台灣觀光產業與旅遊活動近年蓬勃發展，然而觀光地區卻常因交通問題而影響旅遊品質。智慧型運輸系統(Intelligent Transportation Systems, ITS)在台灣已行之有年，近年來更有熱門觀光景點陸續導入ITS以改善交通問題。為了解其導入後如何協助交通管理策略規劃以及導入之成效，本研究以「2012年台灣十大觀光小城─大溪總統鎮」為對象，藉由ITS評估實施交通策略之時機，並探討影響交通策略執行潛在的道路特性與相關影響項目，利用資料探勘法之分類決策樹與關聯關係演算法，進行策略執行之門檻值設定與策略宣導之效益分析。研究結果發現車輛偵測器對於預測壅塞發生助益良多，且在路況良好時，利用資訊發布設備進行交通策略宣導可發揮較大效益。本研究在將來可提供其他觀光地區交通攔截圈之應用，或作為其他都會區交通策略之設計。The Old Street in the Taoyuan Daxi Township is one of the top ten most attractive recreational spots in Taiwan. Therefore, it experiences severe traffic and parking congestion during holidays. The Transportation Bureau of Taoyuan County Government has implemented a series of Intelligent Transportation System (ITS)-related strategies to allievate this issue, including but not limited to realtime traffic flow measurement using vehicle dectors, live traffic information broadcasting using Changeable Message Signs (CMS), etc. In addition, in order to address parking lot shortages while accommodating growing numbers of visitors, a shuttle bus service has been provided. This study evaluated and quantified these transportation management strategies at the Old Street resulting from park-and-ride shuttle service operations and ITS-related data. Several data mining algorithms such as classification and association rules discovery were utilized to help predict when and where the congestion will occur, as well as how the shuttle ridership is affected by use of ITS. Other transportation bureaus can apply the proposed research procedure to analyze of their ITS-related data in order to define appropriate traffic management strategies for tourism places.
