本研究將模糊分類與模糊詮釋資料之概念應用於資訊選粹服務系統的設計上，以更滿足讀者的個人化資訊需求。我們首先對於若干與本研究相關的領域進行探討，包括圖書館自動化、分類編目、資訊選粹服務、詮釋資料、模糊理論等。接著，闡述模糊詮釋資料，並以模糊理論模型解釋其意涵。再者，提出圖書模糊分類與編目，以及將模糊詮釋資料以中國機讀編目格式註錄至書目資料庫之方法。另一方面，本研究亦對讀者進行模糊分類，以便讀者與圖書之模糊關係可藉由模糊合成運算獲得，達成模糊資訊選粹服務之目標。最後，我們發展一套實驗型的模糊資訊選粹服務系統來印證本研究所提出的方法。

This study applies the concepts of fuzzy classification and fuzzy metadata to the design of the library system of selective dissemination of information, so as to further correspond to the reader’s individual
requirement of information. Firstly, we survey some fields related to our study, including library automation,classification and catalogue, selective dissemination of information (SDI), metadata, and fuzzy theory. Secondly,we attempt to define and expound the concept of fuzzy metadata by using fuzzy theory as the theoretical model.
Furthermore, we propose a fuzzy method for classifying and cataloguing books as well as a practical approach for recording the fuzzy metadata to bibliographical databases in CMARC format. The method of fuzzy
classification is also applied to the reader, in order that the fuzzy relations between the reader and books can be explicated in fuzzy composition to achieve the goal of fuzzy SDI. Finally, we develop an experimental fuzzy SDI system to verify the proposed methods in our study.