摘要
本文透過鄰近預測（nowcasting）方法，檢視會議展覽（MICE）對台灣民間消費（service consumption）的影響。運用搜尋趨勢資料庫（Google trends）的大量即時資料，以高頻率的會議展覽資訊鄰近預測出低頻率的台灣民間消費。目的在以豐富的即期資訊提高預測正確性，也欲了解會議展覽對台灣民間消費之影響。本文將搜尋趨勢資料庫的會議展覽關鍵字，架構出會議展覽檢索指標，並強調檢索指標與台灣民間消費之相互影響，欲瞭解何種會議展覽關鍵字與台灣民間消費有最密切相關。實證結果發現14大分類中較常被搜尋的4個關鍵字，分別為incentive（獎勵旅遊）、exhibition hall（展覽館）、conference center（會議中心）、city marketing（城市行銷）。22小分類中較常被搜尋的4個關鍵字，分別為auto show（汽車展）、Taipei exhibition （台北展覽館）、KEC（高雄展覽館）、lantern festival（燈會）。本文結果穩健地指出台灣民間消費顯著受到會議展覽關鍵字的影響，其也顯示台灣會議展覽主要是靠縣市政府推動及民間企業規劃進行之事實。本文價值在於以數量方法成功證明了”會議展覽可帶動民間消費”的直覺想法，我們運用鄰近預測法不僅得到符合直覺想法之結果，而且本文實證結果亦符合既存理論。本文有二大貢獻，第一，目前未有運用會議展覽資訊進行鄰近預測台灣民間消費的相關文獻，本文彌補此類文獻之匱乏。第二，近20年來，台灣政策當局力推會議展覽產業的發展，本文結果歸結出會議展覽產業確實與民間消費具顯著密切相關，也呼應貿易衍生的會展產業確實帶動擴張台灣民間消費與提升台灣經濟之既存事實。
關鍵詞：鄰近預測、台灣民間消費、會議展覽 Abstract
This paper examines service consumption in Taiwan using MICE (Meeting, Incentive, Convention, and Exhibition) data via nowcasting. The aim is to improve the accuracy of service consumption forecasts in Taiwan by utilizing real-time information and to identify the impact of MICE on service consumption. We constructed diffusion indices using MICE keywords obtained from a Google Trends database. High-frequency real-time MICE data was used to nowcast the low-frequency service consumption data. This paper focuses on the interrelations between components to determine which has a greater impact on service consumption. The results show that the keywords "Incentive, Exhibition Hall, Conference Center, and City Marketing" groups, and the keywords "Auto Show, Taipei Exhibition, KEC, and Lantern Festival" classifications, all have a significant impact on service consumption. Our contributions are twofold. First, our empirical results correspond with the fact that "The derivatives of international trade, MICE industries, have expanded service consumption and boosted Taiwan's economy in recent years." Second, this paper fills a gap in the literature on the relationship between MICE and Service Consumption.
Keywords: Nowcasting, Service Consumption, MICE