Supporting policy-making with social media and e-participation platforms data: A policy analytics framework

https://doi.org/10.1016/j.giq.2021.101590Get rights and content

Highlights

  • Policy Analytics refers to the use of data analytics to support policy-making;

  • Data from social media and e-participation platforms are difficult to combine and process;

  • Policy-makers' requirements towards policy analytics are identified;

  • Our framework bundles data analysis techniques for the whole policy-making process;

  • The framework is presented and validated as a dashboard to increase usability.

Abstract

E-participation enables citizens to have an impact on policy-making through electronic means. Two of the most popular channels are social media and dedicated e-participation platforms. However, the ideas, comments, discussions of citizens on these two channels generate a lot of data to be processed by political representatives or public agents afterwards. Despite the existence of various techniques for social media analytics, literature is scarce regarding the analysis techniques to mine e-participation platforms as well as the possible combination of insights between the two channels.

In order to address these gaps, we design a policy analytics framework to leverage insights from e-participation platforms and social media through relevant data analytics to support policy-making. In order to do so, we rely on the Design Science Research methodology. Through the analysis of four different cities in Belgium (Liège, Mons, Marche-en-Famenne, Leuven), we identify policy-makers' requirements and needs of information from platforms and social media. Then, we explore data analysis techniques to address those requirements. Finally, we design an actionable framework, present it as an interactive dashboard and iteratively test it on the case of Liège. This policy analytics framework supports each step of the traditional policy-making process with appropriate data analytics applied to the two sources.

Introduction

The impact of citizens on the decisions taken by political representatives and on policy-making in general is labeled as “citizen participation” (Arnstein, 1969) and is not new. However, this participation can be further stimulated through the use of Information and Communication Technologies (ICT), making it more accessible but also cost-efficient. The use of ICT to support participation is labeled as “e-participation” and defined as “the use of information and communication technologies to broaden and deepen political participation by enabling citizens to connect with one another and with their elected representatives” (Macintosh & Whyte, 2008).

e-participation encompasses different participation channels (Berntzen & Johannessen, 2016; Simonofski, Snoeck, Vanderose, Crompvoets, & Habra, 2017). These channels can be structured into two main categories: government-led initiatives, and citizen-led initiatives (Lee & Kim, 2014; Porwol, Ojo, & Breslin, 2016b). Government-led initiatives are for example e-voting, online surveys or e-participation platforms, whereas citizen-led initiatives could be blogs or forums where citizens can write their ideas and react to ideas of fellow citizens (Lee & Kim, 2014). Porwol et al. (2016b) mention that the most common citizen-led initiatives are discussions on social media platforms such as Facebook or Twitter, since this is the area of the Internet where citizens are most active in general. In this paper, we will focus on two of these channels: e-participation platforms (Government-led) and social media (Citizen-led).

Macintosh, Coleman, and Schneeberger (2009) underline the importance of building synergies between these two channels. Indeed, e-participation platforms are often built or bought to stimulate participation among the population in a structured way but often lack engagement (Macintosh et al., 2009; Toots, 2019). On the other hand, citizens tend to engage quite easily on social media with spontaneous political discussions but in a less structured manner (Porwol et al., 2016b), which makes it harder to exploit for policy-makers. Furthermore, it is particularly interesting for policy-makers to gather insights from both channels as the two channels are not used by the same groups of citizen. On the one hand, social media gather numerous citizens sharing ideas without any real political intent. This channel has therefore the potential to be very inclusive. On the other hand, e-participation platforms gather more motivated individuals who are already politically active (Berntzen & Johannessen, 2016). On top of those respective limitations, there is a gap around the data analysis techniques that could be applied to make a better use of the data available on these two channels, especially for online platforms. The available information can be overwhelming and difficult to process for policy-makers, who would therefore benefit from a better process/system to extract meaningful observations or even recommendations. Several data analysis techniques have been mentioned in the literature such as Lago et al. (2019) for e-participation platforms or Belkahla Driss, Mellouli, and Trabelsi (2019) for social media. Even though previous research focused on data analysis to improve policies (e.g. Policy modeling (Gilbert, Ahrweiler, Barbrook-Johnson, Narasimhan, & Wilkinson, 2018), Policy analysis (Kolkman, 2020)), the possible combination of data analysis from these two specific channels and embeddedness in a broader policy-making process are two questions that remain to be explored. As an answer, this paper suggests an actionable policy analytics framework for policy-makers to harness the ideas, discussions, and feedback provided by citizens. This framework supports each step of the policy-making process with appropriate data analytics to harness insights from the two channels.

