呈現方式對工件分類作業的影響

Abstract

利用工件自動分類系統可以快速地對工件進行搜尋及檢索，若自捩分 類系統與使用者心智模型符合的程度越高，則提取資訊的容易程度越高， 其實用性也越高。為選擇適當的自動分類系統，必須建立以人員的分類規 範做為衡量標準。以人員進行工件分類，如何用適當的方式將工件呈現出 來，才能有效表達工件的形狀和特徵，使人員正確辨認工件，判斷其相似 性，是一個值得探討的議題。 本研究以三十六個工件為實驗材料，以 大學在學男女學生為受試者，利用圖紙等角圖、實體模型及兩種電腦繪圖 等四種呈現方式，探討其對人員在工件分類作業的影響，並且以實體模型 方式的分類結果為基準，利用一致性指標，衡量其它三種呈現方式的分類 結果與其一致的程度。研究結果顯示，三種呈現方式的分類結果間存在顯 著差異；和實體模型方式分類結果的一致性比較，靜態電腦繪圖方式顯著 高於其它兩者，動態電腦繪圖方式的一致性最低，圖紙等角圖方式則介於 兩者之間。 Global shape based automatic workpiece classification systems have been developed for enhancing the effectiveness of part design. The utility of these automatic systems depend primarily on the extent to which its classification results are consistent with users' classification scheme. A user-based benchmark, which was based on users' judgment of the similarity between workpiece shapes, was proposed to evaluate the utility of automatic workpiece classification systems. In developing this benchmark standard, one issue should be taken into consideration is: in what way these workpieces should be presented to users so that their perceptual judgment can be performed accurately? Therefore, the purpose of the present study was to investigate the effect of workpiece presentation styles on workpiece classification. Sixty college students were divided into four groups. In each group, a set of 36 workpieces were presented to subjects in one of the four methods: isometric drawing, computer-based static solid model, rotating solid model, and physical model. Subjects were asked to perform paired- comparison on similarity between global shapes of workpieces. Using the classification result of physical model as the reference, a consistency index was developed to measure the consistency between its result and the results of another method. It was found that the consistency index of static solid model is the highest among the three methods, isometric drawing is the second, and the rotating solid model is the least. Implications of the result to the development of workpiece classification benchmark standard were discussed.