Comparisons of typical algorithms in negative sequential pattern mining
Authors:
- 1 May 2014
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 387-390
Abstract
Negative sequential pattern (NSP), which contains both non-occurring and occurring items, can play more important roles than positive sequential pattern (PSP) in many applications. NSP mining, however, has been just caught attention and very limited methods are available to mine NSP. Furthermore, there is not a unified definition about negative containment, i.e., how a data sequence contains a negative sequence. So in this paper, we select three typical existing methods, PNSP, Neg-GSP and e-NSP, and compare their definitions, runtime and the number of NSPs. Examples and experiments on same dataset clearly show their differences.Keywords
This publication has 8 references indexed in Scilit:
- e-NSP2011
- Mining Positive and Negative Sequential Patterns with Multiple Minimum Supports in Large Transaction Databases2010
- An Efficient GA-Based Algorithm for Mining Negative Sequential Patterns2010
- Mining Positive and Negative Fuzzy Multiple Level Sequential Patterns in Large Transaction Databases2009
- Mining Negative Sequential Patterns for E-commerce Recommendations2008
- Mining Positive and Negative Fuzzy Sequential Patterns in Large Transaction Databases2008
- Mining Negative Sequential Patterns in Transaction Databases2007
- Mining sequential patterns2002