This paper is structured as follows. In Section 2, we discuss the relevant literature about data analysis techniques for e-participation platforms and social media. Furthermore, we explain how these techniques can impact policy-making. In Section 3, we explain the different methodological stages of Design Science Research we followed to build the framework as well as the four selected cases to support this process. In Section 4, we detail the requirements of practitioners regarding the framework, we describe the framework, and we test it on the case of Liège (Belgium). In Section 5, we describe the implications for research and of practice of this paper. We also detail its limitations and introduce further research leads. Finally, in Section 6, we summarize the findings of our research.

Section snippets

Data analytics for policy-making: policy analytics

A public policy is established through a policy-making process: a type of decision process with specific characteristics (Tsoukias, Montibeller, Lucertini, & Belton, 2013). Many theoretical frameworks exist to describe the chronological steps this policy-making process consists of. In this paper, we rely on the broadly accepted work of Howlett, Ramesh, and Perl (2009) that describes the Public Policy-Making Process into five stages:

  • 1.

    Agenda setting: framing of the problem and exploring the needs

Research questions

First of all, several data analysis techniques have been mentioned in the Background to exploit data from both platforms and social media. However, we take the viewpoint of the policy-makers and start from their need for information coming from both channels. This will serve as initial requirements to develop our framework. Therefore, we formulate this first Research Question (RQ):

  • RQ1: Which information do policy-makers find relevant to extract from social media and from e-participation

Identified requirements for policy analytics

In this section, we present the stakeholders' requirements about the extraction and analysis of data from the two channels to support the policy-making process. Table 2 summarizes the requirements extracted from the interviews, the stages of the policy-making cycle they are linked to, the source of data (social media or platform), and the occurrence of the requirements across the interviews.

To complete the insights summarized in the table, we can mention that the Agenda-Setting, Formulation and

Theoretical implications

This paper is an early attempt at formalizing the policy analytics process in the context of e-participation. The suggested framework inserts itself at the crossroads of several research fields: public administration (policy-making steps), digital government (policy analytics), e-participation and data science. Our approach draws its robustness from these different approaches.

In the background, we presented the emerging theory-building of policy analytics with seminal studies such as (

Conclusion

This paper contributes at several levels. First, we identified 10 policy-makers' requirements from four different cities (Liège, Mons, Leuven and Marche-en-Famenne) regarding mining e-participation platforms and social media for policy analytics. These requirements are mapped to the five stages of policy-making. Second, we explored relevant data analysis techniques (document clustering, document filtering, ranking and sentiment analysis) and their combination to address the requirements.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We would like to acknowledge the Belgian Federal Science Policy Office (BELSPO) for their support. The research pertaining to these results received financial aid from the Federal Science Policy according to the agreement of subsidy no. [B2/191/P3/DIGI4FED].

Anthony Simonofski is post-doctoral researcher at the Computer Science Faculty of the University of Namur (UNamur) and at the Faculty of Economics and Business of the KU Leuven. He received his PhD in Business Economics from KU Leuven and in Computer Science from UNamur in 2019. He has published over 30 peer-reviewed scientific papers in international conferences and journals and received the Best Paper Award at the IEEE CBI in 2017. His research focuses on the digital transformation of

References (65)

  • E. Abu-shanab

    E-participation levels and technologies

  • R. Anderson

    Thematic content analysis (TCA): Descriptive presentation of qualitative data using microsoft word

    Descriptive Presentation of Qualitative Data

    (2007)
  • A. Androutsopoulou et al.

    Using social media monitoring for public policy making: An evaluation

  • S.R. Arnstein

    A ladder of citizen participation

    Journal of the American Planning Association

    (1969)
  • M. Asher et al.

    Assessing the effectiveness of e-petitioning through Twitter conversations

  • B. Bengfort et al.

    Applied text analysis with Python: Enabling language-aware data products with machine learning

    (2018)
  • L. Berntzen et al.

    The role of citizen participation in municipal smart city projects: Lessons learned from Norway

  • G.G. Chowdhury

    Natural language processing

    Annual Review of Information Science and Technology

    (2003)
  • P. Cihon et al.

    Tweeting for the cause: Network analysis of UK e-petition sharing activity

  • S. Clay

    The political power of social media: Technology, the public sphere, and political change

    Foreign Affairs

    (2011)
  • F. Cronemberger et al.

    Problem conceptualization as a foundation of data analytics in local governments: Lessons from the City of Syracuse, New York

  • K.A. Daniell et al.

    Policy analysis and policy analytics

    Annals of Operations Research

    (2016)
  • G. De Marchi et al.

    From evidence-based policy making to policy analytics

    Annals of Operations Research

    (2016)
  • J. Desdemoustier et al.

    The state of development of “Smart City” dynamics in Belgium: A quantitative barometer

    (2017)
  • J. Dolson et al.

