本研究採用可以增加時間條件限制之序列型態探勘中的SPADE演算法，試圖找出在視窗限制範圍內的連續事件的關聯性，以2007年12月至2015年9月之臺灣上市櫃公司為研究對象，藉以了解領先指標及企業長短期借款變化之關聯性。 實驗結果發現當相對多項領先指標同步領揚時，景氣看似榮景，但最能代表企業現況的IEO外銷訂單指數及SEMI半導體接單出貨比若在此時走跌時，預測下一期企業長短期借款需求將下降。 ;In order to understand the correlation of leading indicators and corporate changes in the long and short term loans, this study was used SPADE which can increase the time limit of the sequent patterns mining algorithms trying to find relevance in the window limits of consecutive events with objects-listed companies in Taiwan dated from December 2007 to September 2015. It was found that the economy seems booming when the number of leading indicators rise up; however the forecasting of the next period of short and long term corporate loan demand will decline if the most representative indicators-IEO (Index of Export Orders) and the SEMI-conductor book-to-bill ratio fall down.
