際際滷

際際滷Share a Scribd company logo

Sense4us PACITA event presentation

Download as PPTX, PDF
SENSE4US project
SENSE4US project

The document discusses the Sense4us toolkit which aims to help policymakers make more informed decisions by analyzing social media, open data sources, and modeling policy problems. It describes the different components of the Sense4us toolkit, including tools for topic analysis of social media, sentiment analysis, cognitive mapping of policy issues, and simulation of policy options. The document also discusses challenges in using social media and open data to inform policymaking and demonstrates how Sense4us addresses these challenges through various case studies and examples.

1 of 56
Download to read offline
Policy Making in a Complex World: The Opportunities and Risks Presented by New Technologies 2nd European TA Conference (PACITA) Thursday, February 26th 2015 1
Session Schedule General Introduction Analysis of social media to inform policy making Modelling and simulation of public policy problems Case study demonstration of current toolkit Technology assessment: evaluation of design assumptions 2
3 Introduction Steve Taylor, IT Innovation
Contents Sense4us overview Project objectives and how we address them Components of Sense4us Example of how Sense4us helps its users 4
Information Management Challenges faced by Policy Makers Too much information Need to sift through the deluge of information available today Potentially untapped resources of relevant information online Open data, social networks, forums, local blogs, etc Unknown unknowns There may be relevant information policy makers are not aware of The full impact of policy is not always obvious when it is created Unexpected outcomes & affected citizens 5 and How Sense4us Meets them
Sense4us Toolkit Overview 6 Theme Analysis Document Summary & Keywords Related Concepts & Themes Linked Open Data Search User Social Media Search Social Media Comments Sentiment Analysis Opinions & Sentiments Model Builder Policy Model Policy Simulator Simulation Results Policy Document Sense4us tool Data user can see Search Terms & Keywords UI + Infrastructure
Sense4us: An End User Partners Perspective 7
Case Study 8 Navitus Bay Wind Farm Planning Application for wind turbines in the sea off the south coast of the UK Highly controversial
(Real) Case Study & (Fictitious) Examples of Data 9 Theme Analysis Location proposed energy benefits Current analysis of pros and cons Other wind turbine installations (& success or not) Strategies to achieve planning success Concerns (impact on tourism & nature) Benefits (green energy) Linked Open Data Search Search Terms & Keywords User Social Media Search Sentiment Analysis Strong negative sentiment from local residents Positive reaction from green campaigners Model Builder Relationship between turbine location and tourism revenue Policy Simulator If turbines are 10 miles from the coastline, tourism revenue will be hit by 20% Planning Application Document Key Factors for model: Turbine size Turbine location Tourism Revenue Unemployment Reduction in local areas fossil fuel usage Resistance from local communities Disgrace! Good source of energy It will ruin the view Will my fuel bills be cheaper?
Project Status Month 17 of 36 Initial engagement with end users complete Initial prototype complete Plans to demonstrate prototype to end users in next months Gather feedback Update design Update prototype 10
11 PART 1 Using Social Media To Inform Policy Making: To whom are we listening? Miriam Fernandez, Open University Harith Alani, Open University
Introduction Social media A revolutionary opportunity for governments to learn about the citizens and to engage with them more effectively but, When using social media to inform Policy Making To whom are we listening? What are the key policy subject areas of discussion? How do users feel about those policy subject areas? 12
Who is talking about policy in social media? (I) 13 Three main lines of work Statistics about the citizens participation on ePlatforms Not social media participation Statistics about users participating in social media Not narrowed to eParticipation / political discussions Studies of political discussions in social media In the context of political events (elections, revolutions, etc.), not focused on relevant topics for policy making Participation in policy making Participation in social media
