Profit Based Products Recommender System for Seller Zone

Abstract

Finding Common Patterns in Transactional Databases MLP algorithms are considered one of the most important. The MLP algorithm is a good association rule mining algorithm. As the potential for multi-core processors increases, algorithms and applications need to be updated to take advantage of the computing power of multiple cores and find sets of items more frequently in terms of computational resource utilization and CPU performance. Algorithms for parallel mining of sets of frequent elements provide directions for solving the problem of candidate distribution across processors. The dynamic and skillful use of algorithms to recognize common patterns is important in data mining research. The purpose of MLP algorithms is to find connections between different data sets. Each unique record has a set of elements, called a transaction. MLP functionality consists of a set of rules that indicate how often an item appears in a record. In order to find more valuable rules, our underlying goal is to implement the MLP algorithm using a multi-threaded approach. This allows you to intelligently and effectively use the power of your system hardware to improve your algorithms and extract more valuable information. Serial mining consumes time and reduces mining performance. In the proposed system, the MLP algorithm is implemented in serial and parallel manners, and comparisons of both are made based on various support counts and times using parallel programming techniques.