DocumentCode
2369364
Title
An algebra for inductive query evaluation
Author
Lee, S.D. ; De Raedt, Luc
Author_Institution
Inst. fur Inf., Freiburg Univ., Germany
fYear
2003
fDate
19-22 Nov. 2003
Firstpage
147
Lastpage
154
Abstract
Inductive queries are queries that generate pattern sets. We study properties of Boolean inductive queries, i.e. queries that are Boolean expressions over monotonic and antimonotonic constraints. More specifically, we introduce and study algebraic operations on the answer sets of such queries and show how these can be used for constructing and optimizing query plans. Special attention is devoted to the dimension of the queries, i.e. the minimum number of version spaces needed to represent the answer sets. The framework has been implemented for the pattern domain of strings and experimentally validated.
Keywords
Boolean algebra; constraint theory; data mining; query processing; Boolean expression; algebraic operation; antimonotonic constraint; inductive query evaluation; monotonic constraint; query plan optimization; Algebra; Data mining; Database systems; Query processing; Relational databases;
fLanguage
English
Publisher
ieee
Conference_Titel
Data Mining, 2003. ICDM 2003. Third IEEE International Conference on
Print_ISBN
0-7695-1978-4
Type
conf
DOI
10.1109/ICDM.2003.1250914
Filename
1250914
Link To Document