Create your own awesome maps

Even on the go

with our free apps for iPhone, iPad and Android

Get Started

Already have an account?
Log In

JISC CETIS 2010 PARALLEL SESSION 2 LINKED ENTERPRISE DATA by Mind Map: JISC CETIS 2010
PARALLEL SESSION 2
LINKED ENTERPRISE DATA
0.0 stars - reviews range from 0 to 5

JISC CETIS 2010 PARALLEL SESSION 2 LINKED ENTERPRISE DATA

Why

Data Syntax = Mainly Works

Transport = Nearly There

Semantics = Problem

Reconcile Local and Global

How does this fit with a Single Point of Truth?, Aliases

Drivers

Business Intelligence, 1. Finance, 2. Student Data, 3. Information Management, 4. Performance Measurement

National Data Flows, HESA, UCAS, HEFCE, UCISA

BI Requirements, Up to date, Reliable, Data integrity, All major information sources included, Drill down and roll up, Easy to add new sources (internal and external), Allows What If? questions

Panel

PMI adopting linked data for BI

data cleanup, tolerates dirty data

disambiguation

facilitates mashups

handles disparate datasets

What architectures and patterns are best for enterprise use

Practical experiences

Tools

Business Case

1. Disambiguation

Data audit

Speaking a common language

Make data more precise

2. Centralisation

Answering questions from across this institution

Cost of getting/keeping data clean?

3. Inferencing

Others?

Big data?

Data cleanup

What is It?

Triples

Making relationships meaningful

Subject/Predicate/Object, S:O

Disambiguate web links

Bookmarks

http://groups.diigo.com/group/linked-data

http://groups.diigo.com/group/ontology-examples

http://groups.diigo.com/group/education-ontologies

ResearchRevealed

Bristol JISC Project Exploit data silos related to research. Looks like it would be a replacement for PURE .. or could be built on PURE. Add a linked data layer to Equella?

Browse and Discovery across Research Silos

Data Sources

Partners

Publishers

Funding Councils

Finance

Central System

Personal Website

Web of Data

RDF

RDFa, Add attributes to (X)HTML, 1.1, HTML5, Simple vocabularies

SPARQL

Semantic Web/Linked Data

Open Graph?

Steps

Publishing, unfamiliar format, content negotiation, exotic redirects

Governance?

Making Data Work

New technologies can't do everything

It's about solving problems not technology

Examples

Powerset

www.tripit.com

Siri

True Knowledge

metaweb/freebase

Disambiguation on the fly

Scenarios from Southampton

www.semhe.org

Subject related questions

Expert/Research Data

Tutoring/Attendance Monitoring

Building/Room Location

'VLE' as Mashup

Modules

Resources

Assignments

Bookmarks

Emails

Enterprise Linked Data

Attention

MIT CIO 201 symposium, Enterprise 3.0

Open Society Foundation Open Data Study, Top Level Mandate, Middle Layer Bureaucrats, Civil Society, Small number of civic hackers

5 Stars of Linked Data

* available on web

** structured

*** non proprietary formats

**** urls to identify

***** link to other people's data to provide context

Architecture

Community Based Models

Social Semantic Enterprise, Employee focus, www.liferay.com, http://sig.ma, RDFiser!

Semantic Enterprise

Process

Raw Data

Graph-Structured Data

Integrated Data

Actionable Knowledge

Semantics

Ontology

Data Model

Taxonomy

Tags

Values

Open LOD

LOD Cloud, entity consolidation, vocabulary alignment

Identify LOD resources, estimate semantics, match semantics