A Reference Architecture for Ontology Engineering Web Environments/Una Arquitectura de Referencia para Ambientes Web de Ingenieria Ontoiogica.
Ontology authoring, maintenance and use are never easy tasks, mostly due to the complexity of real domains and how they dynamically change as well as different background possessed by modellers about methodologies and formal languages . However, although the needs for ontologies are well-understood, not less important is to provide editing tools to manipulate and understand them .
Ontology Engineering is a set of activities, methodologies, languages and tools related to the authoring, maintenance, refinement, evaluation and use of ontologies. Authoring is key for the ontology engineering and it requires to be backed by proper methodologies for guiding users through the building process. In particular, its complexity increases when such models should be created from scratch or when they should be merged to other ones in order to compose them. All of these activities appear fragmented across several tools adding a more important ingredient to the complexity of the ontologies life cycle. The lack of an adequate and seamless environment affects not only the uptake of ontologies by new users, but also the quality of artifacts created by modellers. Previous surveys [3, 4] reveal that a lot of ontology engineering environments have been developed, but too few of them are webbased systems and in almost all the cases they are mere visualisers, with limited graphical features and lacking of inference services. In fact, some of them claim no well-accepted framework for common authoring tasks exist because of a poor understanding of the effectiveness of tools .
In this direction, new architectures for ontology engineering environments must be prepared for envisioned future changes while being easily extendable and providing basic functionality. Thus, these architectures are for a family of systems and particular members of this family can be constructed based on a reference architecture containing explicit places where extensions can take place. This work proposes and documents a reference architecture for ontology engineering web environments. Moreover, it provides the rationale for boosting the collaborative development of a novel tool based on this architecture, named crowd. Documenting software architecture is essential to work cooperatively in solving larger problems than single groups could solve individually. At the same time, it helps to define and evaluate quality attributes, identify parts where systems could be extended and communicate baselines and design decisions made to let others use it properly [5, 6].
The rest of this article is structured as follows. Section 2 explains the motivation of this research work and previous results. Section 3 presents a survey on related architectures. Section 4 describes our proposed reference web architecture and section 5 gives details about the first version of crowd. Finally, in section 6, we summarise conclusions and discuss limitations and future works.
2 Motivation and Background
Misunderstandings about some property-based distinctions between ontologies and models have been undertaken by a lot of different proposals . These distinctions were mainly based on the needs of formalisation of conceptual models (p.e. UML, ER, ORM) for expressing ontologies and conversely the logic-based reasoning capabilities provided by ontology languages (p.e. OWL ) not captured by models. In particular, despite obvious differences between the expressiveness of conceptual modelling languages (CMLs) and ontology languages, many tools have partially validated this claim [9, 10, 11, 12]. Thus, the effectiveness of using graphical syntax based on these conceptual languages for expressing ontologies, even with complex languages, is still been considered for ontology engineering environments [3, 2]. Evidence for these claims indicate that the common core of UML, (E)ER and ORM 2 can be formalised in the [LAMBDA]CNI Description Logics (DL)  guaranteeing tractable reasoning over them . Even more, each language can express full ALCQI while keeping reasoning capabilities.
At the same time, OWL notations have been also proposed, which are mostly based on graphs [15, 16]. Aiming at the needs for an unified visual notations for ontologies, these approaches claim that switching between tool can be confusing given the diversity of visual representations. In addition, several tools implement just a few ontology engineering activities providing very different user interfaces, graphical and formal languages and reasoning capabilities. These last quotes give a real insight into going towards an unified tool integrating not only the visual notations and their common features, but also supporting other activities of the ontology engineering altogether in a single graphical and logical framework with a set of core functionalities.
