以彈性群聚法建立評估工件自動分類系統之基準

Abstract

群組技術係將相似的工件分門別類組成工件族，利用其相似的特性縮短設計時間或提高生成效率。目前已發展出許多工件自動分類系統，但是如何評估此等系統分類的結果是否能符合使用者需要，尚無相關文獻探討。本研究之目的即在建立一個標準分類系統，用以評估工件自動分類系統之實用性。此標準分類系統具有以下兩種特性：(1)以使用者認知做為分類的基礎，可有效得知工件自動分類系統與使用者心智模型相容的程度；(2)它可依使用者對工件相似程度的認知，彈性提供適當的分類結果當做分類規範使用。這項評估技術，可以選擇最具實用性的工件自動分類系統，用來建立工件分類資料庫，提供各部門人員檢索運用。 但是建立標準分類系統需使用受試者對一組抽樣工件執行兩兩相似評比，以全數實驗數據進行分類，當抽樣二件數增多時成本遞增，方法上有所限制。因此，本文再提出可用局部實驗數據所建立的精簡分類系統，來衡量工件自動分類系統的績效。

In group technology workpieces are categorized into families according to their similarity in desing or manufacturing attributes. This technique can eliminate design duplication and facilitate the production. Much effort has been focused on automated workpiece classification systems development. However, it is difficult to evaluate the performance of thess system. A benchmark calssification system was developed based on golbal shape information to evaluate the utility of workpiece calssification systems. A classification system has a high level of utility if its classification scheme is consistent with users' mental models of the similarity between workpiece shapes. Hence in the proposed method the consistency between a classification system and users' mental models is used as an index of the utility of the system. The proposed benchmark classification has two salient characteristics: (1) it is user-oriented, because it is based on users' judgments. (2) it is flexible, allowing users to adjust the criteria of similarity applied in the automated workpiece calssification. Such benchmark classification is typically established by having subjects to perform complete pair comparisons of all sample workpieces. However, when the number of sample workpieces is very large, such exhaustive comparisons become impractical. This paper also proposes an efficient method, called lean classification, in which data on comparisons between the samples and a small number of typical workpieces are used to infer the benchmark classification results.