Mining Frequent Itemsets in Real Time

Abstract

Data streams are continuous, unbounded usually come in high speed and have a data distribution which often changes with time. They can be analogous to water stream which keeps on flowing on one direction. Data stream is used in many application domains—one of it being the analysis of data in real time also known as streaming data analytics. They vary to a large extent in case of storing in tradition database which needs the database to be a static one. The store and analyze strategy is not applicable in case of streaming data analytics. In order to analyze, it is required to find a certain amount of associativity among the data stream. Such associativity among the data items leads to the generation of association rules. These association rules are an important class of methods of finding regularities/patterns in such data. Using association rule mining all the interesting correlation amongst the data can be used to derive. These relationships in the data items, can go on to a large extent in helping larger transactions records in case of making a decision or drawing a conclusion out of it. This paper makes use of Apriori algorithm in data streams which can discard the non-frequent set of data-items and can finally obtain the frequent itemsets from them.