Who is talking about policy in social media? (II) Goal: Study the characteristics of those users discussing policy in social media and the key topics and sentiment of their discussions Data Policy topics: 76 policy topics collected from 16 PMs, all from different institutions in Germany Generic topics (such as women) were filtered to avoid collecting noisy data -> 42 remaining topics Betreuungsgeld (Care Benefit ) Bildungspolitik (Education Policy) B端rgerrechte (Civil Rights) Topics were monitored in Twitter for a week. The developed algorithms are designed to track discussion dynamics over time 14 Fernandez, M., Wandhoefer, T., Allen, B., Cano, E., Alani, H. Using Social Media To Inform Policy Making: To whom are we listening? European Conference of Social Media (ECSM 2014) Init Date Final Date Num Posts Num Uses 04/01/2014 12/01/2014 17,790 8,296
Who is talking about policy in social media? (III) Top Contributors Less than 6% of the users are responsible of 36% of the generated content 73.4% of top-contributors are NOT citizens but news agencies and other organisations The Average Contributor Is more active, popular and engaged than the average Twitter User Geographic Distribution of Users Higher concentration of users occurs in constituencies of high population density Users engaged in social media conversations around policy topics tend to be geographically concentrated in the same regions than users engaged in eParticipation platforms 15 (a) Distribution of eParticipation projects (b) Distribution of Twitter users Figure 1: (a) Distribution of eParticipation projects in Germany (http://www.politik.de/politik- de/projekte_entdecken/beteiligungskarte) (b) Distribution of Twiter users: yellow are locations with less than 10 users, pink are locations with 10 to 50 users, red are locations with more than 50 users
What are they topics of discussion and the sentiment around those topics? Topic Distribution Few topics are extensively discussed during the analysed period Privacy, Network Policy, Minimum Wage, Copyright, etc. The majority of topics are underrepresented Sentiment Distribution Top Negative topics Genetic Engineering, Immigration, Referendum, European policy and donations to political parties Top Controversial topics Privacy, Fracking and Domestic Policy (High percentage of positive and negative posts) 16 0 500 1000 1500 2000 2500 3000 3500 4000 privacy networkpolicy minimumwage copyright fracking domesticpolicy genetic resin migrants equality femaleratio rightwing referendum leftwing educationand energypolicy europeanpolicy partydonate socialpolicy speedlimit financialpolicy nosmoking caremoney transportpolicy generational debtbrake environmental npdban nonsmoking sociallyticket
SentiCircle: understanding topics and sentiment around political discussions (I) 17 Helps summarising policy discussions A model for extracting the facts (aspects) of a given topic in a tweet collection Classifies public opinion on these facts (aspects) as in-favour or not-in-favour for studied topic Environment ROI Efficiency Weather Installment Maintenance Funding Noise Renewable Energy In-Favor Not-In-Favor Evidence Representation of the topic Renewable Energy using SentiCircles Saif, H., Fernandez, M., He, H., Alani, H. SentiCircles for Contextual and Conceptual Semantic Sentiment analysis of Twitter. European Semantic Web Conference (ESWC 2014) Saif, H., He, Y., Fernandez, M. and Alani, H. (2014) Adapting Sentiment Lexicons using Contextual Semantics for Sentiment Analysis of Twitter, Workshop: Semantic Sentiment Analysis, Crete, Greece. BEST PAPER AWARD! More Less
SentiCirle: understanding topics and sentiment around political discussions (II) SentiCircle: Lexical-based sentiment representation model Assigns sentiment to a term by considering its co- occurrence patterns with other terms 18 The radio T-DOC is computed based on the degree of correlation between the two terms The angle is computed based on the prior sentiment of the term, extracted from an existing lexicon
Discussions Understanding who are the users discussing policy in social media and how policy topics are debated could help PMs assessing how the citizens views and opinions should be weighted and considered to inform policy making This research is being incorporated into the Sense4us toolkit Several problems arise when using social media for this purpose: Data is distributed in multiple social platforms More research is needed to understand how representative is the subset of the population discussing policy in social media Social media -> big data issues: volume, variety, velocity and veracity of the data 19
20 PART 2 Modelling and Simulation of Public Policy Problems Sense4us Model Builder and Simulation Tool Aron Larsson, Osama Ibrahim, Anton Talantsev, eGovlab Department of Computer and Systems Sciences (DSV) Stockholm University
Policymaking process model 21 Prescriptive analysis (Impact Assessment): Carried out at the early stages of policy development), which encompasses the forecasting of consequences and prescriptions about which policies should be implemented. Retrospective analysis (Evaluation): Tries to understand the causes and consequences of policies after they have been implemented.
