This paper extracts and analyses the most frequent natural language expressions used by participants in two real-world collaborative design sessions. The frequent expressions are extracted using the natural language processing technique of word sequence tokenisation. We then select a set of frequent expressions related to design changes. The boundaries of these are expanded by extracting the progression trends to identify specific speaker intents. These are analysed in detail while focusing on ambiguity and vagueness, aiming at identifying speakers’ precision when communicating design changes. We found that frequently used expressions play the role of language markers. In our discussion, we suggest that ambiguous and vague expressions used by participants can initiate mindful explorations of design space.

本文提取并分析了两个实际协作设计会议中参与者使用的最频繁的自然语言表达。使用单词序列标记化的自然语言处理技术提取频繁表达。然后，我们选择一组与设计更改相关的常用表达式。通过提取进展趋势以识别特定的说话者意图，可以扩大这些界限。在重点分析歧义和模糊性的同时，对这些内容进行了详细分析，目的是在传达设计变更时确定说话者的准确性。我们发现，常用的表达起着语言标记的作用。在我们的讨论中，我们建议参与者使用含糊不清的表达方式可以引发对设计空间的认真探索。