Urban IoT Ontologies for Sharing and Electric Mobility

This article describes a modular ontology designed for the domain of sharing and electric mobility.

The ontology is of very high quality and relevance. The article convincingly motivates the work, clearly describes the ontology engineering methodology, and concisely describes each step of the ontology design, development, publication, and maintenance process. The key aspects of the ontologies are very well conveyed, with very good illustrations and textual descriptions.

I definitely do recommend to accept this article.

I have almost no criticism, only a small list of minor issues that the authors can implement rapidly, and an additional pointer to recent literature that could be integrated in future work.
In what follows, I will provide a short summary of the different sections, then list minor issues.

Section 1 motivates the work clearly: municipalities like Milan (IT) face semantic interoperability issues when addressing the challenge of digital information governance. Sharing and electric mobility is especially interesting a domain in data governance, because many competing service providers operate with a growing amount of data generated by IoT devices and available as open data.
Developing an ontology for sharing and electric mobility is an promising approach to solve these issues. The authors led their work in collaboration with the municipality of Milan, IT.

Section 2 lists the most related standardization initiatives, research and development projects, and resulting data models and ontologies. Probably, only the very recent SAREF4AUTO extension of SAREF is missing in the picture. Other related work such as SENSORIS may be found in ETSI TS 103508: "SmartM2M; SAREF extension investigation; Requirements for Automotive", and it may be appropriate to align to the recent SAREF extension SAREF4AUTO: ETSI TS 103 410-7: "SmartM2M; Extension to SAREF; Part 7: Automotive Domain". Both are available online since July 2020.
Reference to the SC 31: "Data communication" of the ISO TC22: "Road vehicles", may be relevant.

The ontology was designed, developed, evaluated, and published, following the recent Linked Open Terms methodology, which is described in Section 3. The structure of the paper follows the main steps of this methodology, and describes precisely each step.

Section 4 discusses how the requirements for the Urban IoT ontology were produced. The process involved a team of domain experts from the Municipality of Milan and the authors' institutions. 3 Use cases for the sharing mobility, 3 use cases for the electric mobility, and a joint use case, were defined. From these use cases, 69 use stories were developed. Then 136 facts and 87 competency questions were listed to define the scope of the ontology, and validated by the domain experts. All the resources are available online in a public git repository of the municipality of Milano.

My only regret is that the Facts and Competency questions are only available in Italian. https://github.com/Comune-Milano/ontologie-iot-urbani/blob/master/requis...

Section 5 describes the ontology, which consists of a core module for Urban IoT, and two extension modules developed for sharing and electric mobility, respectively. These ontologies reuse relevant terms from the most appropriate existing ontologies, without importing them. Each module is very clearly described in the article. The figures are very clear and help to grab the ontology at a glance, with how external ontologies are reused.

A quick check with OWLAPI shows that the ontologies conform to the OWL2 DL and OWL2 RL profiles. (more specifically, I use https://github.com/stain/profilechecker ). This could be a relevant information to add, along with statistics about the number of Classes, Object Properties, Datatype Properties, Instances, in each ontology module and SKOS concept list.

Section 6 describes how the ontology has been evaluated: by reporting the execution of the OOPS! ontology pitfall scanner, and auto-evaluation according to the criteria for ontology evaluation: Accuracy, Adaptability, Clarity, Completeness (report is missing!), Efficiency, Conciseness, Consistency, Organizational fitness

Two minor issues in this section:
- The raw results of the OOPS! scan, and the date of the scan, should be made available on the git repository.
- A report online for the Completeness part is missing. The repository only contains a few SPARQL queries in https://github.com/Comune-Milano/ontologie-iot-urbani/blob/master/exampl...
however, in p14 Section 8.1 the authors claim that every defined competency question could be defined answered.

Section 7 describes how the ontology is published using an open CC-BY license, how the authors chose permanent w3id.org URIs, and made the resources available online in HTML or different RDF formats, using content negotiation. The maintenance can be led using a public github issue tracker on the github repository of the project.

Section 8 provides simple examples of SPARQL queries that answer competency questions, and JSON-LD documents that provide example usage of the ontology. There are a few minor issues with these queries and documents, as listed below, however the content is interesting and illustrates clearly how the ontology can be used in practice. I would personally rather have the examples in turtle instead of JSON-LD, especially as the JSON-LD context is not a real IRI. More online resources such as a synthetic dataset and more queries and examples would be appreciated.

Section 9 discusses potential long term impacts for the ontology, and triggers to-the-point discussion. This section complements nicely the evaluation of the resource.

Minor issues:

page 2 : 2nd column : line 3 (shorter notation: p2:2:3) - is currently lead -> is currently led
p3:1:4: EVSE are more commonly referring to Electric Vehicle Service Equipment, which is a more generic concept than Electric Vehicle Supply Equipment
p3:1:26: do you mean ex., ? or e.g., ? or i.e., ?
p3:2:34: the following reference may be used for the latest version of the SAREF ontology:
> ETSI SmartM2M. SmartM2M; Smart Applications; Reference Ontology and oneM2M Mapping. ETSI Technical Report TS 103 264 V3.1.1 (2020-02)
p7:2:33: there is no link to the repository at this point. A footnote should be added.
Fig 2. kos:schemename - the prefix is undefined
Some of these ontologies should be described (ex. legal:)
Fig 4. the link between uiots:SharingMobilityOffer and schema:Offer is undirected and unlabeled
p14:2:5: prefix missing for the Unavailable state
Listing 1: variable ?a is not bound in the query. Prefixes are not all defined (ex. sh-kos-state:) It would be better to be able to test these queries on a synthetic dataset online
p14:2:29: by a not registered user -> by an unregistered user
Listings 3 to 5: Why isn't the @context value a http(s) IRI? The context at https://tinyurl.com/yy3abvss should be enhanced (aliases for common terms) and made available in the git repository
p16:2:10: national Access Points -> National Access Points (unsure about that one)

This manuscript was submitted as 'Ontology Description' and should be reviewed along the following dimensions: (1) Quality and relevance of the described ontology (convincing evidence must be provided). (2) Illustration, clarity and readability of the describing paper, which shall convey to the reader the key aspects of the described ontology.