Decision support framework approach 22 Enabling policy analysis and decision evaluation where problem structuring is supported by linked open data, topic analysis and sentiment analysis of social media data. Policy problem structuring and modelling (Causal/cognitive map) Design policy options, consequence assessment, generate options (Simulation) Decision evaluation of policy options (Decision analysis) Sentimentanalysis Evidenceextractionfromopendatasources
Why problem structuring? 23 Searching for the right information Capture a policy makers views about a problem Understanding decision making context and communication of problem understanding Structuring more complex cause-effect relationships Identifying where and how interventions have impact Enabling for decision evaluation of policy options
Objectives Cognitive/causal mapping 24 CO2 emissions Means Subsidy bus tickets - Hmm??
Cognitive/causal mapping 25 CO2 emissions Means Subsidy bus tickets - Car traffic + - Objectives Hmm?? LOD LOD
Cognitive/causal mapping 26 CO2 emissions Means Subsidy bus tickets - Car traffic + Objectives Member of council
Cognitive/causal mapping 27 CO2 emissions Means Subsidy bus tickets - Car traffic + Objectives It cannot cost more than 35 a month Its not the price, it is the frequency of buses Member of council
Cognitive/causal mapping 28 CO2 emissions Means Subsidy bus tickets - Car traffic + Objectives Increase bus frequencies - - Member of council
Cognitive/causal mapping 29 CO2 emissions Means Subsidy bus tickets Car traffic + Objectives Increase bus frequencies - - Member of council Traffic planner
Cognitive/causal mapping 30 CO2 emissions Means Subsidy bus tickets Car traffic + Objectives Increase bus frequencies - - Traffic planner Gas driven buses - Bus company CEO Member of council
Cognitive/causal mapping 31 CO2 emissions Means Subsidy bus tickets Car traffic + Objectives Increase bus frequencies - - Gas driven buses - Town commerce + -20% +5% + Stakeholder Traffic planner Member of council Bus company CEO
Evaluation concept: Impact potency 32 CO2 emissions Means Subsidy bus tickets 2 Car traffic + Objectives Increase bus frequencies 2 - - Gas driven buses 1 - Town commerce + -20% +5% + Stakeholder Bus company CEO Traffic planner Member of council
Evaluation concept: Forward analysis 33 CO2 emissions Means Subsidy bus tickets Car traffic Objectives Increase bus frequencies -0.2 1m -0.5 1m Gas driven buses Town commerce -20% +5% Stakeholder -1.0 0m +1.0 0m +1.2 0m +0.6 4m Traffic planner Bus company CEO Member of council
Forward analysis 34 CO2 emissions Means Subsidy bus tickets Car traffic Objectives Increase bus frequencies -0.2 1m -0.5 1m Gas driven buses Town commerce -20% +5% Stakeholder -1.0 0m +1.0 0m +1.2 0m +0.6 4m 20%, t=0 -4%, t=1 7%, t=4 25%, t=12 -29%, t=12 Traffic planner Bus company CEO Member of council
Group decision and game concepts 35 CO2 emissions Means Subsidy bus tickets Car traffic Objectives Increase bus frequencies -0.2 1m -0.5 1m Gas driven buses Town commerce -20% +5% Stakeholder -1.0 0m +1.0 0m +1.2 0m +0.6 4m 20%, t=0 -4%, t=1 7%, t=4 25%, t=12 -29%, t=12 Traffic planner Bus company CEO Member of council
Group decision and game concepts 36 CO2 emissions Means Subsidy bus tickets Car traffic Objectives Increase bus frequencies -0.2 1m -0.5 1m Gas driven buses Town commerce -20% +5% Stakeholder -1.0 0m +1.0 0m +1.2 0m +0.6 4m 15%, t=0 5%, t=4 15%, t=12 -21%, t=1210%, t=6 Bus company CEO Traffic planner Member of council
To sum up We are building a tool that assist public decision processes through modelling a public policy problem, simulating policy consequences for decision evaluation considering multiple objectives and stakeholders. Visually structure the policy problem for increased problem understanding Show an understanding of the relations between policy instruments (funds, taxes, subsidies, prohibition, etc.) and societal effects. Generate different means or combinations of means controlled by different actors to reach similar targets (scenario generation) If we change this policy instrument, according to what we know, what are the effects (over time) on the factors subject to policy targets? Which stakeholders are affected and how? Will they react and if so what is the effect on the factors? 37
Contact Aron Larsson aron@dsv.su.se 38
39 PART 3 Finding and Analysing Online Data to Support Governmental Decision Making Processes - the case of Sense4us Timo Wandh旦fer, GESIS Max Bashevoy, IT Innovation
Persona UK Decision maker Member of the House of Commons Political interest: renewable energy 40
Use Case 41 Similarities? Dissimilarities? Sense4us Topics Topics Debates in the plenum, draft bills, press releases Sense4us ?
