20140902 linda workshop semantincs2014 - bringing lod to smes
TRANSCRIPT
LinDA-project.eu
Bringing LOD to SME – The case of the LinDA pilots business intelligence, environmental sector, media industry
Salvatore Virtuoso [email protected]
Senior Project Manager - PIKSEL
LinDA workshop at
SEMANTICS 2014
+Linda Pilots Role
Setup
• Define Scenarios
• Identify Public & Private datasets needed, Algorythms
• Define Consumption metrics
Execution
• Dataset renovation
• Linked Datasets
• Consumption apps
Evaluation
• User Acceptance
• Workbench assessment
2
To evaluate the efficiency and the business potential
of the LinDA workbench in typical SME setups
Linda Pilot phases:
23/1/2015LinDa workshop @ SEMANTICS 2014, Leipzig
Sept
2014
+The goals of LinDA pilots
2/9/2014LinDa workshop @ SEMANTICS 2014, Leipzig
3
network of business
intelligence management consultants
SME focussing on the application of ICT technologies to environmental
sector issues
Regional broadcaster, with
a mission of reporting about
daily events, political debate
and sports
Bu
sin
ess
Inte
llig
en
ce
to demonstrate innovative and gainful intelligence-based consulting to
customers and strategy planning through the LinDA transformation and analytic tools.
En
vir
on
men
tal
Secto
r To utilize the LinDA solutions for the efficient management and analysis of the
Italian Regions Environmental data M
ed
ia I
nd
ustr
y
to demonstrate the potential of the LinDA workbench to provide advanced tools for investigative journalism
+Business Intelligence Pilot
Two scenarios identified:
Scenario 1 - Identifying actions for the
communication strategy of a client,
operating in the pharma sector
Scenario 2 - Press monitoring reports
and consultation services for a telecom
client
23/1/2015LinDa workshop @ SEMANTICS 2014, Leipzig
4
+Scenario 1 - Pharmaceutical Sector
Issues to be examined:
the OTC liberalisation and how it has affected the:
government [e.g. gov incomes from OTC & drugs, taxes]
industry [e.g. revenues & investments, sales of OTC drugs]
population [e.g. OTC sales, behaviour]
the trends of the OTC prices, compared to the economical status
the trends in the healthcare expenditures
Note: the pilot will be conducted towards the Greek market
2/9/2014LinDa workshop @ SEMANTICS 2014, Leipzig
5
Aim: to assess the positioning of Pharma clients against the issue of
drugs prices liberalisation, mainly for OTC Medicines
+Existing DatasetsExamples of existing public datasets that the pilot intends to use / interlink
European Core Health Indicators (ECHI)
All EU countries 2003-2011, csv format
Demographics and Socioeconomic indicators (e.g. population by education, population below poverty
line)
Healthcare expenditure (e.g. percentage of GDP, percentage of population covered by health
insurance)
Self-reported use of non-prescribed medicines by sex, age and educational attainment level
World Bank Datasets
All countries, 1994-2013, csv format
Out-of-pocket health expenditure (% of private expenditure on health)
Healthcare expenditure (e.g. percentage of GDP) – years 2000 and 2012
OECD Datasets
1980-2013
Health expenditure and financing
Social spending
23/1/20152nd Plenary Meeting (Bonn,DE) | Business Intelligence Analytics Pilot (CP)
6
+Examples of private datasets to be created
Association of the European Self-Medication Industry
All EU countries
2011-2013, online data
Total pharmaceutical market
Non-prescription medicines market
Total self-medication market
The Liberalization of the Retail Market of Non-Prescription Medicines
Pdf file with data regarding liberalization of OTC per country
OTC Distribution in Europe: Meeting the New Challenges - New Expanded 2014 Edition (*)
The Rising Tide of OTC in Europe (*)
Central and Eastern Europe OTC Drugs Industry Outlook to 2017 (*)
(*) Non-open dataset, available on a fee basis
23/1/20152nd Plenary Meeting (Bonn,DE) | Business Intelligence Analytics Pilot (CP)
7
+Scenario 2 – Telecom Sector
Issues to be examined:
Sentiment analysis (e.g. Negative / Positive publicity) on news portals, blog posts, social media, etc
Networking & Electromagnetic fields (EMF) issues
Other relevant: Regulatory, Financial, Marketing, Corporate Social Responsibility (CSR)
Business analysis:
Indication and forecasting regarding the reactions upon an installation of a new antenna
Analysis reports based on comments from residents in specific areas – impact on company’s revenues
Analysis of source types with more negative comments
Impact on health (based on location of antennas and health incidents, examine existence of dependencies by including also areas without antennas)
2/9/2014LinDa workshop @ SEMANTICS 2014, Leipzig
8
Aim: to identify and assess the operating environment, using press
monitoring procedures on specific telecoms parameters that might
concern and affect clients’ agenda
+TelecomPilot
Record of antennas with
geolocation data in Greece
Geolocation data from schools,
hospitals etc.
