부산대학교 도서관

Information and communication technology development has made human resource assessment a top concern for enterprises. However, a key challenge with human resource assessment is adapting to the rapidly changing environment and utilizing various data types for decisionmaking. Unfortunately, the transmission latency of human resource assessment data in cloud computing platforms can negatively impact user experience. The emergence of mobile edge computing provides a new option for computing-intensive tasks like human resource assessment. A new objective function has been proposed to optimize the task offloading strategy, along with a mathematical model that utilizes the deep recurrent Q-network (DRQN) algorithm to leverage multiple Internet of Things (IoT) nodes. As the amount of human resource assessment data collected by IoT nodes continue to grow, extracting the potential laws and assisting human resources in making informed decisions is increasingly important. A fuzzy weighting clustering algorithm has been proposed to address the difficulty of determining the initial cluster number and center in data clustering based on soft subspace combined with a competitive merging mechanism. Simulation results demonstrate that the DRQN algorithm performs well in terms of energy consumption, cost, and latency, proving that the data collected by IoT nodes can meet the needs of human resource assessment and management after real-time processing. The proposed data clustering algorithm also accurately clusters human resource assessment data, meets high-dimensional data clustering needs, and has significant practical application value.