Extending Dynamic Queries to Handle Uncertain Data

Dynamic querying is a technique which has been used successfully to enable novice users to gain access to and insight into data in databases. Some multimedia archives (such as archives of African art) contain data which have vague locations in time and space, that is, although there is some idea of when and where the entity originated, the precise information is unknown. This uncertainty creates problems with the display and querying of the data and so the data is generally not accessible to novice users. In this study we extend dynamic querying techniques to work with African art data with uncertain origins in time and space. We present methods for storing, visualising and querying such uncertain data within the framework of dynamic querying. Results of user tests indicate that our approach was clear to users and that users could successfully perform simple queries using the visual query tools. A similar approach of extending dynamic querying techniques could apply to other domains with any one-dimensional attribute data with probabilistic uncertainty. In this way we show how it is possible for novice users to query large databases with complex uncertain attributes.