42 PART 4 Assumptions to Artefacts: Understanding the Design Choices Underpinning the Sense4Us Project Somya Joshi, eGovlab Department of Computer and Systems Sciences (DSV) Stockholm University
Objective To summarise where we are To understand how we got here To visualise where we are heading Where does Technology Assessment fit into this? 43
Where we are at Milestones End user engagement first leg (Policy makers needs & requirements) Demo of tool stage one Integration of tools 44
Understanding how we got here Design choices & assumptions Who is this for? What are we hoping to impact? Where is the innovation/ added value? How will end users engage with our tool set? 45
What the end users expressed Requirements Provenance of data Transparency Sentiments & Opinions elicitation Summary & Visualisation of raw data sets Localisation of data Customisability of tool set 46
Assumptions of Technology Impact Transparency in policy making Policy makers who want to take on board citizen opinions, discussions, information via social media resources Relevance & Provenance of data Publishers who want to make data accessible to others as well as increase their own knowledge
Assumptions Impacts & Innovation Streamline & improve quality of linked open data searches Citizen solutions & knowledge will be summarised & analysed along sentiment/semantic lines Integration of problem structuring; support for impact assessment; preference elicitation when there are conflicting goals 48
What we hope to achieve Co-evolution of tool set in line with end user feedback Integration of problem structuring and policy analysis tools Extend & deepen knowledge via Linked Open Data Identification of important stakeholders/ actors Reduced cognitive loads on policy makers: Making sense of the noise (reducing complexity) 49
The road ahead Technological integration of the various components within the tool set User interface how will end users visualise, make sense & engage with our tool? Transparency in design (no black boxes), data relevance / provenance (trust) and rich results 50
Gov 2.0: Enhanced policy decision support Social media and linked open data as a rich source of opinions, preferences, knowledge, that will be harnessed Simulation and modeling of policy alternatives & impacts 51 Where we envision were heading
The contextual landscape The vending machine model of Governance: where the citizens only engage with shaking up the system when it doesnt work To Governance as Platform Where the platform is a metaphor for multi-layer decision making in complex, evolving environments 52
Technology Assessment & Political Myths Are we designing a Political construct or a Technological artefact? Participatory Design within Policy context? Stage two of our end user engagement will further test this concept of designing a tool set in line with end user feedback Collaborative approaches have been argued to secure legitimacy . What are the anticipated risks and apprehensions on the part of end users within the Sense4us project? E.g. Political vs. Scientific Fact
Technology Assessment of the Sense4Us Toolkit 54
Discussion on our proposed approach How does one demonstrate new concepts to end users? How does one integrate that feedback into future iterations of the design? Our proposed approach is to demonstrate the various components using examples that are relevant and easy to understand To learn if this is understandable, useful, relevant and what they would like to see done differently 55
THANK YOU! 56

More Related Content

Sense4us PACITA event presentation

  • 1. Policy Making in a Complex World: The Opportunities and Risks Presented by New Technologies 2nd European TA Conference (PACITA) Thursday, February 26th 2015 1
  • 2. Session Schedule General Introduction Analysis of social media to inform policy making Modelling and simulation of public policy problems Case study demonstration of current toolkit Technology assessment: evaluation of design assumptions 2
  • 4. Contents Sense4us overview Project objectives and how we address them Components of Sense4us Example of how Sense4us helps its users 4
