map
political
blog
community
structure.
Adamic and Glance, 2004
Web mining
Using
Facebook as a
source of
social data:
‘webnography’
http://www.thefacebookproject.com/
Web mining in real time
‘Tweet-o-Meter’ – an
example of how we can
capture, visualise and
extract patterns from
mobile communications.
http://www.casa.ucl.ac.uk/tom/
Visualisation
Social
data and
google
maps
mash-up.
http://www.maptube.org/
Visualisation
Using survey tools and
maptube to ‘crowdsource’
opinions.
Sharing methods
Methodbox users
include NHS Public
Health analysts and
Department of Health
Public Health
Observatory analysts,
social scientists and
epidemiologists
Virtual Research Environments
A collaboration
space for social
scientists.
A means to share
scientific resources
across a diverse
community.
www.ourspaces.net
What e-Research means for you
■ Easy-to-use, ‘shrink-wrapped’ tools and services:
 DRS, NeISS, etc.
■ Build your own:
 Create new datasets by mashing up existing data.
 Create ‘workflows’ to discover, extract and analyse
data.
■ Engage in new forms of scholarly
communications:
 Make data and methods freely available so that others
can re-use them.
Creating a research ‘workflow’
Automating extraction
and analysis of
messages in study of
‘social dynamics’ in an
open source software
community.
www.myexperiment.org
New forms of scholarly communications
LogBook
Images
Presentations
Software
Literature
Compute resource
His friends and colleagues
Backup and Archive
Data (files, spreadsheets)
Summary of e-Research
■ Application of advanced digital methods and
tools in all parts of the research lifecycle:
 Locate and access research resources.
 Discover, access, integrate and analyse digital data
on a hitherto unrealisable scale.
 Facilitate sharing and collaboration.
■ Enhanced research practice:
 Reduce ‘time to discovery’, improve robustness,
enable research advances that would not otherwise
be possible.
Where to find out more
http://www.eresearchsouth.ac.uk/uk-esocial-science
http://www.methods.manchester.ac.uk/meth
ods/eresearch/index.shtml
Cyberinfrastructure Vision for 21st Century
Discovery
http://www.nsf.gov/pubs/2007/nsf0728/ind
ex.jsp
Thanks to
■
■
■
■
■
■
■
■
■
■
■
■
Peter Halfpenny
Dave De Roure
Marina Jirotka
Anne Trefethen
Carole Goble
Mark Birkin
Andy Crabtree
Sophia Ananiadou
Andy Hudson-Smith
Richard Milton
Meik Poschen
Alex Voss
Questions
[email protected]
http:www.merc.ac.uk