Modeling smart apiculture ecosystem: An ontology-based approach

This paper addresses the topic of modeling a smart apiculture ecosystem using an ontology-centric approach grounded in Semantic Web standards. The authors propose a domain-specific ontology, BEESNN, as an extension of the W3C SSN/SOSA model, and demonstrate its application within a prototype beekeeping management platform.
The advantage of this work lies in its specificity, as it differs from most smart agriculture ontologies that focus on crops, irrigation, or livestock. This ontology targets the niche domain of beekeeping. It leverages established standards (SSN, SOSA, DUL), which is commendable and facilitates semantic interoperability. The authors go beyond data modeling by incorporating system-level concepts such as sensors, APIs, cloud platforms, alerts, and stakeholder interactions. The inclusion of a functional IT platform prototype adds further practical value.
However, the article suffers from several critical weaknesses that significantly limit its scientific rigor and impact:
The authors do not provide any formal evaluation of the proposed ontology. There are no metrics (e.g., domain coverage, consistency), no expert review, no scenario-based testing, and no validation with end-users. The ontology is presented as-is, without demonstrating that it functions effectively in practice.
Furthermore, the authors do not evaluate the completeness, correctness, or adequacy of the proposed ontology in any formal way. Instead, they merely list potential use cases (e.g., data integration, hive monitoring, alerting), but these scenarios are presented without systematic testing or validation. The assumption that the ontology fulfills its intended purpose is taken for granted. No empirical evidence or comparison is offered to demonstrate that BEESNN improves interoperability, enhances decision-making, or supports standardization more effectively than alternative models.
Use-case illustrations are not a substitute for validation, and do not guarantee that the ontology is complete, semantically accurate, or usable across contexts. This is a significant limitation that undermines the claim of the ontology being foundational for smart apiculture ecosystems.
The article fails to widely survey or compare BEESNN with existing ontologies or conceptual models relevant to apiculture. Notably, initiatives such as the BeeKEEP ontology and the BeeOpen framework, which address aspects of hive monitoring, environmental influence, and open data sharing, are either omitted or only briefly mentioned. The lack of comparative analysis diminishes the perceived novelty and justification for BEESNN.
Also it does not specified any development methodology (e.g., METHONTOLOGY, NeOn, OntoUML) nor justify design choices such as class hierarchies, relationships, or naming conventions. There is no evidence of domain expert involvement or iterative validation, making it difficult to assess the quality and reusability of the ontology.
Although the ontology is presented as a knowledge structure, it lacks semantic expressivity. The authors do not define any formal rules (e.g., OWL axioms, SWRL, SHACL) or reasoning mechanisms that would support automated alerts, recommendations, or logic-based inference—despite this being a key value proposition in a smart IoT environment.
BEESNN’s potential for reuse in broader agricultural domains (e.g., livestock, orchards) or integration with existing ontologies (e.g., AGROVOC, AgOnt, FIWARE agri-models) is not discussed. Nor is there any indication of plans to publish the ontology in public repositories such as AgroPortal or BioPortal, which would enhance its interoperability and discoverability.
The paper does not address the practical limitations of the proposed model. Issues such as data evolution over time, ontology versioning, scalability of deployment, or connectivity challenges (e.g., LoRa reliability in rural contexts) are overlooked. The portrayal of the system is overly optimistic and lacks realistic constraints or risk assessment.
While the article discusses general trends in smart agriculture and IoT integration, the literature review on ontologies specific to apiculture or domain-specific modeling is limited. The omission of key related work in semantic agriculture undermines the theoretical depth of the paper.
In conclusions, the paper presents an interesting and much-needed contribution in the form of a specialized ontology for smart beekeeping. BEESNN is reasonably well-aligned with existing standards and demonstrates strong modeling capabilities across technical and environmental aspects. However, the article remains primarily conceptual, and lacks the necessary methodological rigor, empirical validation, and critical engagement with existing ontological resources. To solidify its impact, future work should validate the ontology empirically (e.g., via expert review or use-case testing), incorporate semantic reasoning mechanisms, extend the literature review to better position BEESNN within the existing landscape, and consider publishing the ontology in an open repository with formal documentation.

The manuscript presents an ontology-based solution for the apiculture domain, which represents an innovative and potentially valuable contribution to this specific field. However, the paper currently combines elements of ontology description with system implementation, and the latter does not clearly constitute a novel research contribution. While the engineering effort behind the system is acknowledged, it remains unclear how this work advances the state of the art in practice. Furthermore, the system itself has not been published as a standalone resource, limiting reproducibility and broader community use (only ontology is).

Regarding the contribution to the semantic web domain, the work appears limited. The ontology primarily reuses existing frameworks such as SSN and DOLCE and introduces several new concepts. The methodology employed for ontology engineering is not explicitly stated. Additionally, for certain concepts related to environment, climate, and software/cloud infrastructures, it is unclear why bee-domain-specific modeling was necessary rather than reusing existing ontologies from general agriculture or software engineering domains. The apiculture-related part itself is quite limited in the resource. Although the resource is available on GitHub, it has not undergone formal validation and evaluation (in practice or even with pitfall scanning tools such as OOPS!).

The manuscript discusses leveraging the ontology to integrate data from multiple domains. However, due to re-engineering of existing ontologies and potential divergence from widely adopted standards, the interoperability of the proposed resource is likely reduced, which in turn limits the practical impact of the work.

Finally, the manuscript’s writing quality requires significant improvement. The narrative is often repetitive and lacks a clear, coherent structure, particularly in the abstract and introduction. These sections should clearly articulate the storyflow, with motivation, research questions, challenges, approach, and contributions in a logical sequence. Terminology use should be precise—for example, avoiding phrases such as “artificial intelligence and machine learning” (the latter is a subset of the former). The text also contains grammatical errors and typographical mistakes (e.g., “acters”), as well as informal expressions such as “smart systems,” which are vague and uncommon in scholarly literature. Abbreviations (e.g. AI, IoT) should be made once at the start and used through the text.

Overall, while the topic holds good promise, the paper would benefit from clearer methodological exposition, larger connectedness to apiculture and agriculture-related domains and larger elaboration of this part of the ontology, stronger demonstration of novelty and impact, improved interoperability considerations, and enhanced writing quality.