FUTURE POLICY MODELLING (FUPOL) MAY 2012 · 2014-03-24 · FUTURE POLICY MODELLING (FUPOL) MAY 2012

www.fupol.eu www.facebook.com/fupol

FUTURE POLICY MODELLING (FUPOL) MAY 2012

http://www.fupol.eu/

http://www.facebook.com/fupol


Overview

Current Status

Q&A, discussion




FP7 Research Project 2011 – 2015 with a budget of 9,0 Mio EUR

Objective: An integrated governance model for cities based on social media policy simulation and related IT tools

Good balance of international ICT companies, universities and political institutions

17 Partners from Austria, Croatia, China, Cyprus, France, Germany, Italy, Latvia,

Romania, Spain, UK

FUPOL: Facts and Figures




Starting point: Urbanization as a 21st century megatrend

1800: 2 percent of the world was urban.

1900: 14 percent urban.

2008, the world crossed the 50 percent threshold.

2050, it is expected that the world will be 80 percent urban.

0

10

20

30

40

50

60

70

80

90

1800 1850 1900 1950 2008 2050




Resulting urban policy challenges and topics as a project focus

Land Use Policy

Urban Sprawl Efficient land use

Housing Policy

Affordable Housing Slums

Integration Policy

Segregation Exclusion

Environment

Pollution Low Carbon

2011/2012

2012/2013 (currently being evaluated)

Tourism Policy

Sustainable Tourism

Budget Policy

Participative Budgeting Impact Forecast




Training Material and Guidelines

Urban Policies Simulation Recommendation Policy Models

Tra

inin

g

Mate

rial and G

uid

elin

es

Tra

inin

g M

ate

rial and G

uid

elin

es

FUPOL

IT TOOLBOX

Objective: A comprehensive solution to support policy design




Scenarios, Simulation

Stakeholder Involvement

„Hot“ Topic sensing

Application Scope: Policy Design Process (simplified)

Automatically collect, analyze and interpret opinions expressed on a large scale from the Internet Involve Stakeholders through Multichannel social computing and crowd sourcing to change the way politicians communicate with citizens Simulate the effects of policies and laws and to assist governments in the whole policy design process and get feedback again




Automatically collect and analyze opinions from social media

What is the current concern of citizens „Hot Topics“

Social Media offer new opportunities complementing „traditional“ channels

such as surveys, face to face meetings




Backend to existing social media presence of a city

Hot Topic Sensing complements existing social media presence

Backend to automatically analyze messages and group them

FUPOL method based on statistical algorithms (LDA) and machine learning –

works with any European language




Select priority topic and initiate public deliberation

Decision maker selects priority topic

Initiates public discussion through various social media

„Single Window“ through central dashboard

Malibu Square Renovation We intend to renovate

Malibu square,. What should be done ? Send us your ideas !




Location based feedback from e-citizens

Rating and selection of best ideas with idea management system

Improves acceptance by e-citizen

© Meipi Prato

Crowdsourcing : Get feedback and ideas from e-citizen




Simulation: Anticipate the impact of policies

More transparency for decision makers, administration and citizen

Land Use Schiphol 2012 2030 after implementing a policy to attract industrial investment

Source: RIKS

Simulate and optimize scenarios




FUPOL-Knowledge database

Identify and visualize urban data (statistics)

Identify related studies

Find best practice

Evaluate scenarios

Find and visualize knowledge to evaluate scenarios



http://www.web-site-scripts.com/wiki-software


User driven innovation means

- thorough understanding of requirements and current challenges in city administrations drives R & D

- close contact to pilot cities and potential users in other cities through F2F meetings and workshops

- Major Cities of Europe (www.majorcities.eu) in the consortium

Implementation strategy (I) User driven innovation



http://www.majorcities.eu/






t

• Incremental delivery of working software

• Continuous multi-level integration

• High level of test automation

t

• Iterative approach

• Short controlling cycles (monthly or 2 weeks, tbc)

• Timeboxed

t

$ $

• Output planning based on user stories

• Priority of work based on „value added“ (business value, risk mitigation, knowledge gained, …)

• Valuable work has a higher priority than less valuable work

Implementation strategy (II) Agile based on SCRUM




Continous release of results

Research & Development

Testing / Pilots

Exploitation

2012 2013 2014 2015 2016

Rel 0.1 July 2012




Overview

Current Status

Q&A, discussion




Overall progressing as planned, software development 3 months ahead of schedule because of the agile methodology

Policy design

Urban policy domains analyzed and prioritized

Analysis of social media in urban politics finalized

Study of data protection issues completed

Policy simulation

Some policy models defined as Fuzzy Cognitive Maps & causal models

Land use SW state of the art analyzed

Land use protoype in NETLOGO implemented

Status and results




Software development

Requirements collected, IT-architecture defined

Enterprise GIS set-up

Software development started in March 2012

Visualization: State of the art study concluded

Hot Topic Sensing & Topic summarization:

State of the Art analysis concluded

Design initiated

Status and results




Land-use, housing, segregation selected for 2011/2012 and policy design

processes, data and indicators of selected domains studied

Domain selection criteria

1. Linkage of the domain to a major urban challenge - this refers to the linkage of the domain to a major urban challenge such as urban sprawl, slums, etc.