This paper presents the Urban IoT suite of ontologies for sharing and electric mobility, which is, as the name suggests, a modular suite of ontologies representing data gathered from Urban IoT devices used in urban mobility.
Main challenge of the work is to handle the heterogeneity and the amount of information derived by the increased use of IoT technologies in the urban mobility context. In fact, both cities have become an inexhaustible source of data, and mobility itself is undergoing a profound change, due to the increasing focus on eco-friendly solutions and the emergence of new mobility habits and modern means of transport.
In this scenario, such information flow offers the opportunity to create data-driven solutions to support decision making and to offer value-added services to citizens. However, it implies two main challenges: the needs of specialised technological infrastructures and the multitude and the variety of the actors involved.
The proposed solution is a conceptual model to harmonise data exchanges between municipalities and service providers, with a specific focus on sharing and electric mobility domains.

As far as the structure and presentation of the topics are concerned, the article appears to be well written and the work well organised.
Challenges and objectives are explained and described in the introduction, which encourages reading and makes it clear from the outset what to focus on. Within this first section, authors also include a short presentation of the structure of the article and this makes the reading much easier. After the description of the work, which is composed of four main phases (Requirements specification, Ontology implementation, Ontology publication, and Ontology maintenance), there are several examples giving a clearer idea of how the implementation works and adding empirical evidence of the usefulness of the work.
Analysing the article in its contents, the results obtained are consistent with the objectives set.
During the implementation of the model, existing relevant vocabularies are considered and re-used to represent identified concepts and relations, but the major innovation lies in the conceptual effort. The result of the work is a modular suite consisting of a main module (Core) that currently extends into just two modules representing the sharing and the electrical mobility context. This choice makes the model extensible in the future. Therefore, not only the final ontology is coherent with the domain of interest, but a general and extensible model is thus presented. This last result is fundamental for the management of heterogeneous data in an ever-expanding context.

In conclusion, even though the evaluation of the article is overall positive, here are some pieces of advice:

- I suggest replacing the title of Section 3 "Ontology engineering methodology" with "Ontology engineering workflow". The section describes the different phases leading to the development of ontology; however, these are already known and correspond to the traditional phases of software development, so there is no implementation of an innovative methodology.

- I suggest adding a clearer specification of why one ontology or vocabulary was chosen to be imported instead of another one (if possible or relevant) and adding bibliographical references or footnotes for each acronym or tool cited. Below there is a list of some of those currently lacking:

- GSM
- OASC
- SEMIC
- SKOS
- AMAT

- I suggest revising Table 1. The table has three headings: "use cases", "user stories" and "competency questions". For each use case the reference domain is indicated (electrical, sharing or mobility area), but it would be useful to specify whether it is a static, dynamic or open data use case, and how this is connected to the reference domain. This aspect is indeed emphasised in the final examples and, although it is described in Section 4.3, it would be useful to show this concept in the table as well.

In addition, some shortcomings in Chapter six "Ontology evaluation" should be underlined.
Even if the choices made are motivated throughout the process, and the good mapping of the domain of interest with the ontology makes it easy to verify the sensibility of these choices, there is little empirical evidence of the satisfaction of the evaluation criteria mentioned. While with regard to Accuracy and Adaptability, it can be sufficient to refer to de-facto standards or line guides of the experts consulted, in the case of Completeness, more evidence supporting the results must be provided. It might therefore be useful to provide at least a reference to a method of validation, as done for Consistency, or to add a reference to the performed checks.

This manuscript was submitted as 'Ontology Description'.

(1) Quality and relevance of the described ontology (convincing evidence must be provided).

This paper presents the Urban IoT ontologies, a set of ontology modules for sharing and electric mobility. The paper also describes the methodology followed for the development of the ontologies and a set of examples related to the instantiation of the ontology and queries. Moreover, the paper includes a section where it is described how the ontology is going to be applied in a real use case.

The motivation of the ontology and its relevance are clearly described in the paper, showing different use cases and involving domain experts within the development activities. Also, all the use cases and requirements are available online in a Github repository so that anyone can access that information. The paper also clearly describes the different activities performed during the development of the ontology (e.g., the requirement specification, implementation, publication...)

However, I would expect from this paper to compare the proposed ontologies with the other ontologies in the state of the art (which are described in 2) to show what are the current domains or issues that are not covered by existing ontologies. This can also be remarked in Section 5, by focusing the examples in key novelties of the proposed ontology (with regards to the state of the art ontologies) to clearly expose the novelty of the work.

I also think that it would be useful for the reader to further explain some decisions or particular patterns proposed in the ontology, e.g., Why if an organisation is a provider and an operator it is only instantiated as an organisation? (p. 12) or Why to use an abstract Geometry concept instead of GeoSPARQL? (p.10)

Minor issues:
- p. 5 Maybe it can be added at the end of the "Ontology publication" subsection that is explained in Section 7, as it is added in the other subsections.
- p.12 near by --> nearby

(2) Illustration, clarity and readability of the describing paper, which shall convey to the reader the key aspects of the described ontology.

The paper is well-structured and it is easy to read. It also includes diagrams of the ontologies to clarify the descriptions. As I mentioned before, Section 2 can be improved to show the novelties of the work in comparison with existing work, and also Section 5 can be improved by further describing the decisions taken to implement the ontologies.