Paper Type
Complete
Paper Number
1727
Description
This paper explores the application of machine learning to analyze citizens' inputs from e-participation to support participation and urban experts. Through eleven semi-structured expert interviews, the collection of four datasets from past participation, and the systematic analysis of them according to the knowledge discovery in the database framework, we identified five fundamental challenges in current approaches and developed an ML-based process model for clustering citizens' inputs. To do so, we created a preprocessing and ML model training pipeline, proved its feasibility by its implementation, and described it in detail as a baseline for further research in this domain. However, comparing expert-generated clusters with those from ML models revealed some distinctions due to differing mental models emphasizing the need for further models and expert-friendly interfaces. With the developed process model, this paper contributes to the citizen design science model by offering practitioners a flexible analysis approach and guidance.
Recommended Citation
Borchers, Marten; Cao, Thu-Bao; and Bittner, Eva A. C., "Toward the ML-based Analysis of Citizens' Inputs from E-Participation in Urban Planning" (2024). PACIS 2024 Proceedings. 6.
https://aisel.aisnet.org/pacis2024/track16_shareecon/track16_shareecon/6
Toward the ML-based Analysis of Citizens' Inputs from E-Participation in Urban Planning
This paper explores the application of machine learning to analyze citizens' inputs from e-participation to support participation and urban experts. Through eleven semi-structured expert interviews, the collection of four datasets from past participation, and the systematic analysis of them according to the knowledge discovery in the database framework, we identified five fundamental challenges in current approaches and developed an ML-based process model for clustering citizens' inputs. To do so, we created a preprocessing and ML model training pipeline, proved its feasibility by its implementation, and described it in detail as a baseline for further research in this domain. However, comparing expert-generated clusters with those from ML models revealed some distinctions due to differing mental models emphasizing the need for further models and expert-friendly interfaces. With the developed process model, this paper contributes to the citizen design science model by offering practitioners a flexible analysis approach and guidance.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
Crowd