Zdeňka Telnarová’s research while affiliated with University of Ostrava and other places

The paper tries to discuss from a slightly higher level a focus oriented towards general properties that a Semantic Web Language (SWL) must have. We state as a ground the following two points of view: to compare (1) Expressivity of the SWL, and (2) a possibility to infer new knowledge from an SWL knowledge base with corresponding properties of the classical First Order Logics (FOPL) as a measure of their basic quality (now prepared). From expressivity, the language for the semantic web must be a common communication tool for computers as well as for people fulfilling an easy-to-use condition by means of adding more semantics directly into the language's syntax. After a discussion of properties and critical recommendations two languages, OWL DL1 and RDF CFL have been proposed here to become the SWL both having the expressivity comparable with the FOPL.

Theoretical development in the area of databases systems has been keeping its dynamics. One of its directions is unfolding the conception of derived data. Models based on derived data are the outcome of the combination of logic programs and databases. They can potentially solve many problems in analysis, design and implementation of databases and information systems and add semantics to the databases. We can use deductive rules for derivation of classes/relations, derivation of attributes, views, for design of query, for expressing integrity constraints.

The paper is devoted to the issue of NoSQL databases and their application in gradually increasing volumes of data in databases. Concerning approaches, three are covered: key-value databases, column-oriented databases, and document databases.

The article aims to process methods and algorithms of computer reading. The authors follow up on the previous work of the authors and describe the synthesis of Text to speech and show the possibilities of using libraries for the future intonation reader with a focus on the Czech language. The result of this study is an understanding of the methodology of reading the Czech text.

How does the human brain work? How can we make sentences and make sentences? These are the questions dealt with already many scientists and several companies developing artificial intelligence. This article presents a study on language networks. At the beginning is the research of the works that have already dealt with this issue. In the next part, the author deals with the application of language networks and semantic networks.

Deductive approaches in database systems are the outcome of the combination of logic programs and databases. They can potentially solve many problems in analysis, design and implementation of databases and information systems. This paper concentrates on deductive rules as a data derivation and integrity maintenance technique. Two database systems and their implementation of deductive rules are mentioned: Chimera and Oracle.

The aim of the article is the analysis and research of readers on the market and readers used at the University of Ostrava. There are many readers on the market, but often with machine translation. The second objective of the paper is an initial proposal intonation reader adapted for reading texts in the Czech language. The Czech language has a special intonation in contrast with the English language. For this reason, the English readers cannot be used to read of Czech texts.

The paper deals with a possibility of using relational databases when storing and managing time series data. It discusses the possibilities of using NoSQL databases as well as the issue of scaling and its implementation into time series. It also mentions advantages of this approach with respect to keeping sophisticated qualities of the SQL language.

This contribution is focused on manipulation with data in two different representations. Data represented by relational tables and data represented by RDF graphs. In terms of manipulation, we consider queries using joins, in relational terminology it means obtaining data from more tables. This contribution discusses the possibility of SQL and SPARQL queries and using mappings of RDF graphs into relational tables with the goal to simulate SPARQL queries in relational databases. There is also an example of Oracle implementation.

The paper is devoted to identification of indeterminacies related with the construction and operation of NoSQL databases and which have a significant impact on the querying possibilities. The problematic issue of NoSQL databases has been selected for its turbulent development of this type of data sources, which are characterised by their diverse architecture, and frequently vague structure of the stored records and, last but not least, a distributed character of this data source. The objective is to identify potential occurrence of indeterminacies and to propose a model which will work with these indeterminacies effectively.