資料挖掘(data mining, DM)被廣泛的應用在許多領域，惟目前較少將其運用在房屋交易媒合上。本研究是將資料挖掘技術中之最近鄰居演算法(nearest neighbor algorithm, NN)運用在房屋交易媒合系統上，最近鄰居演算法是一種實用且準確率高的演算法，運用在媒合系統(matching systems)將有效的提高媒合系統的媒合能力，以輕易的讓使用者媒合到所需要的資訊。
房市的漲跌關乎國家的經濟，許多人一生中或許只有一次購屋機會，購買房屋對大部分的人來說可以說是僅次於婚姻的人生大事，因此選擇房屋對這些人來說就像選人生的另一半一樣，為了改善目前房屋仲介網所採用傳統的條件式搜尋方式來搜尋房屋的缺點，讓使用者不至於浪費找尋時間卻無法搜尋到理想的房屋，本研究預期建立一個以資料挖掘為基礎，能符合使用者需求的線上房屋交易媒合系統，為的是讓使用者可以在最短時間內搜尋到理想的房屋，進而改善原有傳統的條件式搜尋之缺點與效率。
Data mining techniques such as market basket analysis, case-based reasoning, automated cluster detection, and neural networks have been widely used in many areas. However, there is still lack of successful application in the area of house trading and matching so far. This research tries to apply one of the data mining techniques called the nearest neighbor algorithm to house trading and matching in order to strengthen the matching ability in house trading and matching systems. The nearest neighbor algorithm has been viewed as a practical and highly accurate method which is able to increase the accuracy and efficiency of matching systems. It can also help system users obtain the information they need more easily and promptly.
Many people may have only one chance in his or her life to buy a house; therefore, buying a house is as important as getting married with someone. In contrast to traditional house trading and matching systems, which are normally based on whether the input conditions have been matched, this paper proposes a similarity-based house trading and matching system that compares the user’s needs to the objects or cases in the database to decide which house is most suitable for the user.
