Exploratory Semantic Video Search with yovisto

In recent years, especially audiovisual media have become the predominant media of the internet. To enable content based video retrieval, high quality textual metadata have to be provided that describe the content. Keyword-based search in general is particularly applicable if the searcher really knows what she is looking for and how to find it, i.e. to know the appropriate keywords to obtain the desired results But in many cases either the objectives of the searcher are intrinsically fuzzy or she has no idea of the appropriate keywords. Complex search tasks often cannot be answered by retrieving a single document, but require subsequent searches that build on one another. Moreover, if the user tries to achieve an overview of actually available information about a certain topic, today´s web search engines are flooding search results by millions. Thus, giving the user no chance to review but the first few resulting pages. Traditional keyword-based search does not consider the meaning (semantics) of the content of the underlying information and result ranking is mainly based on link popularity.