Incremental association rule mining: a survey by B. Nath, D. K. Bhattacharyya, A. Ghosh. (WIREs Data Mining Knowl Discov 2013. doi: 10.1002/widm.1086)
Abstract:
Association rule mining is a computationally expensive task. Despite the huge processing cost, it has gained tremendous popularity due to the usefulness of association rules. Several efficient algorithms can be found in the literature. This paper provides a comprehensive survey on the state-of-the-art algorithms for association rule mining, specially when the data sets used for rule mining are not static. Addition of new data to a data set may lead to additional rules or to the modification of existing rules. Finding the association rules from the whole data set may lead to significant waste of time if the process has started from the scratch. Several algorithms have been evolved to attend this important issue of the association rule mining problem. This paper analyzes some of them to tackle the incremental association rule mining problem.
Not suggesting that it is always a good idea to model association rules as “associations” in the topic map sense but it is an important area of data mining.
The paper provides:
- a taxonomy on the existing frequent itemset generation techniques and an analysis of their pros and cons,
- a comprehensive review on the existing static and incremental rule generation techniques and their pros and cons, and
- identification of several important issues and research challenges.
Some thirteen (13) pages and sixty-six (66) citations to the literature so a good starting point for research in this area.
If you need a more basic starting point, consider: Association rule learning (Wikipedia).