Comparative Analysis of Pearson and Euclidean Methods for Candidate Recruitment Optimization

Abstract

In the era of globalization and intense business competition, companies desperately need human resources (HR) that fit their culture and goals. High-quality human resources increase productivity and contribute to the innovation and development of the company. Therefore, effective and efficient recruitment is essential to get the best candidates. This research aims to determine the optimal method between Pearson correlation and Euclidean distance in candidate selection. The stages include several systematic steps. First, a simulated dataset with 100,000 candidates is created, each having ten attributes such as skills, experience, education, performance score, age, and certifications. Second, algorithms were applied using Google Colaboratory, with Pearson correlation to measure linear relationships between attributes and Euclidean distance to measure absolute distances between candidates. Third, the similarity results from both techniques were compared using a heatmap table. Analysis showed that Pearson correlation was more consistent and relevant in measuring similarity between candidates. In the visualization of the first 20 datasets, Pearson correlation performed better with a high and stable distribution of correlation values, the lowest value being 0.86. In contrast, Euclidean distance performed less optimally with lower and unstable values, the lowest being 0.26. Considering the stability and consistency of the results, Pearson correlation proved to be more effective and reliable for candidate selection. It provides a clearer and more consistent representation of the similarity between candidates, while Euclidean distance is more suitable as a complementary method or in situations that focus on absolute differences between attributes. Overall, Pearson correlation is recommended as the primary method of candidate selection.