This article analyzes the existing technologies and systems for displaying ontologies, presents the features and shows the need to create a new web-based visualization system of this data structure. The display structures analysis is performed and the ontology representation peculiarities in the form of frame models and application of Simple Knowledge Organization System (SKOS) for the Network router concept are shown. The ontology construction stages of the subject area are outlined and described, which include: formation of the list of concepts included in the ontological structure of the upper level; ordering the list of concepts; creation of a glossary; mapping arcs between created concepts. An example of building an ontology for the subject area «information security» in describing the concept features of «network computer attack». The main functions of the system are highlighted and the stage of its design is performed using the object-oriented approach and UML language tools. The diagram of precedents which visualizes the basic variants of the developed system usage is shown. The construction architecture was chosen in favor of three-tiered client-server architecture and the peculiarities of interaction between levels are described. The development language choice peculiarities are given and the choice of libraries and frameworks is proved, which ensure its effective functioning. The result of the work is a prototype of a software system for visualization of ontologies, which can be used to visualize the specified data structure. The study creates the prerequisites for building a full-featured system of visualization of ontologies, to be part of the decision support system for a particular subject area.
Informatization and information technologies have covered almost all spheres of life in modern
society. It is impossible to name at least one branch of human activity, where their use would not be in
demand. Ontologies are one of the best tools for a formal presentation of knowledge about a particular
subject area. Most often, the concept of ontology is expressed by defining the basic objects (individuals,
attributes, concepts) and the relationship between them, but the analysis of complex concepts may
require considerable effort and time [
Visualization of ontologies is an important aspect of their practical use. Qualitative visualization
provides experts with an understanding of ontology or its fragments (ontology comprehension) [
In the context of formal ontologies visualization, the problem of logical conclusions results visualization is also an important task. Its essence is to create a visual representation that can illustrate the derivation of logical statements, to explain the results of derivation. Thanks to this idea, the developer can better understand the ontology, as well as quickly find and correct problematic axioms that lead to semantic conflicts.
The study showed that heterogeneous knowledge in intelligent information systems is poorly
formalized, which complicates the use of information systems for the complex objects of operational
management and processes and their evaluation in terms of staff. At the same time, the use of visual
images contributes to a sharp increase in perception and allows you to quickly and effectively convey
your thoughts and ideas. Numerous studies [
The authors [
As can be seen from the analysis, this area is extremely important but not fully explored, given that the important and urgent task is to create a system of visualization of ontologies of the subject area, which would allow users to display this structure and adjust playback modes depending on end user requirements.
The aim of the research is to create the subject area ontology visualization system on the example of ontology models. The study will provide tools for dynamic display of different types of ontographs in accordance with the formed criteria of visualization and reproduction of their properties for further use in information systems for operational management of objects and the end user.
To achieve this goal it is necessary to solve the following main tasks: to analyze existing research and systems used in the field of ontology visualization; identify the main tasks that arise, build models of the visualization process and implement the proposed approach using modern software.
The results of the study solve the current scientific and practical problem of creating software for visualization of the ontology of the subject area, which would provide the simplest and most adaptive mechanism for displaying these ontologies of any structure and format.
Traditionally, within computer display, ontology is used as a mean of computer processing
knowledge. However, with the growth of formalized knowledge accumulated on media, there is a
problem of their effective transfer to other systems and user-experts. In this case, the cognitive
properties of this structure come to the fore, which directly affect the simplicity and accuracy of their
interpretation in order to obtain a concept [
Given the analysis, two possible ways of presenting ontologies were identified: in the form of frames
and the use of the Simple Knowledge Organization System (SKOS). If we consider the frame
representation, it can be represented as a visualized fragment of the ontology, which provides the means
to display the structure for further transmission to the stakeholder. In this case, the frame contains two
components - internal content (corresponds to the ontological context) and visual representation
(determines the features of the display to the end user) [
The visual representation itself, according to the founder of the theory of frames work, M. Minsky
[
The basic components of the frame are cognitive components that form an idea of typical events. In this case, any frame consists of filled slots (terminals), the similarity between which is necessary to establish links between the information contained in them. Since there can be many slots in a single frame, their diversity must be hierarchically organized. As an example in Fig.2. shows the taxonomy of the frame for the Network router concept.
The use of this data structure involves a high systematization of knowledge by increasing the number of structural relationships between vertical and horizontal elements.
The SKOS system is a model of knowledge organization developed by a consortium of the World
Wide Web, the main purpose of which is to use in the Semantic Web to facilitate the interaction of
information resources through standardization of thesauri, classifications, taxonomies [
To compare dictionaries, SKOS has developed semantic relationships that express different degrees of closeness of concepts: skos: exactMatch, skos: narrowMatch, skos: broadMatch, skos: closeMatch, etc. 4.2.
As it was shown, the subject area ontology visualization is a difficult stage in the design process, given that its display is proposed to be carried out in fragments. Namely, first to visualize a part of the ontological mapping (fragment) with the subsequent connection and use of the ontology of the categorical level. In this case, the process of creating an ontology of the subject area is proposed to perform as a sequence of the following stages: • formation of a list of concepts that are part of the ontological structure of the subject area; • ordering the list of concepts; • creation of a glossary; • display of arcs between created concepts.
