| Popis: |
Mining frequent patterns are most widely used in many applications such as supermarkets, diagnostics, and other real-time applications. Performance of the algorithm is calculated based on the computation of the algorithm. It is very tedious to compute the frequent patterns in mining. Many algorithms and techniques are implemented and studied to generate the high-performance algorithms such as Prepost+ which employees the N-list to represent itemsets and directly discovers frequent itemsets using a set-enumeration search tree. But due to its pruning strategy, it is known that the computation time is more for processing the search space. It enumerates all item sets from datasets by the principle of exhaustion and they don’t sort them based on utility, but only a statistical proof of most recurring itemset. In this paper, the proposed Enhanced Ontologies based Alignment Algorithm (EOBAA) to identify, extract, sort out the HUI's from FI's. To improve the similarity measure the proposed system adopted Cosine similarity. The experiments conducted on 1 real datasets and show the performance of the EOBAA based on the computation time and accuracy of the proposed EOBAA. |
