This thesis presents novel techniques to index multiple media streams in a digi- tally captured presentation. These media streams are related by the common content in a presentation. We use relevance curves to represent these relationships. These relevance curves are generated by using a mix of text processing techniques and distance measures for sparse vocabularies. These techniques are used to automatically detect slide boundaries in a presentation. Accuracy of detecting these boundaries is evaluated as a function of word error rates.

This masters thesis is dedicated to theme of multimodal databases, especially multimedia databases. The first part contains overview of today's most used data models. Next part summarizes information about content-based search in multimedia content and indexing this type of data. Final part is dedicated to implementation system for storing and managing of multimedia content based on Helix streaming system a PostgreSQL database system.