  • 5. Information Management Challenges faced by Policy Makers Too much information Need to sift through the deluge of information available today Potentially untapped resources of relevant information online Open data, social networks, forums, local blogs, etc Unknown unknowns There may be relevant information policy makers are not aware of The full impact of policy is not always obvious when it is created Unexpected outcomes & affected citizens 5 and How Sense4us Meets them
  • 6. Sense4us Toolkit Overview 6 Theme Analysis Document Summary & Keywords Related Concepts & Themes Linked Open Data Search User Social Media Search Social Media Comments Sentiment Analysis Opinions & Sentiments Model Builder Policy Model Policy Simulator Simulation Results Policy Document Sense4us tool Data user can see Search Terms & Keywords UI + Infrastructure
  • 7. Sense4us: An End User Partners Perspective 7
  • 8. Case Study 8 Navitus Bay Wind Farm Planning Application for wind turbines in the sea off the south coast of the UK Highly controversial
  • 9. (Real) Case Study & (Fictitious) Examples of Data 9 Theme Analysis Location proposed energy benefits Current analysis of pros and cons Other wind turbine installations (& success or not) Strategies to achieve planning success Concerns (impact on tourism & nature) Benefits (green energy) Linked Open Data Search Search Terms & Keywords User Social Media Search Sentiment Analysis Strong negative sentiment from local residents Positive reaction from green campaigners Model Builder Relationship between turbine location and tourism revenue Policy Simulator If turbines are 10 miles from the coastline, tourism revenue will be hit by 20% Planning Application Document Key Factors for model: Turbine size Turbine location Tourism Revenue Unemployment Reduction in local areas fossil fuel usage Resistance from local communities Disgrace! Good source of energy It will ruin the view Will my fuel bills be cheaper?
  • 10. Project Status Month 17 of 36 Initial engagement with end users complete Initial prototype complete Plans to demonstrate prototype to end users in next months Gather feedback Update design Update prototype 10
  • 11. 11 PART 1 Using Social Media To Inform Policy Making: To whom are we listening? Miriam Fernandez, Open University Harith Alani, Open University
  • 12. Introduction Social media A revolutionary opportunity for governments to learn about the citizens and to engage with them more effectively but, When using social media to inform Policy Making To whom are we listening? What are the key policy subject areas of discussion? How do users feel about those policy subject areas? 12
  • 13. Who is talking about policy in social media? (I) 13 Three main lines of work Statistics about the citizens participation on ePlatforms Not social media participation Statistics about users participating in social media Not narrowed to eParticipation / political discussions Studies of political discussions in social media In the context of political events (elections, revolutions, etc.), not focused on relevant topics for policy making Participation in policy making Participation in social media
  • 14. Who is talking about policy in social media? (II) Goal: Study the characteristics of those users discussing policy in social media and the key topics and sentiment of their discussions Data Policy topics: 76 policy topics collected from 16 PMs, all from different institutions in Germany Generic topics (such as women) were filtered to avoid collecting noisy data -> 42 remaining topics Betreuungsgeld (Care Benefit ) Bildungspolitik (Education Policy) B端rgerrechte (Civil Rights) Topics were monitored in Twitter for a week. The developed algorithms are designed to track discussion dynamics over time 14 Fernandez, M., Wandhoefer, T., Allen, B., Cano, E., Alani, H. Using Social Media To Inform Policy Making: To whom are we listening? European Conference of Social Media (ECSM 2014) Init Date Final Date Num Posts Num Uses 04/01/2014 12/01/2014 17,790 8,296