In this context of lacking of a holistic view, a web tool called crowd [17, 18, 19] has been developed. crowd is a multi-view web environment for ontology development being supported by both Universidad Nacional del Comahue (1) and Universidad Nacional del Sur (2) in Argentina. The first intention behinds the tool is to assist users to author and edit ontologies and conceptual models adopting standard CMLs and employing complete logical reasoning to verify the satisfiability of specifications, infer implicit constraints and suggest new ones. The leverage of automated reasoning is enabled by a precise semantic definition of all the elements of the diagrams. Hence, diagrams constraints are internally translated into a logic-based formalism capturing typical features of models. To this end, the tool is fully integrated with a powerful logic-based reasoning server acting as a background inference engine. Moreover, since crowd is based on a deduction-complete notion of reasoning support relative to the diagram graphical syntax, users will see the original model graphically completed with all the deductions and expressed in the graphical language itself. This includes checking class and relationship consistency, discovering implied class and cardinality constraints. Empowered by web technologies, crowd has been designed as a scalable and maintainable architecture for adapting new reasoning, query and documenting engines, graphical languages and design methodologies. It has been conceived from scratch as a graphical-centric tool for ontology modelling, adopting standard languages and considering the possibility to expand its graphical primitives for more expressiveness. The first version is available at http://crowd.fi.uncoma.edu.ar together with its source code.
The road map followed to develop the first version of crowd considered the features and related work explained below:
(F1) Strong Theoretical Background of Graphical Tools. [19, 20] crowd is based on a well-founded structure, a heterogeneous algebra , describing a family of sets such as a metamodel M for a nonempty set of CMLs, a non-empty set L of the languages supported by M, a non-empty set of formal languages FL for ontologies and a non-empty set E of finitary operations for mapping primitives from L to an ontology language from FL (whenever possible).
(F2) Logic-Based Reasoning on Graphical Ontologies. [22, 9, 20] A full-flagged DL encoding of the CMLs enables crowd to offer standard logic reasoning on graphical ontologies as a way to help users along some of the ontology engineering activities. This tool allows to graphically support powerful OWL 2 features such as non-qualified and qualified cardinality restrictions for any n natural number.
(F3) Ontology-Based Metamodelling. Graphical views in crowd require a full specification of the similarities and mismatches of the CMLs. In this context, crowd is backed by a needed metamodel, named KF , for operations of conversion, transformation and approximation among models. Each ontology in crowd is an instance of the KF-metamodel.
(F4) Ontology Multi-View.  Based on the baselines of the KF-metamodel, crowd implements a multi-view GUI for switching between them and thus offering different ways to interact with the current model.
(F5) Export OWL 2. crowd enables exporting full descriptions of graphical ontologies in OWL syntax. It also provides OWLlink  specifications in order to get the inputs offered to off-the-shelf reasoners.
(F6) Import OWL 2. Importing in crowd is a two-step process looking for extracting as many graphical axioms as possible from ontologies in OWL documents. This novel process is supported by logic-based reasoning together with a SPARQL-DL  engine.
(F7) Ontology Documentation. Human-readable documentation of ontologies facilitates their understanding, reuse and adoption by third-parts. In this sense, crowd helps users documenting their ontologies by invoking Widoco  and producing a full, enriched version of their models.
(F8) Namespaces Definitions. crowd handles namespace definitions through a novel treatment of URIs and thus enabling the reuse of other vocabularies in the same graphical model. It provides a modal widget for prefixes and their values. Moreover, each graphical primitive presents a dedicated widget to enter its specific definition.
3 Related Architectures
Through literature and web search we have identified the following ontology editing tools that are, at least to some degree, actively maintained: WebProtege , Protege  -OWLViz (3), OntoGraf (4), SOVA (5), NORMA [10, 30], ICOM , TopBraid Composer , Graphol , OWLGrEd , Menthor , NeOn Toolkit , VOWL , GrOWL , OntoTrack , SWOOP , Hozo  and Graffo . Five of them (VOWL, Graphol, Hozo, GrOWL and Graffo) are mere ontology visualisers, while the remaining ones present (at least in the literature) some degree of interactively and integration of logical support with graphical models. Moreover, OntoTrack, GrOWL and SWOOP seem to be not publicly available to download or deprecated in order to run a demo of the software.
