摘 要 臺灣餐飲業面臨『員工流動率過高』與『人力招募不易』兩個問題。因此,如何降低人員流動率,並且招募到適任的人員是相當重要的課題。故本研究將以火鍋業作為主要研究對象,並藉由訪談不同職責的員工,探討其『人格特質』與『個人與組織契合』是否具有相似性,並透過學習向量量化網路去進行分類預測,以了解『人格特質』與『個人與組織契合』的關係,進而幫助餐飲業在招募不同職責的員工時,能選擇較合適的員工,降低餐飲業中的人員流動。 本研究嘗試將學習向量量化網路分類預測的特性應用於員工招募。利用因素分析將互補性契合、補充性契合、人格特質簡化成數個構面,再將這幾個構面作不同的組合,並透過學習向量量化網路來建立網路模式,以建立最佳的網路模式。 本研究所構建之網路模型以分類正確率為評估指標,其最佳正確率為86%,並透過此模式計算出各特質類型之形心,以供業者作為員工招募之參考。
ABSTRACT In Taiwan the Food and Beverage Service Industry faces two problems: labor turnover rate is high and manpower recruitment is difficult, therefore, how to reduce labor turnover rate and recruit competent labor is a very important issue. The object of this study is hot pot industry, this study explores “Personality Traits” and “Person -Organization Fit” by interview with different responsibilities of labor. This study also uses Learning Vector Quantization Neural Network to classify, forecast and understand the relationship between “Personality Traits” and “Person-Organization Fit”, helps restaurant industry choose suitable labor and reduce labor turnover rate when recruit different responsibilities of labor. This study attempts to use the classification characteristic of Learning Vector Quantization Neural Network for the recruitment of labor. First, we use factor analysis to simplify supplementary fit, complementary fit and personality traits into several dimensions and create different combinations from the dimensions. Then, this study establishes the best network model by Learning Vector Quantization, the assessment indicator of study is correct classification rate, and the best correct classification rate is 86%. This study also calculates the centroid of characteristic type by the model and as the reference when the industry recruits competent labor.