The problem of computational design for menu systems has been addressed in some specific cases such as the linear menu (list). The classical approach has been to model this problem as an assignment task, where commands are assigned to menu positions while optimizing for users’ selection performance and grouping of associated items. However, we show that this approach fails with larger, hierarchically organized menus because it does not take into account the ways in which users navigate hierarchical structures. This paper addresses the computational menu design problem by presenting a novel integer programming formulation that yields usable, well-ordered command hierarchies from a single model. First, it introduces a novel objective function based on information foraging theory, which minimizes navigation time in a hierarchical structure. Second, it models the hierarchical menu design problem as a combination of the exact set covering problem and the assignment problem, organizing commands into ordered groups of ordered groups. The approach is efficient for large, representative instances of the problem. In a controlled usability evaluation, the performance of computationally designed menus was $\sim 25\%$ faster to use than existing commercial designs. We discuss applications of this approach for personalization and adaptation.

在某些特定情况下，例如线性菜单（列表），已经解决了菜单系统的计算设计问题。经典方法是将此问题建模为分配任务，其中将命令分配给菜单位置，同时针对用户的选择性能和关联项目的分组进行优化。但是，我们表明，这种方法无法使用较大的，按层次结构组织的菜单，因为它没有考虑用户导航层次结构的方式。本文通过提出一种新颖的整数编程公式来解决计算菜单设计问题，该公式可从单个模型产生可用的，有序的命令层次结构。首先，它介绍了一种基于信息搜寻理论的新颖目标函数，该函数在分层结构中最大程度地减少了导航时间。第二，它将分层菜单设计问题建模为精确集覆盖问题和分配问题的组合，将命令组织为有序组的有序组。该方法对于大型，典型的问题实例是有效的。在受控的可用性评估中，通过计算设计的菜单的性能为$〜25％$比现有的商业设计更快地使用。我们讨论了这种方法在个性化和适应性方面的应用。