A compositional framework for mining longest ranges (2002)
Download:
by Haiyan Zhao, Haiyan Zhao, Zhenjiang Hu, Zhenjiang Hu, Masato Takeichi, Masato Takeichi
http://www.keisu.t.u-tokyo.ac.jp/Research/METR/2002/METR02-05.pdf
Add To MetaCart
Abstract:
This paper proposes a compositional framework for mining interesting range information from huge databases, in which a domain specific query language is provided to specify the range of interest, and a general algorithm is given to mine the range specified in this language quickly. A wide class of longest range problems, including the intensively studied optimized support range problem [FMMT96], can be solved efficiently and systematically in this framework. Experiments with real world databases have been done and the results show that our framework is efficient not only in theory but also in practice.