2. Availability of statistical data - this is important, because without data simulation is very difficult.

3. Pilot city priority - the priorities assigned by the pilot cities were important in the overall ranking.

4. Overall priority – finally the overall priorities (European and worldwide).

Policy domain selection




Some policy models defined as Fuzzy Cognitive Maps & Causal Models

Building models, which are applicable across different countries and cultures could be a challenge

Policy Models and simulation




Not in widespread use – benefits are not recognized

Weak point: Social media strategy

Weak point: Social media guidelines for employees

Some politicians are afraid of social media, because they cannot

be controlled

City administrations fear additional workload and „PR-disasters“ resulting

from bad communication between civil servants and citizens

Social medias usage model developed as part of the FUPOL policy guidelines

Analysis social media in urban politics, status and best practice




Proposed Model Social Media in Urban Politics




Social Media data (postings etc) can be processed, because the individual has unambiguously given his or her consent

However it has been decided to anonymize them for further processing

Data Protection & Social Media




IT-Platform: Core Platform Architecture decided

FUPOL uses a service oriented architecture (SOA)

Modules communicate over the ESB using standard protocols (WFS/WCS/WMS/SPARQL) if possible and proprietary APIs (FUPOL Core API over SOAP) if no standards apply (i.e. campaign data, user activity,…)




IT-Platform: „Campaign“ in the policy design process

„Campaigns“ maintains the sequence of actions and their results by the user (facilitator) in the policy design process and its results

time

The facilitator adds data from the data/ knowledge base

The facilitator collects social media data using the crawler

The facilitator creates topic trees using the topicalization tool

The facilitator starts the simulation and adds the results

The facilitator writes his final report and closes the campaign. Some generated data might be added to the data base for later reuse

The facilitator creates a new campaign




IT-Platform: Enterprise GIS set-up

Stores all data

Maps names to coordinates

Used by the sysop

Maps INSPIRE data

Provides WFS/WCS/WMS




State of the art and social media monitoring tools evaluated

Data sourcing: How to reduce them to the relevant postings ?

Conventional keyword search available in many social media

monitoring tools is unsuitable

Predefined sources (certain blogs, Twitter hashtags..) is an option

Better option is to use geolocalized data of postings from social networks

as a filter (if available)

Hot topic sensing




Automatic Topic Summarization: State of the art study finalized

We can determine the ‘representative’ sentences in an input with good accuracy. Depending on the domain, we can determine if sentences evoke positive or negative feelings towards a topic

There is some ability to model the structure of a conversation in an online forum and understand the progression of an argument/discussion

There is very little work on true abstractive summarization; most results are simply extracted phrases/sentences from an input

Difficulties when the data is “noisy” (i.e. poor spelling, use of slang, etc.)

Summary

Documents belonging to the same topic




Automatic Topic summarization: Methods under investigation

“Traditional” summaries of citizen comments:

Processed extracted sentences that are representative of citizens’ comments strung together into a shorter representation

“Report”-style summaries of comments: Generated text that describes important discussed issues along with some analysis and example comments

Ranked “topic”-word and -sentence lists of comment themes: Lists of discovered words and sentences that are representative of citizens’ comments

“Opinion” summaries: Summaries that highlight the diverging opinions or sentiment (positive/negative) on different topics/projects




Upcoming….

June 2012 Major Cities of Europe Conference 2012 in Vienna:

Important event – user dialogue through a workshop

July 2012 First release of software core platform

August 2012 Second release of software core platform

October 2012 First set of policy models and other documents (e.g. pilot planning)




Take Part !

Follow us:

Website (www.fupol.eu)

Facebook (www.facebook.com/fupol)

LinkedIn

Take part: Test and evaluate the FUPOL results

Establish relations to exchange views

Dr. Peter Sonntagbauer

Project Director FUPOL cellent AG Tel.: +43 (1) 532 78 03-0 Fax: +43 (1) 532 78 03-33 Mobile: +43 (676) 42 84 724 e-mail [email protected]

www.facebook.com/peter.sonntagbauer

at.linkedin.com/in/petersonntagbauer

Sonntagbauer, Peter




THANK YOU FOR YOUR ATTENTION