  • 15. Who is talking about policy in social media? (III) Top Contributors Less than 6% of the users are responsible of 36% of the generated content 73.4% of top-contributors are NOT citizens but news agencies and other organisations The Average Contributor Is more active, popular and engaged than the average Twitter User Geographic Distribution of Users Higher concentration of users occurs in constituencies of high population density Users engaged in social media conversations around policy topics tend to be geographically concentrated in the same regions than users engaged in eParticipation platforms 15 (a) Distribution of eParticipation projects (b) Distribution of Twitter users Figure 1: (a) Distribution of eParticipation projects in Germany (http://www.politik.de/politik- de/projekte_entdecken/beteiligungskarte) (b) Distribution of Twiter users: yellow are locations with less than 10 users, pink are locations with 10 to 50 users, red are locations with more than 50 users
  • 16. What are they topics of discussion and the sentiment around those topics? Topic Distribution Few topics are extensively discussed during the analysed period Privacy, Network Policy, Minimum Wage, Copyright, etc. The majority of topics are underrepresented Sentiment Distribution Top Negative topics Genetic Engineering, Immigration, Referendum, European policy and donations to political parties Top Controversial topics Privacy, Fracking and Domestic Policy (High percentage of positive and negative posts) 16 0 500 1000 1500 2000 2500 3000 3500 4000 privacy networkpolicy minimumwage copyright fracking domesticpolicy genetic resin migrants equality femaleratio rightwing referendum leftwing educationand energypolicy europeanpolicy partydonate socialpolicy speedlimit financialpolicy nosmoking caremoney transportpolicy generational debtbrake environmental npdban nonsmoking sociallyticket
  • 17. SentiCircle: understanding topics and sentiment around political discussions (I) 17 Helps summarising policy discussions A model for extracting the facts (aspects) of a given topic in a tweet collection Classifies public opinion on these facts (aspects) as in-favour or not-in-favour for studied topic Environment ROI Efficiency Weather Installment Maintenance Funding Noise Renewable Energy In-Favor Not-In-Favor Evidence Representation of the topic Renewable Energy using SentiCircles Saif, H., Fernandez, M., He, H., Alani, H. SentiCircles for Contextual and Conceptual Semantic Sentiment analysis of Twitter. European Semantic Web Conference (ESWC 2014) Saif, H., He, Y., Fernandez, M. and Alani, H. (2014) Adapting Sentiment Lexicons using Contextual Semantics for Sentiment Analysis of Twitter, Workshop: Semantic Sentiment Analysis, Crete, Greece. BEST PAPER AWARD! More Less
  • 18. SentiCirle: understanding topics and sentiment around political discussions (II) SentiCircle: Lexical-based sentiment representation model Assigns sentiment to a term by considering its co- occurrence patterns with other terms 18 The radio T-DOC is computed based on the degree of correlation between the two terms The angle is computed based on the prior sentiment of the term, extracted from an existing lexicon
  • 19. Discussions Understanding who are the users discussing policy in social media and how policy topics are debated could help PMs assessing how the citizens views and opinions should be weighted and considered to inform policy making This research is being incorporated into the Sense4us toolkit Several problems arise when using social media for this purpose: Data is distributed in multiple social platforms More research is needed to understand how representative is the subset of the population discussing policy in social media Social media -> big data issues: volume, variety, velocity and veracity of the data 19
  • 20. 20 PART 2 Modelling and Simulation of Public Policy Problems Sense4us Model Builder and Simulation Tool Aron Larsson, Osama Ibrahim, Anton Talantsev, eGovlab Department of Computer and Systems Sciences (DSV) Stockholm University
  • 21. Policymaking process model 21 Prescriptive analysis (Impact Assessment): Carried out at the early stages of policy development), which encompasses the forecasting of consequences and prescriptions about which policies should be implemented. Retrospective analysis (Evaluation): Tries to understand the causes and consequences of policies after they have been implemented.