Usually, in the process of the first stage, it is proposed to use information from explanatory dictionaries that reflect aspects of the functioning of the subject area. After that, the formed list is specified by the knowledge engineer, on the basis of professional (expert) data in order to determine the final list of concepts that will be used in the process of the subject area ontology visualization. In this case, firstly it is necessary to determine the technical aspects, which include the features of the reflection of concepts such as part of speech, plurality, etc.
In the second stage, the knowledge engineer, based on his own experience, carries out the process
of organizing the list of concepts depending on their «subordination». The definition of «subordination»
can be realized by one of the approaches: «top-down», «bottom-up» or combined [
Creating a glossary involves the formation of a list of definitions for terms that form the concept of the research subject area. Usually, its formation begins with the definition of basic terms and is clarified and supplemented at this stage. The peculiarity of this stage is the creation of formalized descriptions that contain essential features of certain concepts that affect the accuracy of the final ontology.
Then, on the basis of constructed and reflected classes-concepts, the visualization of the conceptual model of the subject area is carried out in the form of an ontograph which is offered to be visualized in the created software.
As an example, consider the process of forming a fragment of the ontology of the subject area “information security”, which reflects the process of describing the characteristics of a network computer attack. The analysis of the subject area allowed to choose the following concepts: “Illegal Activity”, “Social Engineering”, “Computer Network Attack”, “Local Penetration”, “Pretexting”, “Phishing”, “Protocol”, “Quiproquo”, “Tailgating”, “Travel Apple”, “SMTP”, “FTP”, “HTTP”, “HTTPS”, “HTTP Message”, “HTTP Request”, “HTTP Response”, “Request Line”, “Request Header”, “Status”, “Header”, “Content”, “Method”, “Trace”, “Delete”, “Referrer”, “Put”, “Get”, “Post”, “Version”, “URL”, “Port No”, “Domain Name”, ”Query String”, “Protocol Name”. Let's organize the list of terms, having formed the corresponding tuples: • Illegal Activity = < Social Engineering, Computer Network Attack, Local Penetration> • Social Engineering = <Pretexting, Phishing> • Local Penetration =<Quiproquo, Tailgating, Travel Apple> • Computer Network Attack=<Protocol> • Protocol=<SMTP, FTP, HTTP, HTTPS> • HTTP=<HTTP Message> • HTTP Message=<HTTP Request, HTTP Response> • HTTP Response=<Status, Header, Content> • HTTP Request=<Request Line, Request Header> • Request Line=<Method, Version, URL> • Request Header=< Referrer > • Method = <Trace, Delete, Put, Get, Post> • URL = <Port No, Domain Name, Query String, Protocol Name>
On the basis of the ordered list of tuples it is necessary to form a matrix of contiguity on exits and
on the basis of the developed approach to carry out division of the formed concepts on tiers with use of
a tier-parallel form of display [
Further research was aimed at automating this process, given that a decision was made on the software implementation of the described approach. The analysis of the existing visualization systems allowed to form the requirements to the designed system, which will provide opportunities for the correct display of ontologies in accordance with the proposed criteria. The requirements are described using a mechanism of functions and delimitation of work for users and administrators. After defining the requirements, the system was designed using an object-oriented approach. A set of UML diagrams is constructed, in particular the diagram of use cases is presented in Fig.4.
The main functions of users are: • downloading matrix descriptions and their visualization; • search by visualized ontology; • adjustment of visualization parameters, in particular display: distances between visualized classes, sizes of elements used for display; • obtaining information about ontology metadata (title, author, etc.) and statistical data (number of nodes, visualization arcs, etc.).
Users-administrators will be able to: • access the software code, which will allow you to quickly eliminate errors in the work or temporarily block certain functionality; • delete files with visualized ontologies; • implement feedback from users. 4.4.
A studied subject area feature is the presence of different types of calculations and operations with
data that need to be visualized, which requires simultaneous operation with the multiple users system.
Given that, it was decided to implement it in the form of a multifunctional website, which on the one
hand will provide unimpeded access to the resource, and on the other - will facilitate its use on different
platforms [
The design of the software architecture was carried out using an architectural style, which is based on defining the basic elements of the system structure - components of subsystems, objects and relationships between them using three-tiered client-server architecture. This type of software architecture peculiarity is that it consists of three main parts: • the client component of the program, which usually implements the graphical user interface and transmits requests to the application server using the types of data that it can process; • a program execution server that implements business logic and sends requests to a “third-tier” server, such as an external API or a remote database server, to retrieve additional data from it; • database server - manages the database and is responsible for the integrity and storage of data, as well as provides input-output operations when the client accesses information.
When developing a visualization system, the use of a three-tier client-server will allow you to use a number of its advantages, in particular [37]: • reducing the load on the hosting server network, as the links will not exchange significant amounts of information; • the possibility of independent implementation of change architecture at each level; • ensuring the scalability of the system and the possibility of easy updating.
