基於LISA模型之圖書推薦系統探討使用者滿意度之研究

Abstract

在現今數位化的時代中，圖書館經營的目的亦是要能夠滿足讀者的需求，而圖書館個人化服務已經成為是近年來重要的研究課題之一。本論文研究方向是以圖書館推薦系統為核心，探討其系統效能與使用者滿意度評估之研究。以讀者借閱資料為訓練來源，藉由隱性主題發掘技術及社會網路分析(SNA)的過程，以發掘讀者與讀者間借閱興趣的相似度，訂定讀者借閱資料的關聯性之權重高低，藉以得知讀者之最適性書籍推薦清單。 此外，透過系統效能評估結果與讀者滿意度問卷之調查，探討兩者評估結果中間所產生的落差，並找出如何主動發掘讀者的需求，以及提供讀者所需要的資訊。透過此項研究分析，來探討讀者使用圖書館之行為，不僅可提供圖書館經營管理者在決策館藏發展政策、圖書推薦，亦可以提供圖書館界在個人化的主題領域中更加廣泛、實用的服務效能。

In today's digital era, the library is also the purpose of business to be able to meet the needs of readers, and the library is the personal service has become an important research topic in recent years. Libraries in this research is based on recommendation system as the core of the system performance and user satisfaction assessment research. Readers training data to the source of the theme by hidden technology and social network analysis to explore (SNA) in the process, readers and readers to explore the similarity between the loan interest, information on Readers set the weights of the relevance of the level, readers to know the most adaptive book recommendations list. In addition, through the performance assessment system and reader satisfaction survey questionnaire, the results of two assessments produced by the middle of the gap, and find out how to take the initiative to explore the needs of readers, and provide the information readers need. Through this study, to explore the behavior of readers use the library, not only for library managers in decision-making collection development policies, recommended books, libraries can be provided in individual subject areas of more extensive and practical service performance.