However, none of them presents both interactive and web support in the very same ontology engineering environment as proposed by our novel architecture. In the remaining, we conduct a comprehensive analysis over these tools (leaving OntoTrack, GrOWL and SWOOP out of the scope of this work) by addressing the following aspects: (1) operating mode meaning standalone or web (or both); (2) graphical language type - possible categories include ad-hoc node-link notations, UML/EER/ORM 1/ORM 2 or own languages; (3) graphical expressiveness - OWL 1, OWL 2, RDF or subsets of them, (4) non-graphical expressiveness - underlying ontology languages supported OWL 1, OWL 2, RDF or subsets of them; (5) integration of visual models and reasoning aiming at identify tool that graphically complete original models with all the deductions and expressed in the graphical language itself (whenever possible) ([check]x); and (6) multi-view support, where diagrams could be shown in more than one graphical language ([check]x). In particular, the aim of assessing such an integration is to pragmatically justify the needs for the knowledge visualisation process defined in . Specifically, the graphical and reasoning integration refers to a "back and forth" transformation between a graphical model and a logic formalisation capturing the semantics of this model.
As shown by the comparison, only OWLGrEd, VOWL and WebProtege offer some kind of web support. Nevertheless, the first two are mere visualisers of UML-like and VOWL language, respectively, and no editing possibilities. Conversely, WebProtege does provide editing support, laying a strong emphasis on collaborative development although visual notations are absent. Other weak aspect of the surveyed tools is the lack of multi-view. Multiple views of a model allow to interoperate them as well as share and communicate them among different stakeholders . Multiview in the context of a tool fully integrated with logic reasoning also require multiple ways to encode graphical models. This is also provided by the proposed environment and is absent in the related tools. Integration of visual models and reasoning is slightly provided by ICOM and NORMA (subsumptions, m..n, n = 1 cardinalities, disjunctions and equivalences) while in remaining surveyed tools no integration exist or only for subsumptions in few cases. With reference to ontology engineering activities, only the classic standalone Protege is strong in the use of ontologies because is highly extensible thanks to its very sophisticated plugin architecture. Table 1 shows the comparison.
4 crowd Architecture Views
In this work, we document and detail the architecture of crowd offering three views: modules, component and connectors and allocation. Each of them highlights different aspects of implementation and the underlying decisions for future and collaborative support. In each view, we explain its motivation and its intended meaning, and describe its principal parts based on the chosen styles. For all the cases, we use UML notation. Before introducing the architecture, we briefly explain some assumptions made. First, we consider modules as implementation units of software together with a set of responsibilities. Likewise, components represent elements such as objects, clients and servers. Finally, artifacts represent files and software packages to be installed and deployed in a running environment.
4.1 Module View
Motivation. The system is decomposed in implementation units or modules, which can be related by ispart-of, depend-on and is-a relations. Each module provides a responsibility set aiming at defining the role of such module in achieving certain functionality. This view details how these units ensemble to form one longer structure. Its intended use is to provide a blueprint for the source code. Architecture Styles. Decomposition, Uses and Layered.
Description. The module view of crowd is a coherent combination of architecture styles. First of all, crowd is divided into layers, each one providing a set of services to the other layer.
As depicted in Fig. 1, the upper layer groups a set of modules related to the user interface of the tool. At the same time, it is decomposed in implementation units describing a containment relation and the dependencies among them. The main module is the gui, which depends on the packages views and model, all of them for UML, EER and ORM support. The first one provides the definition of templates and their respective event handlers. The latter one manages the underlying graphical models, their primitives and the interaction with the graphical library by means of respective UML, ORM and EER factories. Two specific modules support common features: model.uris and views.common implements orthogonal functionalities for the treatment of namespaces and common templates and events of the GUI, respectively.
