Participatory Value Evaluation as a Tool for Value Extraction and Opinion Mining

Reduce Manual Data Analysis by Automated Value Extraction

More Info
expand_more

Abstract

Many grand challenges like climate change or health security cannot be solved by only the policymakers. Support and expertise of citizens is needed to solve these challenges (Gerton & Mitchell, 2019). Due to the demand for public participation, Participatory Value Evaluation (PVE) fulfils the needs of involving many participants in a scalable way in the process of participation and evaluation. PVE is unique in the way that it allows people to be a virtual decision-maker in a specific context. This method is not a decision-making tool in itself, but it allows participants to make their quantitative decisions and asks them to motivate their answers qualitatively. This results in qualitative and quantitative data about their choices. This research has explored opportunities of Natural Language Processing to automate the qualitative data processing of motivations given by citizens regarding a policy. The focus of this research has been set on human values to categorize the motivations of participants.Results from qualitative data analysis can support the process of policymaking. The outcomes of this exploratory study have been reflected by stakeholders involved in the case study to validate the application within the domain of decision-making. One outcome is that results from qualitative data analysis in PVE should never become a decision-making tool by itself. These outcomes help in creativity and reflection of the policy. Policymakers want to have outcomes of public evaluation and they want to be to act upon them. Analysing the qualitative data based on values give substance to certain values and can give guidelines on how to take these values into account. One way of doing this is visualising the values, norms and design principles in a value hierarchy using Value Sensitive Design. By doing this, knowledge of ‘the public’ can be used to support making policies.