Similarity search refers to the task of finding objects that are similar to a given query in a set of objects. Common DBMS only provide means to efficiently find exact matches to a given query. In case of typing errors, omitted or transposed attribute values or other typical data quality problems in queries, exact search algorithms fail to find all relevant objects in the queried data set.
In this project, we survey existing and develop new algorithms for effective and efficient similarity search. Effective similarity search can be achieved by defining a similarity measure that is well-suited for the given domain. For efficient similarity search, an index structure is required that precomputes similarities of objects to answer queries as fast as possible.
This project is supported by SCHUFA Holding AG.
- Matthias Pohl: Automatisierte Konfiguration des D-Index zur Ähnlichkeitssuche, 2011
- Dandy Fenz: Effiziente Ähnlichkeitssuche in einer großen Menge von Zeichenketten mittels Key-Value-Store, 2011