The bottom layer contains the modules required for the processing of ontologies: translators. builders, users management and interfaces with off-the-shelf tools. The main module of this layer is translator, which has a dual role orchestrating both the encoding of graphical model into logic and processing ontologies extracted from OWL documents using specific strategies and builders. strategies depend on CMLs and how they are encoded in logic, while builders give the structure of documents for different syntaxes. Lastly, translator contains qapackages, which provides queries and parses their answers when ontologies are processed by reasoners. To this end, reasoner connects crowd to reasoning systems and allows to create as many interfaces for other tools as desired. In this sense, reasoner implements connectors for the documentation tool widoco and a query engine sparql-dl. The decoder package is in charge of extracting ontologies from OWL documents by means of SPARQL-DL. It invokes the respective engine set with pattern queries that return structural aspects of the current ontology. This module implements key functionalities for managing URIs and also depend on a strategy and a builder. Altogether, translator, decoder and reasoner are used by the common module, which is responsible for properly combining them to built even more complex functionalities. Fig. 2 illustrates the interaction among objects in the OWL 2 importing scenario.
Finally, crowd includes simple modules for user and model management.
4.2 Component-and-Connector View
Motivation. This kind of views complements the previous one by showing elements (components) that have some run-time presence such as objects, and how these elements communicate each other by means of connectors (communication links, protocols). Connectors define pathways of interaction between components, which have a set of ports to connect them. The resultant view is a graph showing how the system works. Architecture Styles. Call-Return. Description. crowd has been conceived from scratch as a client-server implementation. In this sense, Client provides a user friendly interface to graphical conceptual modelling running on a web browser and requires reasoning services of the Server running on a web server. This tool follows the very same principles of the web, where clients access information from one or more servers distributed using the HTTP protocol, which is a form of request/reply invocation.
The Fig. 3 depicts a Component-and-Connector View (C&C) for the client side, its component and the communication scheme between them and can be described as followed. crowd supports widgets and events for multiples diagram objects for standard languages. They are accessible through the GUI component that handles different instances of GUIIMPL. Each one of such instances is the representation of a specific interface, its events, widgets and diagram. An Adapter component implements the Adapter structural pattern  for providing a bridge between an interface and a diagram (or Model), for instance an UML diagram and its GUI. The Widget Set manages all the widgets needed for a specific diagram and also provides common ones to all the languages. Lastly, a Factory component implements an Abstract Factory pattern . This factory defines an interface for creating the graphical elements of the standard languages of crowd. It delegates this task to the underlying library JointJS (6).
Moreover, the supported CMLs share some common components such as uris, for the full treatment of namespaces. An URI is an unique identifier for each primitive of the ontologies. crowd keeps the track of each primitive by specifying its (graphical) name, an abbreviation (prefix) and an URL. An comprehensive treatment of URIs is essential for the tool because ontologies must be available through internet .
The Fig. 4 schematises the server, its subcomponents and connectors of the environment. For simplicity of the diagram, only Reasoning elements are fully expanded while details of Import ones are hidden because they present similar C&C views. The highlighted part of this scheme (in green) indicates the importance of the logic-based reasoning systems in backing the services of importing OWL ontologies and reasoning over graphical models. In all the cases, off-the-shelf tools provide communicating interfaces or protocols such as OWL API  and OWLlink. Both Traslator and Decoder components connects to external engines. In particular, the OntoExtractor component represents run-time objects implementing the responsibilities defined for the decoder and sparql-dl modules. It is in charge of processing the outputs of SPARQL-DL queries sent to the respective engine, which aims at extracting both intentional (TBox) and assertional (ABox) axioms from OWL ontologies. Such queries are defined in the specification of SPARQL-DL language (7) (see Fig. 5) and are processed by an engine (8) built on top of existing reasoners and that can be fed with ontology documents in any of its linearisations (OWL/RDF, OWL/XML, between others).
