TY - GEN
T1 - A one-phase method for mining high utility mobile sequential patterns in mobile commerce environments
AU - Shie, Bai En
AU - Cheng, Ji Hong
AU - Chuang, Kun-Ta
AU - Tseng, Vincent S.
PY - 2012/8/1
Y1 - 2012/8/1
N2 - Mobile sequential pattern mining is an emerging topic in data mining fields with wide applications, such as planning mobile commerce environments and managing online shopping websites. However, an important factor, i.e., actual utilities (i.e., profit here) of items, is not considered and thus some valuable patterns cannot be found. Therefore, previous researches [8, 9] addressed the problem of mining high utility mobile sequential patterns (abbreviated as UMSPs). Nevertheless the tree-based algorithms may not perform efficiently since mobile transaction sequences are often too complex to form compress tree structures. A novel algorithm, namely UM-Span (high Utility Mobile Sequential Pattern mining), is proposed for efficiently mining UMSPs in this work. UM-Span finds UMSPs by a projected database based framework. It does not need additional database scans to find actual UMSPs, which is the bottleneck of utility mining. Experimental results show that UM-Span outperforms the state-of-the-art UMSP mining algorithms under various conditions.
AB - Mobile sequential pattern mining is an emerging topic in data mining fields with wide applications, such as planning mobile commerce environments and managing online shopping websites. However, an important factor, i.e., actual utilities (i.e., profit here) of items, is not considered and thus some valuable patterns cannot be found. Therefore, previous researches [8, 9] addressed the problem of mining high utility mobile sequential patterns (abbreviated as UMSPs). Nevertheless the tree-based algorithms may not perform efficiently since mobile transaction sequences are often too complex to form compress tree structures. A novel algorithm, namely UM-Span (high Utility Mobile Sequential Pattern mining), is proposed for efficiently mining UMSPs in this work. UM-Span finds UMSPs by a projected database based framework. It does not need additional database scans to find actual UMSPs, which is the bottleneck of utility mining. Experimental results show that UM-Span outperforms the state-of-the-art UMSP mining algorithms under various conditions.
UR - http://www.scopus.com/inward/record.url?scp=84864339610&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84864339610&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-31087-4_63
DO - 10.1007/978-3-642-31087-4_63
M3 - Conference contribution
AN - SCOPUS:84864339610
SN - 9783642310867
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 616
EP - 626
BT - Advanced Research in Applied Artificial Intelligence - 25th International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2012, Proceedings
T2 - 25th International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2012
Y2 - 9 June 2012 through 12 June 2012
ER -