Laboratory for Data Technologies

The Laboratory for Data Technologies was founded in 2009 and is the youngest laboratory at the Faculty for Computer and Information Science. Its members are involved in basic and applicative research in the fields of data analysis, data representation, data visualization and semantic web.


Collaborators


Research Activities

Data analysis

Data analysis is a very broad research area. We focus on the segments of business rules management systems, business intelligence, fraud management and (social) networks. Research is divided between academic research and applicative research. Applicative research is closely connected to the fraud management field and transaction intensive information systems architectures.

Data representation

Data entities are typically represented in an ordinary flat form. However such representation is not adequate when we are interested in relations between entities or in patterns in these relations. In that case one must employ some enhanced representation of data like networks. Networks are the most natural representation of any relational domain (hyper pages, social networks etc.) allowing formulation of complex relations between entities. They allow analysis of entities in context of related entities and not in complete isolation. Networks are currently one of the hot topics in many research areas (network analysis, data mining, bioinformatics, etc.). Our research is mainly focused on analysis and mining of networks data and using networks for fraud detection in different fields.

Data visualization

As the volume and complexity of data increases it becomes very difficult for users to effectively explore large-scale datasets. A possible solution for this problem is visualization (graphical representation of data). Visualizing large amounts of data allows us to see patterns that may otherwise remain hidden and it allows us also to quickly grasp and process large amounts of data that would otherwise require a lot of time to study. Visualizations are used in many fields (medicine, education, geovisualizations, data-mining, financial data analysis etc.) and employ different visualization techniques (graphs, cluster diagrams, volume rendering etc.), but just any arbitrary visualization might not be useful and may even lead to flawed conclusions. An important aspect of visualization is also dynamics of representation and interactivity (e.g. dynamical adjustment of mapping in real-time).

Semantic web

Current version of World Wide Web is consisted of several mutually connected documents that are presented to human users by computers. These documents originated in interconnected systems where every user could contribute. This also results in a fact that information quality cannot always be guaranteed. Current World Wide Web consists of data, information and knowledge, but the role of computers at this stage is only to deliver and represent the content of the documents that describe knowledge. To integrate different information resources users have to manually interpret these data. Semantic Web tends to improve current World Wide Web with computers processing, interpreting, integrating data on the web and with this approach aiding human users in discovering complex knowledge. Semantic Web is focused towards sharing and reusing of data and not documents. The research area emphasizes establishment of common framework to enable sharing and reusing data among applications and enterprises.


Projects

Current Projects

Closed Projects