Similarly, the components Strategy, Builder and QAPackage communicate each other and implement responsibilities associated to the module translator. Translator also connects to external reasoning systems though the OWLlink protocol so that they are fed with OWLlink files containing the OWL encoding of an graphical ontology together with a set of OWLlink queries for sanity checking of the ontology. The output of this reasoning task is post-processed by the QAPackage. Finally, the new OWL ontology is sent to a Decoder instance to be after compared against the original graphical one. In this sense, the decision to integrate two different ways to query ontologies (OWLlink and SPARQL-DL) is in fact pragmatic and is well-grounded in the well-known requirements of interoperability: the use of standard protocols as OWLlink to interconnect to other reasoning systems and the support for the importing of OWL ontologies, which requires of more expressive query languages to extract as much knowledge as possible from OWL documents. Fig 6 presents a sequence diagram describing a complete functionality of system and complements the C&C view highlighting the right interactions in run-time.
To summarise, Table 2 illustrates the mapping between C&C and Module views showing which modules in the Module view contribute to the implementation of components shown in the C&C view.
4.3 Allocation View
Motivation. Allocation views present a mapping between software and non-software elements. The deployment view depicts the physical configuration of the software architecture, in particular, how the components of the architecture are allocated to physical nodes. The install style maps the components of the architecture to the structures in the file system of the running environment; while the work assignment style shows the teams responsible for the development of each of the software modules.
Architecture Styles. Deployment and Install.
Description. crowd is a centralised architecture so that all the PHP artifacts, i.e. ".*php" files, are deployed in a single HTTP server running Apache (9) and MySQL (10) services. Off-the-shelf artifacts, ".*jar" files for reasoning systems and SPARQL-DL engines, are also installed in the same server and correspond to the Reasoner component. Currently, server is x86 Intel 3GHz, running Apache v2.4.10, MySQL v5.6.4 on a Debian GNU/Linux 8 jessie. In addition, PHP v7.0 and Java RE v1.8.0 are also deployed in such server. Finally, multiplicity N.. 1 indicates the number of instances of a client at each end of the communication path. Such clients represent web browsers, which interpret the HTML, CSS and JS artifacts, ".*html", ".*css" and ".*js" files of the user interface.
5 Evaluation: crowd v0.9
First version of crowd has been deployed at http: //crowd.fi.uncoma.edu.ar. Its implementation is the first one so that there are not yet performance assessments of the overall system. The initial steps of the development have been reported in [17, 18].
Currently, the crowd editor presents a workspace for editing visual models in UML: classes, attributes, generalisation and binary associations with n..m cardinalities, for n, m [greater than or equal to] 1. Visual models can be encoded in DL (and serialised in OWL 2) to be sent them to reasoning systems for checking satisfiability and properties refinement. Released version of crowd manages URIs and enables definition of namespaces for re-using of other ontologies available across the web. Users can log-in into crowd to host their diagrams although modelling tasks can be done but no state is saved in our server. The interface is zoomable, that is, the level of detail and size of the icons that represent the model can be smoothly changed by pressing dedicated options, which define a scale from 25% to 125%. This allows the user to visualise big ontologies into the same windows.
Complementary evaluations and applications of our reference architecture beyond crowd have been recently published in [42, 43]. The first one implements a tool for validating visual Orthogonal Variability Models (OVM) in the context of Software Product Lines (SPL). The second one implements a Visual Query Language (VQL) based in UML for the SPARQL-DL language.
6 Conclusions and Future Works
In this article, we have proposed and documented a reference architecture for a web ontology engineering environment. We described in details three complementary views merging different styles for each one: modules (decomposition, layered, uses), componentand-connectors (call-return) and allocation (deployment and install). The architecture described in this paper addresses specific needs. In particular, decreasing the cognitive effort of users when understanding, integrating and using of ontologies. Moreover, it aims at providing the rationale for boosting the collaborative development of a novel tool based on this architecture, named crowd.
