The researchers proposed a comprehensive measurement method for novelty, aiming to quantify potential distances by integrating the networks, semantics, and hierarchical relationships between knowledge units. The study was analyzed and validated based on 142,036 biomedical papers published in PLoS ONE and the validation dataset from the H1 Connect platform. The results showed that the three types of relationships respectively captured different characteristics of potential distances between MeSH terms; compared to the metrics proposed by Uzzi et al. (2013), this measurement method demonstrated stronger consistency with peer review results; moreover, combining all three distance metrics was more effective at identifying novel papers than a single perspective.