Dynamic Real-Time Talent Team Discovery Based on Complex Network Analysis

The real-time and precise identification of talent teams can offer essential support for talent development and augmentation. It plays a key role in efficiently and precisely assembling professional talent and constitutes a significant research focus within human resources. This paper utilizes complex network concepts and artificial intelligence methods to analyze dynamically and mine talent teams in real-time within evolving environments. It proposes a method of dynamic real-time talent team discovery that considers the academic relationships and research direction information of talent simultaneously. Firstly, a unified graph modeling method is proposed to fuse academic relationships and research direction information. Then, we apply non-negative matrix factorization to obtain talent team partition at each time snapshot. Additionally, we employ the evolutionary clustering framework to smooth talent evolution, leveraging historical evolution information to achieve accurate talent team discovery at a continuous time. Extensive experiments demonstrate that the proposed method achieves excellent results on real datasets. The method has proven effective in practical applications, enabling accurate and timely talent team discovery.