虛擬參考資訊服務革新—對話服務平台之比較研究

Abstract

參考服務一直是圖書館提供的服務中相當重要的一環，而虛擬參考服務強調的是以即時回應讀者正確的解答為主要目的。隨著通訊軟體的興起，大數據的盛行，相關服務平台應用也應運而生，聊天機器人便是一例。如何將之導入圖書館中，又如何在不同的平台中找到兼顧館方以及讀者的需求，在有限功能中明確的解決讀者問題。 本研究利用台北市立圖書館常見問題集，在line和messenger兩種不同通訊軟體中，利用chatfuel、botbonnie和自行建置等三種不同的服務平台，分別製作出兩種不同操作模式的聊天機器人，一為檢索式，另一為選單式。並透過實際製作了解其服務平台差異性，並將建置好的聊天機器人請位於台北地區大一新生實際操作使用並評估其滿意度。 使用者評估的標準包含資訊品質、系統品質、服務品質與系統比較，藉此了解到使用者的滿意度和對聊天機器人的價值認定，並透過以上幾點來了解使用者的使用意圖。Referencing has always been an important part of the library services, and virtual referencing service emphasises on the immediate and correct response to the readers. As the communication software and big data have been developing prosperously nowadays, relevant applications ,such as chatbot, bocome common. When applying chatbot onto library services, it is important to meet both the readers' and the librarians' needs by providing various resources and solve readers' problem efficiently within limited functions.

This study adopts a innovatively programmed chatbot and two programmed by chatfuel and botbonnie. These chatbots are categotised by search type and pulldown-menu type. This study applies the chatbots onto two communication softwares --- Line and Messenger, and bases its data on the FAQs of the Taipei Public Library. This study implements a user satisfaction survey regarding these chatbots among the first-year university students in Taipei. This study assesses the chatbots with the qualities of informations, systems, and services. In addition, this study does a comparison among the three chatbots. Through these assessments and the comparison, this study analyses the user satisfaction and the user's experiences with the chatbots. Also, this study analyses the user's expectation toward chatbots in library services.