weclat: Mining Associations from Weighted Transaction Data with Eclat...

In arules: Mining Association Rules and Frequent Itemsets

Mining Associations from Weighted Transaction Data with Eclat (WARM)

Description

Find frequent itemsets with the Eclat algorithm. This implementation uses optimized transaction ID list joins and transaction weights to implement weighted association rule mining (WARM).

Usage

weclat(data, parameter = NULL, control = NULL)

Arguments

data	an object that can be coerced into an object of class transactions.
parameter	an object of class ASparameter (default values: support = 0.1, minlen = 1L, and maxlen = 5L) or a named list with corresponding components.
control	an object of class AScontrol (default values: verbose = TRUE) or a named list with corresponding components.

Details

Transaction weights are stored in the transactions as a column called weight in transactionInfo.

The weighted support of an itemset is the sum of the weights of the transactions that contain the itemset. An itemset is frequent if its weighted support is equal or greater than the threshold specified by support (assuming that the weights sum to one).

Note that Eclat only mines (weighted) frequent itemsets. Weighted association rules can be created using ruleInduction().

Value

Returns an object of class itemsets. Note that weighted support is returned in quality as column support.

Note

The C code can be interrupted by CTRL-C. This is convenient but comes at the price that the code cannot clean up its internal memory.

Author(s)

Christian Buchta

References

G.D. Ramkumar, S. Ranka, and S. Tsur (1998). Weighted Association Rules: Model and Algorithm, Proceedings of ACM SIGKDD.

See Also

Other mining algorithms: APappearance-class, AScontrol-classes, ASparameter-classes, apriori(), eclat(), fim4r(), ruleInduction()

Other weighted association mining functions: SunBai, hits()

Examples

## Example 1: SunBai data
data(SunBai)
SunBai

## weights are stored in transactionInfo
transactionInfo(SunBai)

## mine weighted support itemsets using transaction support in SunBai
s <- weclat(SunBai, parameter = list(support = 0.3),
		       control = list(verbose = TRUE))
inspect(sort(s))

## create rules using weighted support (satisfying a minimum
## weighted confidence of 90%).
r <- ruleInduction(s, confidence = .9)
inspect(r)

## Example 2: Find association rules in weighted data
trans <-  list(
    c("A", "B", "C", "D", "E"),
    c("C", "F", "G"),
    c("A", "B"),
    c("A"),
    c("C", "F", "G", "H"),
    c("A", "G", "H")
)

weight <- c(5, 10, 6, 7, 5, 1)

## convert list to transactions
trans <-  transactions(trans)

## add weight information
transactionInfo(trans) <- data.frame(weight = weight)
inspect(trans)

## mine weighed support itemsets
s <- weclat(trans, parameter = list(support = 0.3),
		       control = list(verbose = TRUE))
inspect(sort(s))

## create association rules
r <- ruleInduction(s, confidence = .5)
inspect(r)

arules documentation built on Sept. 11, 2024, 8:15 p.m.