A talent portrait construction and matching method and system based on a knowledge graph

Abstract

This invention discloses a method and system for constructing and matching talent profiles based on knowledge graphs. The method includes: collecting multimodal talent-related data from multiple sources using a standardized interface and performing formatted preprocessing on the multimodal data; extracting modal features from the preprocessed multimodal data and aligning and fusing features through a cross-modal attention mechanism to generate a unified multimodal fusion feature representation; extracting multi-dimensional labels based on the multimodal fusion feature representation to generate a fine-grained talent profile; constructing a knowledge graph in the human resources domain; receiving job requirement description information, performing relational reasoning and intelligent matching based on the human resources domain knowledge graph, and generating a comprehensive matching score, a recommended talent list, and explanations of matching reasons. This invention achieves multimodal data fusion-based talent profile construction and intelligent job matching based on knowledge graph relational reasoning in a multi-source heterogeneous human resources data scenario.

Images