Образовательный портал «Кафедра онлайн» НИЯУ «МИФИ», Обнинск"

Главная » Рубрики » Блоги преподавателей » Свежая публикация по семантическому вебу

Свежая публикация по семантическому вебу

В.П. Тельнов. Национальный исследовательский ядерный университет МИФИ, ИАТЭ, Россия, Обнинск.

Telnov V, Korovin Y (2019) Semantic web and knowledge graphs as an educational technology of personnel training for nuclear power engineering. Nuclear Energy and Technology 5(3): 273-280. http://doi.org/10.3897/nucet.5.39226

Аннотация. The technologies of knowledge representation and inference in an artificial intelligence system focused on the domain of nuclear physics and nuclear power engineering are considered. The possibilities of description logics and graph databases of nuclear knowledge for the generation of cognitive hypotheses, using in addition to deduction and other ways of reasoning, such as inductive inference and reasoning based on analogies, are discussed. The use of adequate description logic and measures of semantic similarity is substantiated. Interactive visual navigation and reasoning on the knowledge graphs are performed by means of special retrieval widgets and the smart RDF browser. Operations with semantic repositories are implemented on cloud platforms using SPARQL queries and RESTful services. The proposed software solutions are based on cloud computing using DBaaS and PaaS service models to ensure scalability of data warehouses and network services. Example of use of the offered technologies and software has been given.

Ключевые слова: Nuclear education, semantic web, knowledge graph, cloud computing

Введение

Since the 1960s, in the framework of research on artificial intelligence, various formalisms for knowledge representation (semantic networks, frame systems, etc.) have been developed (Harmelen et al. 2008). In 2019, the ontology description languages RDF, OWL (W3C 2012), knowledge graphs and description logics (Baader et al. 2010) provide a modern theoretical basis for the creation of systems and methods of acquisition, presentation, processing and in-tegration of problem–oriented knowledge in computer systems, which, in particular, is con-firmed by the current standards W3C in the field of semantic web.

The reports of International Conference on Semantic Systems, International Workshops on Description Logic noted the growing interest of giants of the IT industry (Google, Facebook, Wikimedia) to graph models of knowledge representation and description logics. As of 2019 educational web–portals of universities, national centers for the exchange of scientific infor-mation, world nuclear data centers underused semantic web technologies. As for the inductive inference rules in graphs, the following considerations make them useful. First, inductive in-ference rules based on consideration of possible alternatives (precedents) allow to generate cognitive hypotheses (fuzzy knowledge) that cannot be obtained directly by deductive reason-ing on the graph. Secondly, inductive inference is one of the basic technologies of semantic annotation of network content, when it is necessary to redesign, expand and update existing graphs with new knowledge. With the help of inductive inference the problems of classifica-tion and clustering of new concepts and individuals in the semantic base of nuclear knowl-edge are solved.

The aim of the work presented in the paper is to create a semantic web portal of knowledge in the domain of nuclear physics and nuclear power engineering based on ontology and using graph databases deployed on cloud platforms. The task of the study was to create the follow-ing graphs of nuclear knowledge:

  • World nuclear data centers;
  • Nuclear research centers;
  • Events and publications from CERN;
  • IAEA databases and network services;
  • Nuclear physics at MSU and MEPhI;
  • Nuclear physics journals;
  • Joint nuclear knowledge graph.

The potential beneficiaries of information solutions and technologies that are proposed in the paper are students, teachers, experts, engineers, managers and specialists in the domain of nuclear physics and nuclear power engineering (target audience).

Щелкните здесь мышкой, чтобы увидеть полный текст публикации.

Figure 1. Search widgets designed for a quick immersion in knowledge graphs.

Figure 2. Fragment of the knowledge graph titled «Nuclear Physics at MSU, MEPHI» as an example of the implementation of inductive reasoning on graph.

Комментариев нет

Похожие публикации
 
 

Комментариев нет

Вы можете быть первым, кто оставит комментарий.

 
 

Вы можете оставить комментарий

 





 
 

Выполните простое задание (антиспам). Картинки можно сибирать приблизительно, без точной подгонки фрагментов.


 
 
 

Наверх