  • 22. Decision support framework approach 22 Enabling policy analysis and decision evaluation where problem structuring is supported by linked open data, topic analysis and sentiment analysis of social media data. Policy problem structuring and modelling (Causal/cognitive map) Design policy options, consequence assessment, generate options (Simulation) Decision evaluation of policy options (Decision analysis) Sentimentanalysis Evidenceextractionfromopendatasources
  • 23. Why problem structuring? 23 Searching for the right information Capture a policy makers views about a problem Understanding decision making context and communication of problem understanding Structuring more complex cause-effect relationships Identifying where and how interventions have impact Enabling for decision evaluation of policy options
  • 27. Cognitive/causal mapping 27 CO2 emissions Means Subsidy bus tickets - Car traffic + Objectives It cannot cost more than 35 a month Its not the price, it is the frequency of buses Member of council
  • 30. Cognitive/causal mapping 30 CO2 emissions Means Subsidy bus tickets Car traffic + Objectives Increase bus frequencies - - Traffic planner Gas driven buses - Bus company CEO Member of council
  • 31. Cognitive/causal mapping 31 CO2 emissions Means Subsidy bus tickets Car traffic + Objectives Increase bus frequencies - - Gas driven buses - Town commerce + -20% +5% + Stakeholder Traffic planner Member of council Bus company CEO
  • 32. Evaluation concept: Impact potency 32 CO2 emissions Means Subsidy bus tickets 2 Car traffic + Objectives Increase bus frequencies 2 - - Gas driven buses 1 - Town commerce + -20% +5% + Stakeholder Bus company CEO Traffic planner Member of council
  • 33. Evaluation concept: Forward analysis 33 CO2 emissions Means Subsidy bus tickets Car traffic Objectives Increase bus frequencies -0.2 1m -0.5 1m Gas driven buses Town commerce -20% +5% Stakeholder -1.0 0m +1.0 0m +1.2 0m +0.6 4m Traffic planner Bus company CEO Member of council
  • 34. Forward analysis 34 CO2 emissions Means Subsidy bus tickets Car traffic Objectives Increase bus frequencies -0.2 1m -0.5 1m Gas driven buses Town commerce -20% +5% Stakeholder -1.0 0m +1.0 0m +1.2 0m +0.6 4m 20%, t=0 -4%, t=1 7%, t=4 25%, t=12 -29%, t=12 Traffic planner Bus company CEO Member of council
  • 35. Group decision and game concepts 35 CO2 emissions Means Subsidy bus tickets Car traffic Objectives Increase bus frequencies -0.2 1m -0.5 1m Gas driven buses Town commerce -20% +5% Stakeholder -1.0 0m +1.0 0m +1.2 0m +0.6 4m 20%, t=0 -4%, t=1 7%, t=4 25%, t=12 -29%, t=12 Traffic planner Bus company CEO Member of council
  • 36. Group decision and game concepts 36 CO2 emissions Means Subsidy bus tickets Car traffic Objectives Increase bus frequencies -0.2 1m -0.5 1m Gas driven buses Town commerce -20% +5% Stakeholder -1.0 0m +1.0 0m +1.2 0m +0.6 4m 15%, t=0 5%, t=4 15%, t=12 -21%, t=1210%, t=6 Bus company CEO Traffic planner Member of council
  • 37. To sum up We are building a tool that assist public decision processes through modelling a public policy problem, simulating policy consequences for decision evaluation considering multiple objectives and stakeholders. Visually structure the policy problem for increased problem understanding Show an understanding of the relations between policy instruments (funds, taxes, subsidies, prohibition, etc.) and societal effects. Generate different means or combinations of means controlled by different actors to reach similar targets (scenario generation) If we change this policy instrument, according to what we know, what are the effects (over time) on the factors subject to policy targets? Which stakeholders are affected and how? Will they react and if so what is the effect on the factors? 37
  • 39. 39 PART 3 Finding and Analysing Online Data to Support Governmental Decision Making Processes - the case of Sense4us Timo Wandh旦fer, GESIS Max Bashevoy, IT Innovation
  • 40. Persona UK Decision maker Member of the House of Commons Political interest: renewable energy 40
  • 41. Use Case 41 Similarities? Dissimilarities? Sense4us Topics Topics Debates in the plenum, draft bills, press releases Sense4us ?
