PerfectO: An Online Toolkit for Improving Quality …
163
cations by modeling a specific domain. Many academic and industrial communities
now understand the need for ontologies.
The design, integration, and re-usability of ontologies are widely encouraged by
the semantic web community. However, in other communities (e.g., IoT, Sensor
Web, Semantic Sensor Networks), reusing ontologies is still a significant challenge
where most of the time, similar ontologies are re-designed regularly, and not aligned
with existing ontologies. There is a need to reference and classify those ontologies.
Ontology catalogs such as Linked Open Vocabularies (LOV) [4], LOV for the Internet
of Things (LOV4IoT) [5–7], and BioPortal for medical ontologies encourage the
reuse of ontologies. However, interoperability issues remain due to the lack of proper
guidance to acquire knowledge for best practice and the lack of time.
Helping ontology designers in improving ontologies is essential for the following
reasons: (1) Reusing domain knowledge in other projects/domains because most of
the time the knowledge is already conceived and implemented as ontologies. (2)
Reducing the exhaustive job of re-inventing ontologies. (3) Reducing development
time. (4) Increasing semantic interoperability among systems when machines share
the same ontology. (5) Providing cross-domain interoperability, the same ontology
can be used in different domains. For instance, a smart home ontology is relevant
in healthcare for Ambient Assisted Living (AAL) to help elderly people to stay
independent at home.
Improving the quality of knowledge representation is gaining more attention,
existing surveys are: (1) knowledge graphs in [1], (2) Linked Data in [8], and (3)
ontology quality in [9–12]. Fernandez-Lopez et al. [13] study the LOV ontology
catalog and demonstrate that: (1) 36% of the ontologies registered in LOV are not
appropriately be loaded, (2) 78% of ontologies reuse W3C endorsed ontologies, and
(3) 27% of the LOV ontologies do not have English-language identifiers such as
labels and comments, or within URIs. Easy-to-use ontology improvement tools are
still needed to ease ontology designers’ life. We focus on the ontology quality by
foster a subset of tools to encourage ontology reuse.
Contributions: Our proposed solution, called PerfectO (stands for Perfect
Ontology), relies on organizing and helps improve the existing knowledge rather
than generating more. PerfectO’s primary objective is to increase the ontology
engineer/designer experience and facilitate the ontology quality improvement process by selecting a subset of useful validation tools providing web services or online
GUIs. PerfectO enhances ontology experience, accessibility, ontology classification, ontology interlining, and the optimization of the available ontologies. In
this chapter, we use PerfectO implementation in the Internet of Things domain
enhancing ontologies from Linked Open Vocabularies for IoT (LOV4IoT) ontology
catalog as a demonstration of how PerfectO can be used beyond the semantic web
domain. PerfectO LOV4IoT use case is included, and it is deployed as the proof
of concept and experimental validation in other fields.
Structure of the chapter: The rest of the chapter is structured as follows: Sect. 2
highlights the need to help ontology designers in improving ontology quality, accessibility, and knowledge classification. Section 3 investigates the literature to enhance
ontologies by covering complementary research topics: quality, evaluation, method-
163
cations by modeling a specific domain. Many academic and industrial communities
now understand the need for ontologies.
The design, integration, and re-usability of ontologies are widely encouraged by
the semantic web community. However, in other communities (e.g., IoT, Sensor
Web, Semantic Sensor Networks), reusing ontologies is still a significant challenge
where most of the time, similar ontologies are re-designed regularly, and not aligned
with existing ontologies. There is a need to reference and classify those ontologies.
Ontology catalogs such as Linked Open Vocabularies (LOV) [4], LOV for the Internet
of Things (LOV4IoT) [5–7], and BioPortal for medical ontologies encourage the
reuse of ontologies. However, interoperability issues remain due to the lack of proper
guidance to acquire knowledge for best practice and the lack of time.
Helping ontology designers in improving ontologies is essential for the following
reasons: (1) Reusing domain knowledge in other projects/domains because most of
the time the knowledge is already conceived and implemented as ontologies. (2)
Reducing the exhaustive job of re-inventing ontologies. (3) Reducing development
time. (4) Increasing semantic interoperability among systems when machines share
the same ontology. (5) Providing cross-domain interoperability, the same ontology
can be used in different domains. For instance, a smart home ontology is relevant
in healthcare for Ambient Assisted Living (AAL) to help elderly people to stay
independent at home.
Improving the quality of knowledge representation is gaining more attention,
existing surveys are: (1) knowledge graphs in [1], (2) Linked Data in [8], and (3)
ontology quality in [9–12]. Fernandez-Lopez et al. [13] study the LOV ontology
catalog and demonstrate that: (1) 36% of the ontologies registered in LOV are not
appropriately be loaded, (2) 78% of ontologies reuse W3C endorsed ontologies, and
(3) 27% of the LOV ontologies do not have English-language identifiers such as
labels and comments, or within URIs. Easy-to-use ontology improvement tools are
still needed to ease ontology designers’ life. We focus on the ontology quality by
foster a subset of tools to encourage ontology reuse.
Contributions: Our proposed solution, called PerfectO (stands for Perfect
Ontology), relies on organizing and helps improve the existing knowledge rather
than generating more. PerfectO’s primary objective is to increase the ontology
engineer/designer experience and facilitate the ontology quality improvement process by selecting a subset of useful validation tools providing web services or online
GUIs. PerfectO enhances ontology experience, accessibility, ontology classification, ontology interlining, and the optimization of the available ontologies. In
this chapter, we use PerfectO implementation in the Internet of Things domain
enhancing ontologies from Linked Open Vocabularies for IoT (LOV4IoT) ontology
catalog as a demonstration of how PerfectO can be used beyond the semantic web
domain. PerfectO LOV4IoT use case is included, and it is deployed as the proof
of concept and experimental validation in other fields.
Structure of the chapter: The rest of the chapter is structured as follows: Sect. 2
highlights the need to help ontology designers in improving ontology quality, accessibility, and knowledge classification. Section 3 investigates the literature to enhance
ontologies by covering complementary research topics: quality, evaluation, method-