Hybrid Maximum Clique Algorithm Using Parallel Integer Programming for Uniform Test Assembly

Educational assessments often require uniform test forms, for which each form has equivalent measurement accuracy but with a different set of items. For assembly, an important issue is the increase number assembled tests. Although many automatic assembly methods exist, maximum clique algorithm (MCA)-based method known to assemble greatest tests highest based on item response theory. In that method, graph constructed by sequentially adding randomly formed as vertex without considering structure. However, difficulty its high space complexity, interrupts search cliques more than hundred thousand vertices. To overcome this difficulty, article proposes new algorithm: hybrid using parallel integer programming. The first step searches large possible up computer memory limitations state-of-the-art MCA low time complexity complexity. second repeatedly connected all vertices current from remaining programming proposed constructs larger traditional do. Finally, we use simulation and actual data experiments demonstrate effectiveness method. Results show our assembles 1.5–2.7 times greater can.