Running projects are listed below, see “Completed projects” for completed projects.

Automated support of manual testing of web software systems based on formal model of the application

Researcher: Ing. Karel Frajták

The aim of the project is development of automated support of manual testing of web applications. This support is aimed for situations, when testers don’t have prepared manual test scenarios. The approach we take is direct guidance of tester based on formal model, generated from standard design model of web application. The formal model describes the tested application and contains also testing coverage and special metadata, allowing to increase the reliability and efficiency of performed tests. Based on this formal model, methods and algorithms for tester navigation will be formulated. The formal model and the whole concept will be verified by pilot implementation of navigation application for testers and by application of the solution on several selected real projects of web applications development.

A General Model for Personalized Web Design

Researcher: Ing. Martin Balík

The project aims to devise a model for personalized web design. This model needs to be suitable for universal use in applications increasing efficiency of user access to information on the web.

Výzkum inovativních testovacích metod a prostředků pro zvýšení efektivity testování

Researcher: Ing. Miroslav Bureš, Ph.D.

Se zvyšující se komplexitou současných webových aplikací a zvyšujícím se množstvím dat, které jsou těmito aplikacemi zpracovávány je kladen i větší důraz na spolehlivost a obecnou použitelnost těchto aplikací Tím rostou i požadavky na disciplíny testování softwaru a quality assurance tuto spolehlivost zajistit. Současný běžný způsob testování software přestává být v mnoha aspektech vyhovující a v rámci disciplíny je zapotřebí výzkumu a vývoje nových metod a efektivnějších postupů testování. V rámci výzkumu se zabýváme tématy jako automatizace revize analytické a návrhové dokumentace, zaměření testů na pokrytí skutečných požadavků zadavatele na vyvíjený software, automatizované ověřování kvality a relevance připravených testovacích scénářů, testy generované z modelu aplikace přizpůsobené reálným podmínkám stávajících projektů a související.


Researcher: Ing. Radek Malinský, Ph.D.

Web 2.0 (social networking, blogging, etc.) can serve as a data source for social science research because it contains vast amount of information from many different users. The volume of information on the web has been growing at a very high rate and becoming a network of heterogeneous data, this makes things difficult to find and is therefore not almost useful. It is necessary to design suitable metric for such volume of information, which would better separate
the useful data from the useless.

This project aims to make a theoretical definition of a novel web metric, which improved the content representation of a web page, its overall semantic meaning and its relations with similar web pages. The metric based on combination of two different research areas: Webometrics and Sentiment Analysis. The project further deals with a proposal of an appropriate citation index structure and algorithm for the information obtained from it. The index stores an output of the content analysis and it allows quick and easy searches with regard to users’ queries. The main goal is to develop a complex system, which improves the quality of search engines on the principle of webometrics. That will lead to a better machine understanding of website content and user specified queries, and thereby the reduction of irrelevant web search results.

Original webometric techniques improved searching and provided trend detection. However, they are not able to distinguish a polarity of the text, its semantic meaning and the emotional state of the author. Webometrics is purely a quantitative approach to the web, which can be enhanced by qualitative methods and thereby allows us to expand the possibilities of a study problem. Sentiment analysis may be used as a qualitative complement to quantitative approach. Extension of the webometric techniques of sentiment analysis methods leads up to gaining insights into a public opinion with respect to some topic and to a better machine understanding of the text. Better understanding of the content on the web site could have a significant impact on the quality of site evaluation.

Semantic Web Access Prediction by Means of WordNet

Researcher: Ing. Lenka Hapalová

Web access prediction as the part of web page prefetching belongs among the most important techniques to reduce web access latency observed by users of the Internet. Prefetching hyperlinks presented on current page is one of the most common. It have an advantage of their simplicity and easy of use. In comparison with the history based methods prefetching hyperlinks can make predictions of actions that have never been taken by the user, but are surpassed by history based models in accuracy. The principal aim of the project is to suggest, implement and test semantic web access predistion algorithm that will use semantic techniques. To be able to use the semantic techniques the semantics of the web pages will be represented by ontology that provides us not only terms and their definitions but also the relationship between the objects.

Semantic P2P Search Engine

Researcher: Ing. Ilya Rudomilov

The goal of this project is to design and implement intelligent Semantic Peer-to-Peer (P2P) search engine as multi-agent system (MAS) by using open-source framework JADE (Java Agent Development Framework). JADE framework implements FIPA standards (The Foundation for Intelligent Physical Agents) and modern techniques for developing multi-agent systems. The main idea of the proposed project is to use semantic model for P2P search engine, where peers are specified as semantic meta-models by the standardized OWL language from The World Wide Web Consortium. Using semantic model improve the quality of communication between intelligent peers in this P2P network. Undoubtedly proposed semantic P2P network has all advantages of normal P2P networks and in the first place allow deciding a point with bottle-neck effect (typical problem for client-server applications) by using a set of peers for storing and data processing.

Methods and Technologies for Testing of Web Software Systems

Researcher: Ing. Martin Filipský

Main goals of our research are to propose and validate a formal layer among requirements on automated testing, an application under and test scripts for the automated testing. On the proposed layer, algorithms and approaches will be defined for a solving of problems with a test script aging, with an uncertain test automation of web applications and finally with a heterogeneity of input requirements and platforms for automated testing.