    Explaining variation in the e-Government features of municipal websites: An analysis of e-Content, e-Participation, and social media features in Canadian municipal websites

    Canadian Journal of Urban Research

    (2012)
  • E. Drever

    Using semi-structured interviews in small-scale research. A Teacher’s guide

    (1995)
  • L.N. Gerston

    Public policy making: Process and principles

    (2014)
  • N. Gilbert et al.

    Computational modelling of public policy: Reflections on practice

    Journal of Artificial Societies and Social Simulation

    (2018)
  • J.R. Gil-Garcia et al.

    Policy analytics: Definitions, components, methods, and illustrative examples

  • G. Guest et al.

    How many interviews are enough?: An experiment with data saturation and variability

    Field Methods

    (2006)
  • L. Hagen et al.

    Open data visualizations and analytics as tools for policy-making

    Government Information Quarterly

    (2019)
  • S.A. Hale et al.

    Investigating political participation and social information: Using big data and a natural experiment

  • Cited by (29)

    • Government Digital Transformation: Understanding the Role of Government Social Media

      2023, Government Information Quarterly
      Citation Excerpt :

      Thus far, the existing studies mainly focus on the communication effects of government information under crisis management or government-related topics. For example, citizen participation in government social media (GSM) during the crisis (Guo, Liu, Wu, and Zhang, 2021), GSM and citizen engagement during the Covid-19 pandemic (Chen et al., 2020; Mansoor, 2021b), social media and E-governance (Criado, Sandoval-Almazan, and Gil-Garcia, 2013), citizens' government-led E-participation through social media (Alarabiat, Soares, and Estevez, 2021; Simonofski, Fink, and Burnay, 2021) as well as social media and E-government services (Aladwani and Dwivedi, 2018; Dwivedi et al., 2017; Khan, Umer, Umer, and Naqvi, 2021). There is still a lack of understanding of government information dissemination effect on cultivating innovation acceptance.

    • Capricious opinions: A study of polarization of social media groups

      2022, Government Information Quarterly
      Citation Excerpt :

      Eventually, like-minded individuals form homogenous groups exhibiting attributes connected through weak ties (Villodre & Criado, 2020), referred to as online communities on SMP. Finally, users within these groups start forming, sharing, or changing their social media-induced opinions (SMIO) towards a policy (Simonofski, Fink, & Burnay, 2021) or event. Some individuals being strongly opinionated (Xu, Xu, & Li, 2016) change the rest of the individuals' original views, also referred to as social media-induced (SMI) influencers leading to polarization.

    • Deep embedded median clustering for routing misbehaviour and attacks detection in ad-hoc networks

      2022, Ad Hoc Networks
      Citation Excerpt :

      To analyse and have profound insights into the data, scientific modeling and syntax are used. Additional insights in this area are needed to keep up with the exponential growth of manufacturing, business, and medicine [27]. Computer science makes use of data mining, big data, and machine learning techniques to acquire such in-depth knowledge of the above areas.

    View all citing articles on Scopus

    Anthony Simonofski is post-doctoral researcher at the Computer Science Faculty of the University of Namur (UNamur) and at the Faculty of Economics and Business of the KU Leuven. He received his PhD in Business Economics from KU Leuven and in Computer Science from UNamur in 2019. He has published over 30 peer-reviewed scientific papers in international conferences and journals and received the Best Paper Award at the IEEE CBI in 2017. His research focuses on the digital transformation of (public) organizations through several lenses, including: e-participation, smart cities, e-government, digital governance and open data.

    Jerôme Fink is received a master's degree in computer science from the University of Namur in 2019. He is currently working as a PhD Researcher at the Computer Science Faculty of the University of Namur. His research interests include computer vision, gesture recognition and natural language processing. His research project focuses on the translation of sign language video to text.

    Corentin Burnay is Associate Professor in Information Management at the Business Administration Department of the University of Namur, in Belgium. He is a member of the PReCISE Research center and the Namur Digital Institute. His research focuses on using software and requirements engineering methods to better specify and design the content and features offered by decision support systems (DSS) interfaces, like reports, dashboards or corporate cockpits. He is particularly interested in studying requirements elicitation and how it helps to detect recurring defects of DSS interfaces like informational overloads and structural defaults. Recently his work is focused on mobilizing alternative technologies like blockchain and big data to improve the practice of DSS interface design. He has published over 20 peer-reviewed scientific papers in international conferences and journals and is the recipient of the 2014 CAiSE conference distinguished paper award. Corentin Burnay is also regularly solicited as a reviewer for journals like “Empirical Software Engineering” or “Journal of Systems and Softwares”. He has organized and chaired the series of International Workshops on Modeling and Reasoning for Business Intelligence (MORE-BI).

    View full text