Lukas Radvilavičius

Clustering visually similar web page elements for structured web data extraction

We propose a novel approach for extraction of structured web data called ClustVX. It clusters visually similar web page elements by exploiting their visual formatting and structural features. Clusters are then used to derive extraction rules. The experimental evaluation results of ClustVX system on three publicly available benchmark data sets outperform state-of-the-art structured data extraction systems.

Visualization of Mapped Security Standards for Analysis and Use Optimisation

—There are many security standards which the company can use. While security standards can differ in purpose and covered area, more than one standard can be used at the same time which leads to overlap and potential conflicts in requirements of different standards. In such cases, deep analysis of used standards has to be done to ensure optimal usage of companys resources while implementing these security requirements. In this paper we analyze existing solutions for standard harmonization, describe concepts of adaptive mapping and present new ideas for visualization of mapped security standards. These ideas were integrated into developed tool to visualize mapped security standards.

Raising effectiveness of access control systems by applying multi-criteria decision analysis: part 1 – problem definition

AbstractCurrently, control of access to information and physical resources has become extremely important. Numerous methods and solutions for architecture of systems aimed at controlling physical access are available; however, there is little information about application of Multi-Criteria Decision Analysis methods when evaluating separate logical components, needed for the design of access control systems and their interconnection in the final architecture.This paper is the first part of a two-part article, discussing application of multi-criteria decision making for architecture of access control systems. The first part defines the problem and discusses the possibility to use Multi Criteria Decision Making techniques when designing access control systems, including risk analysis for specific criteria and practical application of the developed model. In the second part, the possible solution model will be presented.

Database Level Honeytoken Modules for Active DBMS Protection

Data and information are sometimes called the “crown jewels” of an organization [3]. Modern company cannot exist without flexible, fast and secure data management system. Insuring data security is of vital importance for any company.