Mining Rules From Distributed Big Data Repositories: A Novel Scheme Fo Valuable Association Rule Detection
Rapid advancements made in technology have increased the growth of information on the internet. It becomes a challenging process for the users to suggest the right decision at the right time. The real-time issues are to be explored. It is observed that some items are not classified properly and thus leads to improper recommendation processes under a different context. Henceforth, the enhancement of the contextual information will improve the performance of the distributed databases. In our research study, this scalability of big transactional database issues is resolved by proposing a novel consistent and inconsistent rule detection using frequent pattern mining under a big data context. The big sales data is collected from a public repository and’s preprocessed by the elimination of null transactional records in the databases. Then, it is formulated as zone-wise transactional databases. The apriori algorithm is employed to find out the patterns which are frequent from the sales data. With the estimated interestingness value, the rules are framed, detected, and classified into consistent and inconsistent rules for each zone. Experimental results have shown the efficiency of the proposed technique.