Adaptive Beam Search Outperforms Mainstream School Timetabling Methods

A two-stage adaptive beam search framework significantly outperforms mainstream algorithms and traditional manual scheduling methods in both solution quality and computational efficiency, according to a study published in Complex & Intelligent Systems. Developed to address the novel timetabling challenges arising from China’s “3+1+2” college entrance examination reform, the computational framework resolves the heightened complexity of personalized education requirements by decomposing the problem into two interrelated stages: class formation and course timetabling.

In the first stage, mathematical optimization models and tailored adaptive beam search algorithms are established to manage class formation. The research team implemented a greedy strategy integrated with a state-aware beam search to optimize student assignment and paired-class matching under the structural requirements of the reformed curriculum.

For the subsequent course timetabling stage, the framework introduces a multi-neighborhood search structure that explicitly incorporates paired-class constraints. This timetabling algorithm operates alongside a dynamic neighborhood selection mechanism designed to enhance the operational efficiency and search effectiveness of the beam search process.

Experimental evaluations using real-world data from a representative high school demonstrated that the proposed method significantly exceeds both mainstream algorithms and manual timetabling across solution quality and computation speed. Subsequent ablation studies confirmed the effectiveness of the adaptive mechanisms, indicating that the framework provides an efficient approach for personalized timetabling alongside a scalable paradigm for broader resource scheduling problems.

Subject of Research: Personalized school timetabling optimization using adaptive beam search

Article Title: A two-stage adaptive beam search framework for personalized school timetabling under complex curriculum reform constraints

Article References: Song, T., Chen, M., Xu, Y., Dang, H., Lv, Y., & Li, P. (2026). A two-stage adaptive beam search framework for personalized school timetabling under complex curriculum reform constraints. Complex & Intelligent Systems. https://doi.org/10.1007/s40747-026-02578-9

Image Credits: AI Generated

DOI: 10.1007/s40747-026-02578-9

Keywords: adaptive beam search, school timetabling, curriculum reform, class formation, course timetabling, paired-class matching, student assignment, multi-neighborhood search, dynamic neighborhood selection, mathematical optimization, combinatorial optimization, resource scheduling, computational efficiency, greedy strategy, state-aware search, educational scheduling, ablation study