Some limitations have been identified. CMLs languages are not enough expressive considering the underlying expressiveness of reasoning systems. Users could not visually represent disjoint and equivalence constraints, among others logical axioms. Due to this, we have considered two possible solutions: allowing users to define logical expressions, which requires knowledge about formal systems, or providing new visual primitives with a precise semantic definition. In this sense, crowd implements a text area for introducing OWL 2 statements as a temporary solution.
Future work involves releasing a new version of crowd implementing the remaining primitives of UML and (E)ER and full reasoning over these diagrams. Moreover, we expect to extend crowd to other engineering activities such as alignment and merging of ontologies as well as managing of instances and increasing compliance with W3C recommendations.
The authors have declared that no competing interests exist.
 R. Mizoguchi, Ontology Engineering Environments, pp. 275-295. Springer, 2004.
 M. Vigo, S. Bail, C. Jay, and R. Stevens, "Overcoming the Pitfalls of Ontology Authoring: Strategies and Implications for Tool Design," International Journal of Human Computer Studies, 2014.
 N. Achich, B. Bouaziz, A. Algergawy, and F. Gargouri, "Ontology visualization: An overview," in Intelligent Systems Design and Applications - 17th ISDA.
 N. M. Meenachi and M. S. Baba, "Article: Web Ontology Language Editors for Semantic WebA Survey," International Journal of Computer Applications, 2012.
 D. Garlan, F. Bachmann, J. Ivers, J. Stafford, L. Bass, P. Clements, and P. Merson, Documenting Software Architectures: Views and Beyond. 2nd ed., 2010.
 H. Gomaa, Software Modeling and Design: UML, Use Cases, Patterns, and Software Architectures. Cambridge University Press, 1st ed., 2011.
 C. Atkinson, M. Gutheil, and K. Kiko, "On the Relationship of Ontologies and Models," in Proceedings of the 2nd WoMM, 2006.
 P. Hitzler, M. Krotzsch, B. Parsia, P. F. PatelSchneider, and S. Rudolph, eds., OWL 2 Web Ontology Language: Primer. W3C Recommendation, 27 October 2009. Available at http: //www.w3.org/TR/owl2-primer/.
 P. Fillottrani, E. Franconi, and S. Tessaris, "The ICOM 3.0 intelligent conceptual modelling tool and methodology," Semantic Web, 2012.
 M. Curland and T. A. Halpin, "The NORMA Software Tool for ORM 2," in CAiSE Forum, Lecture Notes in Business Information Processing, Springer, 2010.
 K. Cerans, J. Ovcinnikova, R. Liepins, and A. Sprogis, "Advanced OWL 2.0 Ontology Visualization in OWLGrEd," in DB&IS, 2012.
 R. Hodrob and M. Jarrar, "On Using a Graphical Notation in Ontology Engineering," 2012.
 F. Baader, D. Calvanese, D. L. McGuinness, D. Nardi, and P. F. Patel-Schneider, eds., The Description Logic Handbook: Theory, Implementation, and Applications. Cambridge University Press, 2003.
 P. R. Fillottrani and C. M. Keet, "Evidence-based languages for conceptual data modelling profiles," in Proceeding of the 19th ADBIS, 2015.
 S. Negru, F. Haag, and S. Lohmann, "Towards a unified visual notation for owl ontologies: Insights from a comparative user study," in Proceedings of the 9th ICSS, 2013.
 M. Console, D. Lembo, V. Santarelli, and D. F. Savo, "Graphol: Ontology representation through diagrams," in Informal Proceedings of the 27th International DL, 2014.
 C. Gimenez, G. Braun, L. Cecchi, and L. Fillottrani, "crowd: A Tool for Conceptual Modelling assisted by Automated Reasoning - Preliminary Report," in Proc. of the 2nd SAOA@JAIIO, 2016.
 G. Braun, C. Gimenez, P. Fillottrani, and L. Cecchi, "Towards conceptual modelling interoperability in a web tool for ontology engineering," in 3rd SAOA@JAIIO, 2017.
