一、 研究動機及目的:本研究旨在應用類神經網路建構一阻礙推論模型,協助休閒業者透過家庭生命週期及人口統計變數來推論民眾可能遭遇的休閒阻礙;並搭配阻礙協商策略,探討民眾對阻礙的協商意願,以供業者於擬定差異化行銷策略時作為參考依據。 二、 研究方法及設計:本研究以國立自然科學博物館為研究主題,使用問卷調查法蒐集樣本後,透過集群分析將民眾阻礙因素分為不同的阻礙族群;再應用類神經網路建構家庭生命週期以及人口統計變數與阻礙族群間的連結關係;另外再以敏感度分析來了解各變數對阻礙族群推論的影響力;最後觀察各阻礙族群對阻礙協商策略的採用意願。 三、 研究結果:阻礙族群可分為時間阻礙、外在阻礙以及無阻礙三群。由阻礙推論模型之推論結果得知,該模型確實可透過家庭生命週期以及人口統計變數來推論出民眾所屬阻礙族群。而性別、教育程度及家管對休閒阻礙的影響符合文獻的研究結果,但年齡及收入對阻礙族群推論的影響力皆偏低。本研究亦發現到時間阻礙族群的民眾較有意願採用「節省金錢花費」以及「尋找到達科博館的方法」這兩個策略進行阻礙協商;而外在阻礙族群則多採用「說服同伴一起參與」以及「減少其他活動花費的時間」這兩項協商策略。 四、 研究貢獻:本研究結果可望對博物館業者進行市場行銷時達到輔助效果,藉以提升行銷策略的達成率,吸引更多民眾參觀博物館。
Purpose: The aim of this study is to develop constraints inference model applying neural network and to assist leisure operators infer what leisure constraints people experienced through family life cycles and demographic variables. Besides, this study also explores the willingness of constraints negotiation of the respondents. Design/methodology/approach: This paper focuses on the barriers to visiting The National Science Museum in Taiwan, and collects the samples by questionnaire survey method. First, we use K-means to classify the respondents into various groups. And then, we applying neural network to explore the relations of family life cycle and demographic variables to leisure constraints. Furthermore, we attempt to understand the influence of each variable on inferring constraint groups with sensitivity analysis. Finally, we observe the adoption intention with negotiation strategy of each constraint group. Findings: There are three groups we clustered from respondents, namely time constraints, external constraints, and no constraints. Factually, this model can effectively infer the constraint group to which an individual belongs to base on his or her identification. The influence of gender, education level and housewife conform to the literature, but the influence of age and income are low. “reducing costs” and “searching for the ways to museum” are with higher adoption intention in time constraints. “Persuading friends to participate together” and “reducing the time from other activities” are with higher adoption intention in external constraints. Originality/value: Museum operators can use this model and the results as an aid to develop effective negotiation strategies to induce participation of potential visitors. Keywords: Leisure constraints, constraint groups, neural network, constraint negotiation