Comparing the structure of modern visualization systems [38-40], in particular the main components and purpose of their subsystems, with a three-tier client-server architecture, we can see that they have much in common (purpose, relationships, etc.), which means that its use will be optimal for both program developers and users.
The next stage of the study was the choice of implementing the software system means. The analysis of both the required functionality and the interactivity of the application contributed to the choice as a programming language - JavaScript, which uses the user's web browser as the execution environment. The main technologies that were used:
1. JQuery - a library, the use of which allowed to create a dynamic web page that supports animation, event handling and the use of AJAX-applications;
2. HTML5 (HyperText Markup Language, version 5) - this language has improved (compared to previous versions) multimedia support, provides adaptability of the Internet resource while maintaining backward compatibility and ease of analysis for software spiders;
3. CSS 3 (Cascading Style Sheets, version 3) - the use of the third version of cascading tables provided the means to quickly describe the appearance of a web page created using HTML; 4. Bootstrap 3 - the use of styles in this library has created a product that is adaptable to the screen size and is supported by all modern web browsers.
The Webpack library was used to build individual JavaScript modules, which combines all the software modules of the system and creates static resources that represent them. In Fig. 6 presents a diagram of the Webpack.
In addition to Webpack, we used the Grunt task manager, written in JavaScript and designed to automate a wide range of tasks such as source code compilation, binary code collection, and system deployment. The combination of «grunt-webpack» and «webpack-dev-server» allowed implementing live-server technology, which provides automatic start of the web server after each code update. The main module of the developed system is a visualization unit, based on the D3.js library. The name D3 stands for Data-Driven Documents, which focuses on the use of JavaScript, SVG and CSS, which provide flexible tools for creating ontological structures with animation elements and interoperability. Compared to pixel graphics, SVG has many advantages, namely XML-based, which makes the generated code easy to read and helps reduce the size of the generated images. These features have led to the fact that today, this framework is the most popular for graphical data processing and creating various kinds of graphical interpretations. With the help of this framework, a frame model is built in the created prototype, which adapts to the user's requirements. 4.5.
The described architecture formed the software implementation basis of the subject area ontology visualization system prototype, which was implemented in JavaScript. Fig.7 shows the main page of the website which presents a visualized ontograph of the example that was shown in paragraph 4.2.
As can be seen from the figure, the main window of the system consists of two parts: the left contains a hierarchical tree that displays the classes of the ontograph and the right - the main window in which the visualized ontograph is displayed. The formed ontograph is presented in a tier-parallel form in which there are no relationships within the tier. This form of presentation fully meets the established criteria of reproduction and is easy to evaluate and further study. A separate function of the system is to search for the ontograph node by key term, and the search result is displayed in an enlarged form with «toning». An example of a search for «HTTP-Request» is shown in Fig.8:
Currently, the system operates in the form of a demonstration prototype with limited functionality, which will be expanded in accordance with the described functionality.
As a result of the conducted research the existing technologies and systems of ontologies display on the given features which motivate the creation of a new web-oriented system of visualization are analyzed. The analysis of possible means of ontology representation in the form of frame models and use of SKOS is carried out. The main stages of building an ontology of the subject area are described and an example of visualization of the concept of «network computer attack» is given. Further research was based on the creation of a project using an object-oriented approach and the use of UML language tools. The practical implementation of the system is based on the use of three-tiered client-server architecture using the Javascript language. The result of the work is a prototype of an ontology visualization system that can be used to display a given structure and describes a specific subject area.
Further research will be aimed at creating related software modules, their verification and coordination of operation.
[36] T. Basyuk, The Popularization Problem of Websites and Analysis of Competitors”. in: Shakhovska N., Stepashko V. (eds) – Advances in Intelligent Systems and Computing II. CSIT 2017. Advances in Intelligent Systems and Computing, vol 689. Springer, Cham pp. 54-65, doi:10.1007/978-3319-70581-1_4. [37] T. Basyuk, A. Vasyliuk, V. Lytvyn, Mathematical Model of Semantic Search and Search Optimization, Proceedings of the 3rd International Conference on Computational Linguistics and Intelligent Systems (COLINS-2019). Volume I: Main Conference, Kharkiv, Ukraine, April 18-19, 2019, Vol-2362: pp.96-105. [38] S. Narayanan, S. Mcllraith, Analysis and simulation of Web services, Computer Networks: The International Journal of Computer and Telecommunications Networking. Volume 42 , Issue 15 (August 2013). Special issue: The Semantic Web: an evolution for a revolution. - New York, NY, USA Publisher Elsevier North-Holland, Inc., 2013. - pp. 675-693. [39] Z. Cui, M. Cox, D.Jones, An Environment for Managing Enterprise Domain Ontology, 2000. URL.
http :/www.btexact.com/-proj ect/-ibsr/publications.html. [40] N. Casellas, Ontology Engineering: Methodologies, Modelling Trends, and the Ontology of Professional Judicial Knowledge, Springer Science & Business Media, 2011.