New Technologies for the Information Society, Faculty of Applied Sciences, University of West Bohemia Plzeň, Czech Republic ; Department of Computer Science and Engineering, Faculty of Applied Sciences, University of West Bohemia Plzeň, Czech Republic.
Front Neuroinform. 2015 Feb 25;9:3. doi: 10.3389/fninf.2015.00003. eCollection 2015.
The article deals with and discusses two main approaches in building semantic structures for electrophysiological metadata. It is the use of conventional data structures, repositories, and programming languages on one hand and the use of formal representations of ontologies, known from knowledge representation, such as description logics or semantic web languages on the other hand. Although knowledge engineering offers languages supporting richer semantic means of expression and technological advanced approaches, conventional data structures and repositories are still popular among developers, administrators and users because of their simplicity, overall intelligibility, and lower demands on technical equipment. The choice of conventional data resources and repositories, however, raises the question of how and where to add semantics that cannot be naturally expressed using them. As one of the possible solutions, this semantics can be added into the structures of the programming language that accesses and processes the underlying data. To support this idea we introduced a software prototype that enables its users to add semantically richer expressions into a Java object-oriented code. This approach does not burden users with additional demands on programming environment since reflective Java annotations were used as an entry for these expressions. Moreover, additional semantics need not to be written by the programmer directly to the code, but it can be collected from non-programmers using a graphic user interface. The mapping that allows the transformation of the semantically enriched Java code into the Semantic Web language OWL was proposed and implemented in a library named the Semantic Framework. This approach was validated by the integration of the Semantic Framework in the EEG/ERP Portal and by the subsequent registration of the EEG/ERP Portal in the Neuroscience Information Framework.
本文探讨了构建电生理元数据语义结构的两种主要方法。一方面是使用传统的数据结构、存储库和编程语言，另一方面是使用来自知识表示的本体的正式表示形式，例如描述逻辑或语义网语言。尽管知识工程提供了支持更丰富语义表达的语言和先进的技术方法，但由于其简单性、总体可理解性和对技术设备的低要求，传统的数据结构和存储库仍然受到开发人员、管理员和用户的欢迎。然而，选择传统的数据资源和存储库引发了一个问题，即如何以及在何处添加无法使用它们自然表达的语义。作为一种可能的解决方案，可以将这些语义添加到访问和处理基础数据的编程语言的结构中。为了支持这个想法，我们引入了一个软件原型，使用户能够将语义更丰富的表达式添加到面向 Java 对象的代码中。这种方法不会给用户带来对编程环境的额外要求，因为使用了反射性的 Java 注释作为这些表达式的入口。此外，不需要程序员直接将额外的语义写入代码，而是可以使用图形用户界面从非程序员那里收集。还提出并实现了一个名为 Semantic Framework 的库，用于允许将语义丰富的 Java 代码转换为语义网语言 OWL 的映射。通过将 Semantic Framework 集成到 EEG/ERP 门户中，并随后将 EEG/ERP 门户注册到 Neuroscience Information Framework 中，对该方法进行了验证。
