Wireless Communications and Mobile Computing

Research Article

Mining Profitable and Concise Patterns in Large-Scale Internet of Things Environments

The designed GMR-Miner algorithm.

Input:, a quantitative database; ptable, a profit table of all items; , a minimum utility count.
Output: a set of discovered closed high-utility itemsets (CHUIs)
perform -means to cluster as ().
perform exploration function {
for each {
set key-value pair as (tid, -itemset).
for each -itemset {
calculate .
}
write.
}
for each in {
.
}
write.
}
perform exploitation {
build tidset.
for each {
set key-value pair as (tid, -itemset).
for each -itemset {
calculate by CLS-Miner.
}
write
}
for each in {
.
}
write.
}
perform integration {
project -itemset as utility-list (UL).
build EUCS of 2-itemsets.
for each in {
check -itemset.
if appears in tidset and {
write a pair .
else {
write a pair .
}
}
}
for each in {
.
if {
write.
}
}
}