| 摘要: | 現有之住宅文化，受台灣近代發展歷史之波折而有很深的影響，其中最明顯的觀念即是”住者有其屋”。受此思考影響，即使房價自1987年後持續上漲，住宅自有率仍持續上升；但在此同時，相較於平均收入房價已超出合理之範圍，因此低收入戶之住宅自有率呈現負成長。也因如此，在無法強制介入住宅之自由市場或快速改善狀況的前下，住宅補貼即成為照顧弱勢族群不可或缺的政策。
政府原先按照辦理職業別、身分別給予各項政策性房屋貸款措施不同額度與利率，且同質之政策性房屋貸款措施分散在數個部會各自辦理，造成住宅資源及人力之有效利用之效率折減。行政院經濟建設委員會於94年8月3日提出：「有關各部會以職業身分別辦理之住宅補貼計畫，自95年度起不再新增，94年度以前已核定之各項住宅補貼措施，仍繼續執行至結束，未來住宅補貼業務統一由內政部辦理。」。
此補貼方案自開辦以來，每年申請人數皆逾萬人，為一龐大之資料庫。除利用傳統統計方式分析基本特性外，更因資料數量豐富，適用於資料探勘之分析方法，因此本研究即利用資料探勘中之群啟發演算法對住宅補貼申請資料進行分群，使其資料自行群聚，顯現各群組所特有之性質，進行人為解釋，提供後續相關政策修改與執行之參考。
研究結果可看出住宅補貼審查之最大重點為經濟狀況，受補貼者為經濟弱勢，但不一定為弱勢族群：即代表，透過行之有年的眾多補助與福利協助後，弱勢族群已漸漸不完全等於經濟弱勢族群。最後，研究成果可提供社會局、內政部營建署等住宅補貼相關機關，作為後續住宅政策調整研擬之參考。
;The existing housing situation in Taiwan –The history, development and modern changes of Taiwan’s economy have had a strong impact on Taiwan’s housing situation effecting the idea of home ownership. Although prices of homes have risen steadily since 1987, home sales have also continued to increase. At the same time average family income has risen to levels beyond the scope of reasonable pricing; and therefore home ownership rates for lower-income households has dropped. This is also the reason why a ‘free market’ system cannot be enforced during these periods of rapid expansion, so the establishment of housing policies to support the disadvantaged is essential.
In accordance with the original structure, various policy measures were established based on occupation and the background of the applicant, with homogeneous interest rates and policy measures. These were dispersed and handled separately by several ministries, resulting in an efficient and effective use of resources and manpower; but not completely efficient. On 3 August 2006, the Council for Economic Affairs proposed that ‘Relative departments’ would conduct a professional review of the Residential Subsidy Program, since the program was based on subsidies that were approved before the year of 2005; and in the future would operate at the same level, and will be managed by the Ministry of the Interior.
Since the inception of the subsidy program, the number of applicants exceeds 10,000 annually. This is a sizeable database. In addition to using traditional, statistical methods to analyze the basics, and also because of the vast amount of data for analyzing the results of the data mining, this study is in the use of data-mining algorithms - inspired by the group to apply for housing subsidy. Data for ‘clustering’ by individual groups reveals the unique nature of each group, and provides follow-up for operational and policy changes to each related reference.
In its broadest scope, results can be realized regarding the economic situation of housing subsidies in a weak economy; but not necessarily aimed at ‘vulnerable’ groups. Within the program, assistance through a number of grants and benefits has existed for years without equality necessarily being directed to the economically disadvantaged groups. Finally, this research can provide data for the establishment of social benefits such as housing subsidies to the relevant authorities of the Ministry of the Interior to assist in the development of housing policy adjustments. |
