Home

Legal analysis

Privacy issues

IPR of the datasets

Licenses

Limitations on public access

Economical conditions

Temporary aspects

Ethical analysis

The ethical analysis of the datasets used in this project is based on two key frameworks: the principles outlined by the Data Ethics Institute document and the questions posed in the Data Ethics Canvas developed by the Open Data Institute (ODI). These frameworks help examine aspects such as transparency, privacy, bias, and public benefit. In several instances, however, the absence of critical metadata or methodological documentation has limited the depth and clarity of the ethical evaluation.

Bologna

Air Quality Bologna (ID - DS1)

This dataset, sourced from Arpae (Agenzia regionale per la prevenzione, l’ambiente e l’energia dell’Emilia-Romagna), is accessible via the Bologna City Council’s Open Data platform. It contains no personal or human-related data, as it is collected through environmental monitoring instruments at three fixed stations. The data is openly licensed under CC BY 4.0 and updated annually.

From an ethical perspective, the dataset demonstrates strong transparency and public accessibility. Its release supports public awareness, academic research, and policy decisions. However, being limited only to three stations and neighbourhoods in Bologna, raises potential concerns about spatial representation and environmental justice, as some zones of the city may be underrepresented in the data. Overall, the dataset aligns well with the principles of openness and public benefit, though further clarification on data coverage could enhance equity considerations.

Private Vehicles Registered in Bologna by Power Type and CAP (ID - DS2)

This dataset is published on Bologna’s Open Data website and originates from the Automobile Club of Italy (ACI). It details vehicle registrations by postal code and includes anonymized demographic information such as sex, age group, and vehicle fuel type. The data is updated annually and made available under a CC BY 4.0 license.

Although the dataset is presented as anonymized and it is grouped under numbers of different groups in postal code area. So, there is no theoretical risk of re-identification of the data. In this dataset, being free from bias and discrimination requires critical examination. For example, how vehicle data is aggregated and categorized could inadvertently reflect or reinforce socioeconomic inequalities [Less wealthier zones, have fewer electric cars, and this may interpret as they are responsible for air pollution more than the others]. This dataset is useful for environmental research and mobility planning but may also be exploited for commercial interests, such as targeted marketing. Overall, while the dataset serves important functions, it would benefit from enhanced documentation regarding privacy safeguards and data processing methods.

Public Transportation by Fuel Type (ID - DS3)

Collected and published by the Bologna City Council and TPER (Emilia-Romagna’s public transport provider), this dataset includes information about the types of fuels used in the city’s public bus fleet. It contains no personal data.

The dataset appears ethically sound, promoting sustainability and enabling researchers and policymakers to assess progress toward low-emission mobility goals. However, the licensing terms are not explicitly stated. While the data is freely accessible, clearer legal guidance on reuse would enhance its ethical transparency and utility. The dataset does not introduce any evident risk of harm or discrimination and supports public interest objectives well.

Public Transport Passenger Numbers (ID - DS4)

This dataset reports the number of public transport trips in Bologna by month and year and is published by the Bologna City Council and TPER. It captures aggregated human activity but contains no personal or identifiable data.

Ethically, the dataset poses minimal privacy risks and holds significant potential for improving urban mobility and infrastructure planning. However, the absence of detailed information on how data is collected (e.g., through ticketing systems or automated sensors) limits full transparency. Additionally, while the dataset does not appear biased, it would be helpful to analyze it in conjunction with socioeconomic data to assess transport equity. The lack of a clearly stated license is another area for improvement, though the dataset is available for public use with attribution.

Milan

Air Quality Milan (ID - DS5)

This dataset is maintained by AMAT (Agenzia Mobilità Ambiente Territorio) and published by the Metropolitan City of Milan. It consists of environmental air quality measurements and does not include any personal or sensitive data.

The data is updated annually and is publicly accessible, reflecting a high level of transparency. The data is derived from seven stations in Milan, which is quite fair but as with Bologna’s air quality data, a more detailed explanation of the station network would help assess whether all neighborhoods are equally represented. The dataset supports public health and environmental policy and aligns well with ethical principles of transparency, beneficence, and sustainability.

Low Emission Bus Fleet in Milan (ID - DS6)

This dataset contains statistics on the number of low-emission buses operating in Milan. It is non-personal in nature and does not include sensitive data.

The data contributes positively to understanding urban environmental progress and supports research and public policy.

Buses by Emission Class (ID - DS7)

Published by the Council of Milan, this dataset categorizes buses used for local public transport by emission class. It is non-personal and regularly updated.

This dataset has a clear environmental and administrative purpose. It does not raise ethical concerns regarding privacy or discrimination. Its utility in policy-making, sustainability tracking, and infrastructure planning makes it ethically robust, although greater methodological detail would be welcome.

Annual Passengers per Inhabitant (ID - DS8)

This dataset presents the number of annual public transport trips per inhabitant in Milan. While it is a high-level indicator, it lacks transparency regarding its construction and data sources.

From an ethical standpoint, the absence of methodological clarity limits its trustworthiness. Misinterpretation of such indicators can lead to misleading conclusions about mobility access or public transport usage. To enhance transparency and accountability, the data providers should publish a clear explanation of how the indicator is calculated.

