[r] agrawal95, "Mining Sequential Patterns"

@inproceedings{ agrawal95mining,
   author = "Rakesh Agrawal and Ramakrishnan Srikant",
   title = "Mining sequential patterns",
   booktitle = "Eleventh International Conference on Data Engineering",
   publisher = "IEEE Computer Society Press",
   address = "Taipei, Taiwan",
   editor = "Philip S. Yu and Arbee S. P. Chen",
   pages = "3--14",
   year = "1995",
   url = "citeseer.ist.psu.edu/agrawal95mining.html" }

Motivation:
To mine patterns of sequences.
Contributions:
Clearly defined the sequence mining problem. Defined support in this problem. Gave the whole process for sequence mining and 3 algorithms for the sequencing phase.
Methods:
Sequence mining can be done in 5 phases: sort, large itemset, transformation, sequence and maximal. Details can be found in paper, pay attention to the large itemset finding part and don't need to spend much on the 3 sequencing algorithms.
Discussions:
This algorithm can find all maximal frequent sequences. However, the apriori nature makes its performance low. More recently, J.Han et al did research on sequence mining performance.