疫情前後國人國內旅遊狀況初探研究──以電信數據探勘技術為例

Insights into Domestic Tourist Profiles through the Analysis of Cellular Vehicle Probe Data–Using Profiles of Activity Before and After the Onset of the Pandemic

為瞭解國人國內旅遊樣態、觀光效益等重要指標，國人旅遊狀況常態性調查有其必要性，但傳統以電訪、問卷調查之耗時、高成本、無法瞭解在國內旅遊區域分布，即時取得國旅現況等限制，往往造成調查結果有所限制，為彌補傳統調查不足之處，本研究以電信數據為基礎，加上定位技術演進及設備提升，可取得大量去識別化位置資訊，將定位資料應用於洞察旅客旅遊狀況，並建立多種指標進行評估，且根據大數法則(law of large numbers)可證明大樣本對母體參數可以得到良好的估計結果。本研究為國內首度以電信數據進行全臺國人國內旅遊狀況的分析，並且區分Covid-19疫情爆發期及疫情穩定期進行旅客旅遊狀況差異化分析，驗證觀光策略推動成效，此方法可提供政府檢視策略效益參考。

To estimate the number of domestic trips and the benefits of tourism, regular surveys of travel by Republic of China (Taiwan) citizens are important tools. However, conventional approaches, such as phone surveys and paper questionnaires, are time-consuming and costly. Moreover, such surveys at times have insufficient sample sizes, which can reduce the accuracy of estimations. To avoid such shortcomings, this study utilized “cellular vehicle probe” (CVP) data. The CVP data approach evolved from the development of new positioning technologies and equipment. This approach collects large amounts of anonymous signaling data, which can then be analyzed to model tourist patterns. We developed several indicators that can describe tourist behavior, and these indicators have a strong ability to estimate behavior patterns using the “law of large numbers.” This study is the first domestic travel survey conducted using mobile phone data. As part of the study, we conducted a differentiated analysis between two different phases of the Covid-19 pandemic. This provided insight into how different strategies impacted the tourism environment when the outbreak conditions were severe and when concerns were at a lower level. Our study demonstrated that the CVP approach can provide useful feedback, such as assessing the impact of government tourism strategies on tourism conditions.