 G. Braun, C. Gimenez, L. Cecchi, and P. Fillottrani, "Towards a visualisation process for ontology-based conceptual modelling," in Proc. of the IX ONTOBRAS, 2016.
 G. Braun, L. Cecchi, and P. Fillottrani, "Integrating graphical support with reasoning in a methodology for ontology evolution," in Proc. of the 1st JOWO@IJCAI, 2015.
 G. Birkhoff and J. D. Lipson, "Heterogeneous algebras," Journal of Combinatorial Theory, 1970.
 D. Berardi, D. Calvanese, and G. De Giacomo, "Reasoning on UML class diagrams," Artif. Intell., 2005.
 C. M. Keet and P. R. Fillottrani, "An ontologydriven unifying metamodel of UML class diagrams, eer, and ORM2," Data Knowl. Eng., 2015.
 T. Liebig, M. Luther, O. Noppens, and M. Wessel, "Owllink," Semantic Web, 2011.
 E. Sirin and B. Parsia, "Sparql-dl: Sparql query for owl-dl," in In 3rd OWL Experiences and Directions Workshop (OWLED-2007, 2007.
 D. Garijo, "WIDOCO: A Wizard for Documenting Ontologies," in ISWC 2017, 2017.
 G. Guizzardi, Ontological foundations for structural conceptual models. PhD thesis, University of Twente, 2005.
 T. Tudorache, C. Nyulas, N. F. Noy, and M. A. Musen, "Webprotege: A collaborative ontology editor and knowledge acquisition tool for the web," Semantic Web, 2013.
 H. Knublauch, R. Fergerson, N. Noy, and M. Musen, "The Protege OWL plugin: An open development environment for semantic web applications," 2004.
 F. Sportelli, "NORMA: A software for intelligent conceptual modeling," in Proceedings of FOIS, 2016.
 TopQuadrant, "TopQuadrant--Products--TopBraid Composer," 2011.
 J. L. R. Moreira, T. P. Sales, J. Guerson, B. F. B. Braga, F. Brasileiro, and V. Sobral, "Menthor editor: An ontology-driven conceptual modeling platform," in JOWO@FOIS, 2016.
 P. Hasse, H. Lewen, R. Studer, and M. Erdmann, "The NeOn Ontology Engineering Toolkit," 2008.
 S. Lohmann, S. Negru, and D. Bold, "The protegevowl plugin: Ontology visualization for everyone," in The Semantic Web: ESWC 2014, 2014.
 S. Krivov, R. Williams, and F. Villa, "GrOWL: A tool for visualization and editing of OWL ontologies," J. Web Sem., 2007.
 T. Liebig, "Ontotrack: Fast browsing and easy editing of large ontologies," in In Proceedings of 2nd EON@ISWC, 2003.
 A. Kalyanput, B. Parsia, E. Sirin, B. Grau, and J. Hendler, "Swoop: A 'web' ontology editing browser," Journal of Web Semantics, 2005.
 K. Kozaki, Y. Kitamura, M. Ikeda, and R. Mizoguchi, "Hozo: An Environment for Building/Using Ontologies Based on a Fundamental Consideration of "Role" and "Relationship"," in 13th EKAW Conference, 2002.
 R. Falco, A. Gangemi, S. Peroni, D. M. Shotton, and F. Vitali, "Modelling OWL ontologies with graffoo," in The Semantic Web: ESWC 2014, 2014.
 E. Gamma, R. Helm, R. Johnson, and J. Vlissides, Design Patterns: Elements of Reusable Object-oriented Software. Addison-Wesley Longman Publishing Co., Inc., 1995.
 M. Horridge and S. Bechhofer, "The owl api: A java api for owl ontologies," Semant. web, 2011.
 n. Oyarzun, G. Braun, L. Cecchi, and P. Fillottrani, "A Graphical Web Tool with DL-based Reasoning," in Proc. of the XXVICACIC, 2018.
