求翻译：如果计算时间很长, 我们将通过如 Beam 的局部搜索来优化词义组 合. 这种方法背后的操作是通过应用启发式的技巧 来缩减搜索空间. 在搜索过程的每个阶段, Beam 搜 索器将限制 k 个最有前景的候选, 而 k 是预先定义的数.是什么意思？

If the calculated time is long, we will adopt such as local search to optimize the combination of the meanings of the Beam. behind this method of operation is to apply heuristic techniques to reduce the search space. At each stage of the search process, Beam Finder will restrict k-most promising candidate, and k is the number of predefined.