Title :
Weighted sequential pattern mining algorithm research based on well completion business process
Author :
Ruishan Du ; Fuhua Shang
Author_Institution :
Sch. of Comput. & Inf. Technol., Northeast Pet. Univ., Daqing, China
Abstract :
Focusing on the users is more interested in the sequential patterns that accord with the well completion business process habit overall in the web access mode of well completion mobile platform. This paper proposed a weighted sequential pattern mining algorithm based on the well completion business process. With analyzing the business process models and web access log of well completion, confirm the business dependency strength calculation model of well completion as the sequence weight, at the same time, using the technology of k-minimum weighted support in the weighted mining to improve the PrefixSpan algorithm. The algorithm discards a lot of access sequence that dissatisfied the needs of the business process in the weighted mining, avoids the happening of the candidate combination explosion effectively. Experiments showed that the algorithm can rapidly excavate meaningful well completion sequential patterns.
Keywords :
Internet; business data processing; data mining; mobile computing; pattern classification; PrefixSpan algorithm; Web access mode; business dependency strength calculation model; business process model; k-minimum weighted support technology; weighted sequential pattern mining algorithm; well completion business process; well completion mobile platform; Algorithm design and analysis; Analytical models; Business; Data mining; Databases; Educational institutions; Mobile communication; Data mining; K-minimum weighted support; Weighted sequence pattern; Well completion;
Conference_Titel :
Intelligent Control and Automation (WCICA), 2014 11th World Congress on
DOI :
10.1109/WCICA.2014.7053605