 C. Gimenez, G. Braun, L. Cecchi, and P. Fillottrani, "Towards a Visual SPARQL-DL Query Builder," in Proc. of the XXVI CACIC, 2018.
Citation: G. Braun, E. Estevez and P. Fillottrani. "A Reference Architecture for Ontology Engineering Web Environments", Journal of Computer Science & Technology, vol. 19, no. 1, pp. 22-31, 2019.
German Braun (1,3), Elsa Estevez (2,3), and Pablo Fillottrani (2,4)
(1) Universidad Nacional del Comahue, (email@example.com
(2) Universidad Nacional del Sur, (ece, firstname.lastname@example.org
(3) Consejo Nacional de Investigaciones Cientificas y Tecnicas (CONICET), Argentina.
(4) Comisi0n de Investigaciones Cientificas (CIC), Provincia de Buenos Aires, Argentina.
Received: July 18, 2018 Accepted: November 26, 2018.
(3) http://protegewiki.stanford.edu/wiki/OWLViz accessed July 2018
(4) http://protegewiki.stanford.edu/wiki/OntoGraf accessed July 2018
(5) http://protegewiki.stanford.edu/wiki/SOVA accessed July 2018
Table 1: Overview of ontology editors considering the criteria proposed. Web Graphical Graphical Standalone Language Expressiveness WebProtege web x x Protege standalone ad-hoc Node-link subsumption OWLViz Protege standalone ad-hoc Node-link subsumption OntoGraf Protege standalone ad-hoc Node-link OWL 1 SOVA NORMA standalone ORM 2 ALCQI ICOM standalone EER ALCI TopBraid standalone ad-hoc Node-link/ OWL 2 subset Composer UML OWLGrEd both Extended-UML OWL 2 subset eddy - standalone Grapho SROIQ(D) Graphol *own language Menthor standalone OntoUML OntoUML *own language () NeOn Toolkit standalone ad-hoc Node-link OWL 2 VOWL both VOWL OWL 2 *own language Hozo standalone ad-hoc Node-link role-concepts & wholeness/relation concepts Graffo standalone ad-hoc Node-link OWL 2 Non-Graphical Graphical and Multi Expressiveness Integration View WebProtege OWL 1/OWL 2 x x Protege OWL 1/OWL 2 [check] x OWLViz *only subsumption Protege OWL 1/OWL 2 x x OntoGraf [check] Protege OWL 2 *only x SOVA subsumption NORMA ALCQI [check] x ICOM ALCQHI [check] x TopBraid OWL 2 [check] [check] Composer *only subsumption OWLGrEd OWL 2 x x eddy - SROIQ(D) x x Graphol Menthor UFO  x x [check] NeOn Toolkit OWL 2 *only x subsumption VOWL OWL 2 x x Hozo role-concepts & x x wholeness/relation concepts Graffo OWL 2 x x Table 2: Mapping between C&C and Module Views for both client and server. (*) represent modules as part of more than one run-time component. C&C View Module View Model, Factory, Uris model GUI, GUIIMPL, Adapter gui Widgets Set, Events Set views *:Translator common, translator *:Decoder common, decoder OntoExtractor decoder Strategy strategies *:Builder builder QAPackage qapackages Off-the-shelf widoco Docs Generator Off-the-shelf Reasoner, reasoner SPARQL-DL engine User Management users
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||ORIGINAL ARTICLE|
|Author:||Braun, German; Estevez, Elsa; Fillottrani, Pablo|
|Publication:||Journal of Computer Science & Technology|
|Date:||Apr 1, 2019|
|Previous Article:||Wavelets Defined Over Non Nested Tetrahedral Grids: A Theoretical Approach/Wavelets Definidas sobre Grillas Tetraedricas no Anidadas: un Abordaje...|
|Next Article:||Burst Error Analysis Introduced in Multiple Traffic of Protocols TCP Reno, Cubic, Westwood and Vegas on a Model of Hybrid Topology/Analisis de Error...|