A simple implementation of the apriori algorithm. A method for extracting frequent substructures in a set of sequences of ordered events.
Apriori takes a list of strings, representing sequences, and an integer, representing the percentage of sequences the pattern must match for being considered.
In [1]: from apriori import *
In [2]: data = ["ABCDEFGHIJKL","ZOPQABCDLMNOP","REWQZOPQAB"]
In [3]: patterns = Apriori(data, 34)
In [4]: patterns
Out[4]:
{'AB': 3,
'ABC': 2,
'ABCD': 2,
'BC': 2,
'BCD': 2,
'CD': 2,
'OP': 2,
'OPQ': 2,
'OPQA': 2,
'OPQAB': 2,
'PQ': 2,
'PQA': 2,
'PQAB': 2,
'QA': 2,
'QAB': 2,
'ZO': 2,
'ZOP': 2,
'ZOPQ': 2,
'ZOPQA': 2,
'ZOPQAB': 2}-
"Mining Frequent Patterns, Associations, and Correlations" (Chap. 5) in Han, J., Kamber, M., & Pei, J. (2006). Data mining: concepts and techniques. Morgan kaufmann.
-
Mooney, C. H., & Roddick, J. F. (2013). Sequential pattern mining--approaches and algorithms. ACM Computing Surveys (CSUR), 45(2), 19.