A viable systems perspective for managing urban complexity: collective perception based on fuzzy and semantic approaches in the decision-making process

Purpose. The growing complexity of urban landscapes and the faster technology evolution make central a rethink of urban governance, in order to understand how the application of smart technologies and automatized research techniques to big data management reframe and boost decision-making processes. The adoption of a systems perspective permits to focus on the definition of new model of decision-making for urban context, based on a participatory logic, which can increase actors’ engagement, which populate the same context and harmonize their objectives with community’s overall goal. In this regard, adopting the interpretative lens of the Viable Systems Approach (VSA), this work aims to propose a decision-making support model for the urban governing body to frame the urban collective perception of the actors (e.g. citizens and tourists) regarding the city and its assets. Thus, the adhesion to VSA can enhance the awareness of the interventions that could be aimed at solving urban problems. Methodology. The proposed model follows a robust and innovative methodological framework based on a big data-oriented approach. Firstly, following the Lynch’s theory, the most relevant urban elements that contribute to create the common perception of a city are defined. After, starting from crawling texts from online sources, the methodology foresees to apply an Aspect Based Sentiment Analysis (ABSA), an advanced sentiment analysis, to evaluate the sentiment expressed in the reviews by online users regarding the urban elements. Finally, scenario analysis is performed by using Fuzzy Cognitive Map (FCM) to analyse the impact of users’ opinion about city issues. Findings. A large-scale text analytics study has been conducted on two selected Italian cities. The results lead to an exposition of shared evaluations on the levels of “sentiment” as perceived by the community in relation to urban points of interest through summary sheets. Furthermore, carrying out a What-If simulation, it is determined how the current collective perception affects other important urban issues and how, changing the collective perception through targeted interventions, the urban context will react. In this way, the information variety endowment of the decision makers is increased and a series of interventions, aimed at establishing the conditions for a context consonance, by obtaining an overall view composed by the different perceptions of the community, can be implemented. Research limits. Beyond the advantages offered by big data analysis (primarily the possibility of analyzing a huge amount of data in real time), the automated collection of people’s reviews is characterized by a certain superficiality, since it does not allow going deep into the understanding of the people’s opinions. In fact, although the sample was particularly large (a huge amount of reviews extracted in a period span of 12 months), could be interesting a deeper analysis of users’ complete through in-depth interviews. Practical implications. The model can support urban decision-makers by offering some insights on the level of sentiment and some relevant results deriving from scenario analysis to understand how the collective perception of the city can influence important urban questions and how the governing body should intervene for aligning to an ideal city. In this way, it is possible to reduce the delay which characterizes the urban systems in relation to a critical situation perceived by the community and the interventions by the institutions. Another important aspect regards the capacity to depict graphically the results obtained: processing information through the visual system can significantly increase managerial capability to address complexity. Furthermore, the proposed model promotes a common language between urban decision makers and stakeholders by guaranteeing a greater awareness of the interventions to carry out. Originality. The originality of the paper lies in combining, in a single model, the VSA, interpretative lens of reality, with an innovative methodology followed a big data-oriented approach. In particular, the work has utilized knowledge from three different fields, i.e. urban management, computing science and statistics, which have been synergistically integrate for customizing, implementing, and using IT tools capable of automatically identifying, selecting, categorizing and analyzing the collective perception of a city and urban assets in it through people’s reviews.