Similarity search in uncertain databases
Matthias Renz · 2011
Querying and mining uncertain data has received a lot of attention from the research community in recent years. The reasons are advances in collecting data, whereas collection methodologies are imprecise and yield incomplete or inaccurate information. Traditional query processing approaches are often inapplicable or may extract misleading or plain wrong information when applied to uncertain data. Therefore, modern data management solutions coping with uncertain data are very important. The incorporation of the uncertainty enables us to increase the quality of query results, but yields new problems such that novel query processing methods are required. In this talk I will address this research field from the efficiency point of view. I will give an overview of modeling uncertain data in feature spaces and illustrate diverse probabilistic similarity search methods which are important tools for many modern similarity search applications.