Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Adaptive Knowledge Portal for Education Domain

59 views

Published on

Abstract: https://arxiv.org/submit/2064092/preview

Published in: Data & Analytics
  • Be the first to comment

  • Be the first to like this

Adaptive Knowledge Portal for Education Domain

  1. 1. Adaptive Knowledge Portal for Education Domain Mikhail Navrotskiy ITMO University, Informatics and Applied Mathematics Department m.navrotskiy@gmail.com Abstract The Scientific Web is a context-dependent projection of scientific knowledge on the users interests, department and university goals. Scientific knowledge is a collection of integrated data from the University databases, the University open data portal and open data portals of articles, courses and other types of educational content. It also contains such external databases as DBPedia and WikiData related to the Universities activities. For the scien- tific data representation and management there is developed the ontology. It is based on FOAF, VIVO, and BIBO ontologies. Introduction In the last ten years there has increased the amount of published online content which includes sci- entific and educational data. These data main sources include universities, scientific publishers, aca- demics and researchers[2]. They share information about publications (articles), research projects, open educational courses, scientific results, and so on. These sources include Linked Open Data (LOD) Sources and non-LOD sources[1]. These data allow to solve a lot of such problems as: calculation and evaluation of performance indicators (employees and departments); development of grant applications, conference and events, education. Since a large number of described data have already been published in the clear form, it is proposed to integrate data from these sources to eliminate the described problems. Thus, employees can use various sources to obtain the necessary information[3]. On the other hand, the sources may be simple catalogues and large and complex portals with linked data, then integration could create too much knowledge. Main Objectives 1. Using LOD in the University (Department) staff and students for their daily tasks. 2. Develop ontology model for representation data from LOD sources. 3. Develop knowledge portal for eduaction domain. Ontology Model When developing the ontological model, the following open ontologies were used: FOAF, DBO (DB- pedia ontology), BIBO, Teach, OU (Open University). Classes developed within the framework of the work have the prefix ”ifmo”. The developed ontology (the data structure of the term description) is shown in the figure 1. Figure 1: Ontology structure Ontologies includes the following classes: Publication - a scientific publication; University - an university; Department - university department; Project - a scientific or research project; Person - an employee or student; Student - a student in the current university; ResearchArea - the research area; DataSource - LOD or non-LOD data source; Resource - an educational resource; Article - an article in scientific journal; and others. The System Architecture The application includes the following modules (fig. 2): Client - lightweight client (HTML5, iOS, Android), used for user interaction with the system; Glossary server - a dictionary that describes domain; Portal server - the main portal server; Configuration module - mod- ule for system configuration; Local ontology - local ontology in OWL format; Adaptation module - the adaptation module of the scientific data; Sources base - base of data sources; Department’s semantic database - semantic portal of department; Search module - module to build queries to sources; Merge module - the merge module data from sources; External data source - LOD sources. Figure 2: The application architecture Case study Consider an example in which a student learns the ontologies course material. At the same time he is interested in the definition of the term ”Semantic Web” which he is not familiar with. The user goes to the portal and begins to search for the term definition. This example is performed for the Informatics and Applied Mathematics Department of the ITMO University. The rest of the queries can be viewed in more detail in the project repository1. The result is shown in the figure 3. Figure 3: Page with data about “Semantic Web“ (Fragment) Using all data sources, the user can obtain the following data about the domain: • Definition of the term in several languages (now Russian and English are used). • References to articles about this and related terms in Wikipedia. • Famous scientists working list in this domain. • Publications on this domain. • Projects in this domain. Conclusions The applications were deployed (to cloud) and students, postgraduate students and employee of ITMO University Computer Science and Applied Mathematics Department who are using wiki application and lodGlossary in their educational processes. It is a closed beta testing process. The authors want to deploy these apps for all the university departments in the future. Forthcoming Research Some problems which raised in the project require additional research and further development. The most challenging problems are: 1. To add more data sources (for instance, springer, other universities from Linked University organi- zation and others). 2. To develop algorithms and methods for helping the specialists to choose the data sources for each department and query (we can use SWRL rules, for example). 3. We need to optimize queries to external data sources. Some terms require a very long time for searching data and rendering a page. References [1] Tayfun G¨okmen Halac¸, Bahtiyar Erden, Emrah Inan, Damla Oguz, Pinar Gocebe, and Oguz Dikenelli. Publishing and linking university data considering the dynamism of datasources. In Proceedings of the 9th International Conference on Semantic Systems, pages 140–145. ACM, 2013. [2] Nikolay Karmanovskiy, Dmitry Mouromtsev, Mikhail Navrotskiy, Dmitry Pavlov, and Irina Rad- chenko. A case study of open science concept: Linked open data in university. In Digital Trans- formation and Global Society, pages 400–403. Springer, 2016. [3] D.I. Mouromtsev, J. Lehmann, I.A. Semerhanov, M.A. Navrotskiy, and I.S. Ermilov. Study of current approaches for web publishing of open scientific data. Scientific and Technical Journal of Informatics Technologies, Mechanics and Optics, 15(6):1081–1087, 2015. 1 https://github.com/LODIFMO/wiki-app

×