Vehicle Fleet Milan (ID - DS9)

This dataset, produced by ACI and published by the City of Milan, shows the number and types of vehicles registered in the city. It is derived from personal registration data but has been fully anonymized.

Ethically, the dataset appears to pose no direct privacy risks if proper anonymization protocols are in place. The dataset is valuable for urban planning, environmental analysis, and transportation policy, but safeguards should be in place to prevent misuse, such as commercial profiling or discriminatory policy applications.

Conclusion

Across both cities, most datasets align well with key ethical principles such as transparency, openness, sustainability, and beneficence. However, common areas for improvement include clearer licensing, detailed methodological documentation, and transparency about anonymization processes. Applying the Data Ethics Principles from DataEthics.eu and the ODI Data Ethics Canvas reveals that while these public data initiatives are generally well-intentioned and valuable, greater critical reflection and improved documentation could significantly enhance their ethical robustness and trustworthiness.

Sustainability of the update of the datasets over time

All of the datasets we used in our project are time dependent and represent data over time. This obliged them to be updated yearly. Unfortunately, some of the datasets are not being updated very regularly.

Air Quality Bologna (ID - DS1)

The data is annual and is being updated every January with care. Therefore, the dataset is up-to-date until 2024. Good to know that the daily data is also available in another dataset, provided by the city council of Bologna.

Private Vehicles Registered in Bologna by Power Type and CAP (ID - DS2)

This data set is being updated yearly and by the end of each year, and it is up-to-date until 2024.

Public Transportation by Fuel Type (ID - DS3)

It is mentioned in the information section of the dataset that data shows the situation on 31/12 of each year. Unfortunately, the data for 2024 is not updated and it is up-to-date until 2023.

Public Transport Passenger Numbers (ID - DS4)

It is mentioned in the information section of the dataset that data shows the situation on 31/12 of each year. The data for 2024 is not updated and it is up-to-date until 2023.

Air Quality Milan (ID - DS5)

The data is being updated by the information released by Arpa Lombardia on a daily basis. The data is maintained well and is updated regularly. Although due to a reason mentioned in the source website the data from 1 May to 30 September in 2024 is not available.

Low Emission Bus Fleet in Milan (ID - DS6)

The dataset has not been updated since 2018, and due to unknown reasons the new data on Bus Fleet in Milan is not inserted into the dataset.

Buses by Emission Class (ID - DS7)

The dataset has not been updated since 2018, therefore, the dataset will get a poor score with respect to sustainability of the update over time.

Annual Passengers per Inhabitant (ID - DS8)

The dataset contains data until 2018 and has not been updated ever since.

Vehicle Fleet Milan (ID - DS9)

This dataset should be updated every year and it is up-to-date until the end of 2022 and the last two years are not inserted in the dataset. The data is provided by ACI and their open policy but it is not updated by the city council of Milan.

Quality analysis

The quality analysis of these datasets has been carried out following the europan standards applied by Italy UNI ISO/IEC 25012:2014.

Technical analysis

Source datasets

Following is the evaluation of the source datasets based on the AGID metadata model, classifying metadata quality on a range from 1 to 4, based on the detail level, and the bond between data and metadata.

RDF assertion and metadata for the mashup datasets

All producedd mashup datasets have been described with metadata following the specification of DCAT-AP. While a point may be made, that since all our data are Italian in origin, we could use DCAT-AP_IT instead, we decided on using more flexible (and an European standard) DCAT-AP instead. Additionally, PROV-O, the provenance ontology, is used to described sources and activities.

Additional metadata for source datasets

The metadata for the source datasets has been gathered directly from the original sources. Where information was missing, it was inferred and supplemented following the same guidelines used for mashup datasets (e.g., assigning a theme to a source dataset was done according to the same European authority).

Catalogue

Original datasets' metadata

Mashup datasets' metadata

FAIR principles

As a part of the technical analysis, we have performed an analysis of the FAIR principles of all the source datasets. These are a set of recommendations proposed by the GO FAIR Initiative, a consortium of scientists and organisations aimed to support Findability, Accessibility, Interoperability and Reusability of digital data.

Findability

Accessibility

Interoperability

Reusability

As can be seen, the datasets are practically grouped in three groups, corresponding to where they were originally published. Datasets DS5 to DS9, coming from the Open Data service of Comune di Milano, comply with all FAIR principles. Datasets DS1 and DS2, coming from the Open Data service of Comune di Bologna, comply with all but one FAIR principle, as they and their metadata lack qualified references to other (meta)data. Datasets DS3 and DS4, on the other hand, are very lackluster when it comes to compliance with FAIR principles. These two datasets come from the metropolitan statistical portal made as part of an agreement between Comune di Bologna and Citta’ Metropolitana di Bologna. The datasets are basically provided as is, with no metadata, no API access and lackluster licensing information; as they are only provided with a copyright saying that the data comes from Comune di Bologna and can be freely consulted, used and reproduced on condition of citing the source. Worth noting is the fact that these two datasets are not available on the Open Data service of the Comune.