Efficient Apriori Rank Pruning Model Based Novelty Detection with One-Class SVM IN Online Pharmaceutical Platforms

In the dynamic landscape of eCommerce, the ability to swiftly identify and mitigate anomalies within product transactions holds paramount importance. This work presents a novel approach to outlier detection through the integration of a rank pruning model with significance itemsets. Leveraging advanced data mining techniques, this methodology strives to enhance the efficiency and accuracy of anomaly detection in eCommerce transactions. The proposed framework commences with the collection and preprocessing of e-pharmacy transaction data, encompassing essential attributes such as customer and product identifiers, transaction amounts, and timestamps. The EARPM(Efficient Apriori Rank Pruning Model) presents an innovative approach to detecting outliers in eCommerce transaction data using the combination of Apriori algorithm and rank pruning techniques. By Implementing One-Class SVM Model as outlier detection mechanism, that uses the rank information to identify transactions with low ranks as potential outliers. With a focus on practicality and real-world applicability, the proposed system culminates in deployment within the eCommerce platforms backend.