植基於3NN+1投票機制的案例式推論系統及其在筆記型電腦市場區隔之應用

案例式推論常用於商業領域的顧客分群之研究，更是專家系統領域中發展迅速的一環。早期的案例式推論可能會有不佳的執行績效，其原因來自於未對案例或相關特徵進行篩選的動作。後續則有學者提出實例選擇、特徵選擇與同步最佳化選擇等方法對案例式推論系統進行改良。案例式推論的應用研究中，投票的機制通常是假設顧客分群為兩群體之情況為主，當顧客分群為三群體時，將無法有效解決三群體的推論問題。本研究希望提出一套3NN+1投票機制的案例式推論系統，結合基因演算法同時選擇最具代表性的參考案例與相關特徵。最後，本研究以筆記型電腦銷售產業進行實證，結果顯示本研究所提出之推論系統的分群準確度與實用性應具有協助企業進行策略性行銷之能力。

Purpose – Case-based reasoning system (CBR) has been widely applied to the issue of market segmentation. Most of previous studies focused on dividing customers into two groups. The voting method used in CBR may become inappropriate when one would like to divide customers into three groups according to some segmentation variable. Thus, a new voting method called 3NN+1is proposed to bridge the gap.
Design/methodology/approach – To make the inference of the 3NN+1 based CBR more efficient, the features and instances (or cases) for reasoning is selected simultaneously by means of genetic algorithms.
Findings – From the case of notebook market, the new CBR system is proved useful for segmenting the market of customer. The accuracy percentage of the case reasoning is about 57%.
Research/Practical implications – The new system presented here would be valuable when enterprises divide customers into three groups in compliance with their purchasing behaviors for developing marketing strategies.
Originality/value – This study would be a pioneer in the three-group classification problem with respect to CBR systems.