購屋決策支援系統開發之研究

Developing A Housing Purchase Decision Support System

鄭敬錞(Ching-Chun Cheng)

指導教授 ： 彭建文

國立臺北大學/公共事務學院/不動產與城鄉環境學系/碩士(2010年)

摘要

本研究發展一套以消費者決策程序模式(Consumer Decision Process model, CDP model)整合簡單加權法(Simple Additive Weighting, SAW)與分析層級程序法（Analytic Hierarchy Process, AHP），建立符合購屋決策歷程的購屋決策支援系統，讓購屋決策者有邏輯地來思考購屋決策的各項歷程，並改善過去研究將AHP法評估準則權重由專家決定的缺點，改為讓決策者自訂權重，同時考量購屋者的居住需求、負擔能力、居住品質偏好，進行整體系統規畫，並結合智慧型代理人程式進行搜尋物件，設計看屋記錄結合決策方法輔助決策者做比較評估。以web-based方式開發，並透過整合CDP模式及決策方法，建構可以擴充決策方法的雛型系統架構。
本文進一步探討SAW與AHP兩種決策方法於購屋決策上的適用性，其結果為，SAW法適用在「決策者較不熟悉評估準則」或是「需要一次評估大量準則」時；AHP法適用在「決策者對評估準則有相當熟悉程度」、「決策者對於決策結果要求較為精準」時。當AHP法應用於比較方案條件差異較小時，挑選準則的方法頇做調整，其結果才能較為準確。本文發現AHP法在購屋決策上有一致性檢定之問題，歸納為六大類原因：「決策者對於評估準則的重視程度還不是很清楚」、「因描述尺度的差異」、「決策者所挑選出的評估準則中，有非常極端的情況」、「決策者在兩兩比較時，不會考量到其他準則間的相對關係」、「面對同一面項評估時，同時考量兩種房屋類型市場」、「一致性檢定的門檻值可能太高」。
個案研究中發現，購屋者間的偏好結果是屬於「評估準則集中，但個案間重視程度不一」，房屋內部品質集中在「房屋面積大小」、「通風與採光」、「格局與使用舒適性」、「風水」，社區環境品質集中在「居民素質」、「社區規模大小」、「管委會與規約健全」、「社區門禁保全與清潔」、「消防設備與逃生動線」，鄰里環境品質集中在「大眾運輸的便捷性」、「空氣品質與噪音」、「治安與附近居民素質」、「家人至就業、就學距離」、「購物便利性」。
購屋決策支援系統開發面臨問題，有三大面向，分別為系統限制面：「不同房屋市場類型的評估準則不同」、「評估準則構面選擇不同，必頇依準則間的關係選取適合的決策方法」；專家輔助面：「房貸方式建議」、「區位轉換建議」、「偏好市場類型轉換」、「特殊需求」；功能輔助面：「客製不同智慧型代理人程式技術困難」、「不易動態決定議價空間」、「定義評估準則之細部操作型指標困難」、「詳細評估內容如何呈現」、「尚無完整房屋真實交易資訊提供整合」，建議未來設計購屋決策支援系統必頇優先考量此三大面向進行改善。

關鍵字

CDP模式,購屋決策,系統開發,AHP,簡單加權法

並列摘要

is only suitable for use when decision makers are unfamiliar with the evaluation criteria or when they must evaluate numerous criteria; while AHP is suitable for use when decision makers are relatively familiar with the evaluation criteria and when they prefer relatively precise decision results. If the significance of the compared items is weak using the AHP method, the criteria selection method must be modified to improve the precision of the results. This study found that consistency test may become an issue regarding house-purchasing decisions using the AHP method. The six major causes are as follows: (1) decision makers are unfamiliar with the weighting of the evaluation criteria; (2) differences in the description scales; (3) the evaluation criteria the decision maker selected contain extreme conditions; (4) decision makers tend to neglect the relation to other criteria than the two criteria decision makers are
iii
comparing; (5) decision makers simultaneously assess two types of housing markets in one evaluation criterion; and (6) the threshold values of consistency test are too high.
The results of the case study reveal that house buyers prefer the evaluation criteria to be centralized, but the subjects to have different weightings. Additionally, they prefer that the inner quality of houses concentrates on the “size of the houses,” “ventilation and lighting,” “configuration and residential comfort,” and “feng shui;” the environmental quality of the community focuses on the “behavior of neighbors,” “size of the community,” “management committee and comprehensive regulations,” “security system and clean community environment,” and “fire-fighting equipment and evacuation routes;” and finally, that the environmental quality in the neighborhood centers on the “convenient access to public transport,” “air quality and noise level,” “public security and neighbor behavior in neighborhood,” “distance from homes to employment/school,” and “the convenience of shopping.”
There are three obstacles for developing a support system for house-purchasing decision. First, the dimension of the system limitations includes “different evaluation criteria for various types of housing markets” and “various evaluation criteria—decision makers must select suitable decision-making methods according to the relations between evaluation criteria.” Second, the expert dimension includes “suggestions for housing loan types,” “suggestions for housing area change,” “change of preferred housing market types,” and “special demands.” Finally, the functional dimension includes “technical difficulty of customizing intelligent agent system,” “fixed and unchangeable bargaining zone,” “difficulty defining the detailed operating indicators of evaluation criteria,” “method for presenting detailed evaluation,” and “lack of complete actual trading information for integration.” Therefore, this paper suggests that these three major obstacle dimensions must be addressed first when developing a decision-making support system in terms of house-purchasing.

並列關鍵字

CDP model,housing decision-making,system-developing,AHP,SAW

參考文獻

張嘉純，（2009），「台灣房地產價格與房屋貸款之關聯性研究」，臺灣大學國家發展研究所碩士論文。

連經宇、陳彥仲，（1999），「模糊語意變數法應用於住宅消費決策行為之初探研究」，住宅學報，第8期，頁69-90。

陳彥仲，（1997），「住宅選擇之程序性決策模式」，住宅學報，第5期，頁37-49。

張珩、邢志航，（2004），「永續理念下「居住空間規模」推估模式建構及實證之研究－以公寓大廈式集合住宅為例」，建築學報，第45期，頁99-120。

藍晟豪，（2007），「消費者購屋決策與網站服務品質影響因素之研究」，元智大學企業管理學系碩士論文。