  • 42. 42 PART 4 Assumptions to Artefacts: Understanding the Design Choices Underpinning the Sense4Us Project Somya Joshi, eGovlab Department of Computer and Systems Sciences (DSV) Stockholm University
  • 43. Objective To summarise where we are To understand how we got here To visualise where we are heading Where does Technology Assessment fit into this? 43
  • 44. Where we are at Milestones End user engagement first leg (Policy makers needs & requirements) Demo of tool stage one Integration of tools 44
  • 45. Understanding how we got here Design choices & assumptions Who is this for? What are we hoping to impact? Where is the innovation/ added value? How will end users engage with our tool set? 45
  • 46. What the end users expressed Requirements Provenance of data Transparency Sentiments & Opinions elicitation Summary & Visualisation of raw data sets Localisation of data Customisability of tool set 46
  • 47. Assumptions of Technology Impact Transparency in policy making Policy makers who want to take on board citizen opinions, discussions, information via social media resources Relevance & Provenance of data Publishers who want to make data accessible to others as well as increase their own knowledge
  • 48. Assumptions Impacts & Innovation Streamline & improve quality of linked open data searches Citizen solutions & knowledge will be summarised & analysed along sentiment/semantic lines Integration of problem structuring; support for impact assessment; preference elicitation when there are conflicting goals 48
  • 49. What we hope to achieve Co-evolution of tool set in line with end user feedback Integration of problem structuring and policy analysis tools Extend & deepen knowledge via Linked Open Data Identification of important stakeholders/ actors Reduced cognitive loads on policy makers: Making sense of the noise (reducing complexity) 49
  • 50. The road ahead Technological integration of the various components within the tool set User interface how will end users visualise, make sense & engage with our tool? Transparency in design (no black boxes), data relevance / provenance (trust) and rich results 50
  • 51. Gov 2.0: Enhanced policy decision support Social media and linked open data as a rich source of opinions, preferences, knowledge, that will be harnessed Simulation and modeling of policy alternatives & impacts 51 Where we envision were heading
  • 52. The contextual landscape The vending machine model of Governance: where the citizens only engage with shaking up the system when it doesnt work To Governance as Platform Where the platform is a metaphor for multi-layer decision making in complex, evolving environments 52
  • 53. Technology Assessment & Political Myths Are we designing a Political construct or a Technological artefact? Participatory Design within Policy context? Stage two of our end user engagement will further test this concept of designing a tool set in line with end user feedback Collaborative approaches have been argued to secure legitimacy . What are the anticipated risks and apprehensions on the part of end users within the Sense4us project? E.g. Political vs. Scientific Fact
  • 54. Technology Assessment of the Sense4Us Toolkit 54
  • 55. Discussion on our proposed approach How does one demonstrate new concepts to end users? How does one integrate that feedback into future iterations of the design? Our proposed approach is to demonstrate the various components using examples that are relevant and easy to understand To learn if this is understandable, useful, relevant and what they would like to see done differently 55

Editor's Notes

  • #10: http://en.wikipedia.org/wiki/Navitus_Bay_wind_farm Use case is real, all example data is entirely made up. This is just to give an idea of what information could be found by the toolkit. Citizens could also use the tool to mobilise opposition to the turbines there are already challenge websites set up, and the tool could be used to find related cons against other wind turbine installations, or to find evidence.
  • #13: Policy areas / policy subject areas (electric cars)
  • #14: Policy subject areas
  • #17: A more advance version of the sentiment analysis later on
  • #18: More / less
  • #22: Intelligence Design Choice Identification Development Selection
  • #23: (i) model a public policy problem situation using a causal semantic network or a causal map, defined by a single user, the policy analyst or a domain expert, or developed as a joint model of the problem through a synthesis analysis of multiple users cognitive understanding of the problem; (ii) simulate change transfer on the causal model by quantifying the links connecting the model variables and generating change scenarios; (iii) design alternative policy options based on a forward looking impact assessment in terms of economic, social, environmental and other impacts; and (iv) reach a policy decision after evaluation of policy options using a MCDA model.