A Semantic Meta-Model for Data Integration and Exploitation in Precision Agriculture and Livestock Farming

Tracking #: 3087-4301

Dimitris Zeginis
Evangelos Kalampokis
Raúl Palma
Rob Atkinson
Konstantinos A. Tarabanis

Responsible editor: 
Guest Editors Global Food System 2021

Submission type: 
Full Paper
At the domains of agriculture and livestock farming a huge amount of data are produced through numerous heterogeneous sources including sensor data, weather/climate data, statistical and government data, drone/satellite imagery, video, and maps. This plethora of data can be used at precision agriculture and precision livestock farming in order to provide predictive insights in farming operations, drive real-time operational decisions, and redesign business processes. The predictive power of the data can be further boosted if data from diverse sources are integrated and processed together, thus providing more unexplored insights. However, the exploitation and integration of data exploited at precision agriculture is not straightforward since they: i) cannot be easily discovered across the numerous heterogeneous sources and ii) use different structural and naming conventions hindering their interoperability. The aim of this paper is to: i) study the characteristics of data used in precision agriculture & livestock farming and ii) study the user requirements related to data modeling and processing from nine real cases at the agriculture, livestock farming and aquaculture domains and iii) propose a semantic meta-model that is based on W3C standards (DCAT, PROV-O and QB vocabulary) in order to enable the definition of metadata that facilitate the discovery, exploration, integration and accessing of data in the domain.
Full PDF Version: 

Minor Revision

Solicited Reviews:
Click to Expand/Collapse
Review #1
Anonymous submitted on 14/Apr/2022
Minor Revision
Review Comment:

The responses that were provided by the authors answered most of the doubts that I had in the initial review.

In general, this paper is well written and in the main is precise, concise and information dense. I could not find any obvious grammar errors or typos. However, occasionally the paper resorts to hyperbolic language such as ‘huge volume ‘, try not to use this type of language and substitute these terms for more academic language.

The use of industry partners should make the resources applicable to precision agriculture, however, as the consortium does not have any of the major players in the agribusiness market there is some doubt whether this standard will be adopted after the project has been completed.

The only concern I have at the moment is the selection of the case studies for the paper. It would aid the quality of the paper to have some text explaining why these case studies were chosen. The selection as it stands is it a little ad-hoc. It could be argued that the methodology is suited to these studies, but not to others. Consequently, some text is required to justify these choices. Simply stating that they are real cases is not sufficient.

It should be noted that it was not possible to evaluate the resources, as the server returned a 500 error. The URL posted is for redirection only, so I can’t be sure of the suitability of the storage location, as I suspect it is a private server (possibly in a university) with a redirection. Try to have a more permanent home for the resources, as they will need to outlive the project and the people who worked on it.

Apart from the quibble about the case studies, I think that the revised paper is suitable for publication.

Review #2
By Christopher Brewster submitted on 23/Apr/2022
Review Comment:

The authors have made significant improvements which justify acceptance at this stage.
A couple of minor issues to be looked at by the authors are:
1. Although the English is generally well written there are many minor irritations (e.g. misuse of preposition "at" (where it should be "in" or "on" often)
2. Web links do not work e.g. http://w3id.org/cybele/model (see Long Term Stable Link to Resources) but alos other links in the paper.
3. Please check all references as being complete (some are not).

Review #3
Anonymous submitted on 02/May/2022
Major Revision
Review Comment:

This paper has made some improvements over its original form. Principally, it added more backgrounds of the way of proposing the model, e.g., involved institutions and survey table, and more technical discussions on interoperability. I read through authors' responses to all reviewers' comments as well as the revised paper. Frankly speaking, I do not think the authors fundamentally addressed my, as well as the other two reviewers'(especially the first reviewer's) comments. Plus, there are many duplicated responses across the letter, and they are very general, even though IMHO the comments asked different things.

More concretely, I am barely satisfied by the response to my comment related to how "specified/special" such a model is for precision agriculture and livestock farming. My comment was more to challenge the motivation and innovation of the model. Put differently, if one was asked to create a meta-model for, say urban-related data, I think a very similar model can be proposed. Then what is "new" about the proposed meta-model? In my humble opinion, the problem might boil down to the fact that the meta-model is largely a "merging" of different ontologies. There is nothing wrong about it. I totally support reusing rather than creating something new if there already exits something we can apply. But I am skeptical about the originality of such a paper if its main contribution is this meta-model. Simply put, I believe this paper can be regarded as (maybe) a "best practice" of using DCAT, PROV, and QB for metadata modeling in agriculture and livestock farming, with some minor model modifications (the SHACL shape to connect DCAT and QB for example). Then, the authors might consider to more focus on the use cases rather than the detailed technical modeling. Plus, I believe my concern echoes Reviewer 1's comment. The authors' very general responses using (1) the four categories of metadata (already widely known), (2) the support of "data alignment" (did not specifically discuss though), and (3) the support of "interoperability" (similar to alignment actually; too general that one can fairly say all semantic models have the same goal), are hardly satisfactory in my opinion.

This concern is also reflected in the fact that many contents (sections) in the paper are loosely connected with the proposed meta-model. For example, even though the survey form is provided, how does it specifically help to build the model? For example, is there any useful answer that persuade you to make some specific modeling decisions in the domain of agriculture and livestock farming while using DCAT/PROV (the paper actually mentioned a little bit on it, but I am still curious how different it would be if the domain changes)? How does the category of different data sources in Section 4.2 related to the modeling? The authors states that "... to understand the nature and structure of the data... not intended to provide a taxonomy of the data". Then how does such an understanding helps build the model?

Plus, in related work, the four categories of metadata have already been discussed, why in Section 5.1, it is discussed again? Also, I disagree with the author that a diagram of the involved conceptual model is unnecessary and can already be seen from Fig. 2. I believe they should be quite different things. For example, what is the conceptual contribution of this paper rather than merging several different existing models (like Fig 2)? In section 4.3 (line 39-46), aren't the three identified scenario/cases to summarize the benefits/motivations of the proposed model the same (i.e., integrate and query heterogeneous data)?

In summary, this paper introduced many detailed and technical aspects of the project, which is exciting indeed. However, I really encourage the authors to think what they really want to highlight in this one paper and what is the major contribution they want to present to the community of agriculture and livestock farming specifically.