A transactional database is a set of transactions.
Each transaction contains a transaction-identifier (TID) and a set of items.
Example:
A sample transactional database containing the items from a to f is shown in below.
TID Transactions 1 a, b, c
2 d, e
3 a, e, f
Rules to create a transactional database:
- Since the TID of a transaction directly represents its row number in a database, we the algorithms in PAMI ignore the TID information to save storage space and processing time.
- The items in a transactional database can be integers or strings.
- All items in a transaction must be seperated with a separator.
- ‘ Tab space ’ is the default seperator used by the mining algorithms in PAMI. However, transactional databases can also be constructed using other separators, such as comma and space.
Format:
>>> item1<sep>item2<sep>...<sep>itemN
Example:
- >>> a b c
a d e f b d
frequent1
relativeFrequent
FrequentPatternWithMultipleMinimumSupport1
CorrelatedPatternMining1
FaultTolerantPatternMining1
CoveragePatternMining1