Abstract
In the last decade there has been much interest in preference query processing for various applications like personalized information or decision making systems. Many attributes from the queried datasets are typically organized in hierarchies of tree-shaped structure. In database systems, this fact is commonly represented by a connection of each tuple to its parent node. SQL dialects provide various constructs for processing this kind of data. We introduce a novel technique to handle hierarchies in preference queries by extension of preference algebra. For this purpose we define new preference constructors to specify wishes for attribute values associated with hierarchies. Our approach allows the intuitive definition of a transitive closure in database systems and overcomes the massive recomputation required by standard methods.
Authors
Patrick Roocks, Florian Wenzel, Olena Rudenko, Markus Endres, Werner Kießling
Publication
9th Multidisciplinary Workshop on Advances in Preference Handling, in conjunction with IJCAI-15, Buenos Aires, Argentina, 2015