Geodata.gov.gr
Telecoms revenues, market
share, etc
Monitoring of publications in
media relevant to population’s
comments on antennas
List of media
2/9/2014LinDa workshop @ SEMANTICS 2014, Leipzig
9
Existing dataset examples Private dataset to be created
+Environmental Management Pilot
Specific case study: the area comprising the municipalities of Acerra, Nola and Marigliano in Campania, Italy.
The region has recently experienced increasing deaths caused by cancer and other diseases that exceeds the Italian national average.
Objective: To analyse the impact on the health of the residents based on the variance on specific environmental parameters and pollution indicators
Process: examination of the frequency and the type of occurrence of diseases in specific geographical areas within Italy with the environmental conditions and indicators for pollution in the area
23/1/2015LinDa workshop @ SEMANTICS 2014, Leipzig
10
+Environmental Mgmt Scenario
Generic datasets (ISTAT)
Waste collection, noise, vehicle rate, air quality
Population and Households: (Demographics, mortality, projections)
Health statitics (health conditions, incidence, prevalence and mortality)
Cancer registries
Environmental datasets
Waters, etc.
International datasets
pollution, emissions, wastes, policies per country
11
Examples of dataset to be used:
2/9/2014LinDa workshop @ SEMANTICS 2014, Leipzig
+Media Industry Pilot
Two scenarios identified:
Scenario A – Adding data mashups and
analytics to the toolbox of investigative
journalists (and potentially, citizens-
reporters)
Scenario B – Tapping into the
knowledge reservoir of Post Production
Scripts (media-rich and detailed
storyboards supporting programme
broadcasting)
2/9/2014LinDa workshop @ SEMANTICS 2014, Leipzig
12
+Scenario – Media Industry A
Topics of interest
Advanced analysis over data from multiple sources,
Interconnection of georeferenced information to other
parameters as well as international datasets,
Preparation of infographics and creation of
higher‐end multimedia communication tools, and
Integration of LINDA tools in the daily routine of
journalistic work – as well as the supporting
(standard) IT infrastructure
2/9/2014LinDa workshop @ SEMANTICS 2014, Leipzig
13
Aim: to support investigative journalists engaged in evidence search,
collection and commenting/reporting.
+Scenario – Media Industry B
Topics of interest
Migration of existing Post Production Scripts in rdf
format
Advanced management of multimedia content items
(esp. music score and video clips)
Collection of HTML & plain text from WWW and staff
authoring, and
Integration of LINDA tools in the daily routine of
journalistic work – as well as the supporting
(standard) IT infrastructure
2/9/2014LinDa workshop @ SEMANTICS 2014, Leipzig
14
Aim: to facilitate the retrieval of information stored in previously recorded
audio/video clips through mining of Post Production Scripts.
+Additional business casesExamples of LINDA tools application to the investigative journalism scenario (#5)
of potential interest for SMEs
International comparison of the quality of universities worldwide
An orientation service for students wanting to move abroad
Including location related aspects (logistics, job opportunities, individual grants etc.)
Reconstruction of where aid money goes to
Possible clients: public/private donors and international organisations (e.g. OECD,
World Bank)
Sensitive issue but with huge political impact
A Dbpedia of patents worldwide
Integrating text and visual information
Potentially extended to public domain solutions
Documented interest by a local SME in the patent attorney business
2/9/2014LinDa workshop @ SEMANTICS 2014, Leipzig
15
+Conclusions
Many linked data projects are promoted by technology enthusiasts (or deliberate experimenters) keen to explore and use their own approach rather than carefully selecting the best tool for the job.
In other words, linked data projects have not, as yet, been based on business cases.
The benefits of linked data are most often assumed or implied by the implementers; there is little measurement of them.
Many projects using linked data struggle to express their benefits (although it may be too early for most of them).
There are far more expressions of technical benefit (for example it is easier to work across systems) than business benefit (say, better service quality), although one might lead to the other.
There is a lack of cost-benefit analysis for linked data projects and a lack of comparison with other technical approaches.
From: http://repository.jisc.ac.uk/559/1/JISC_Linked_Data_Review_Oct2011.pdf
2/9/2014LinDa workshop @ SEMANTICS 2014, Leipzig
16