資料探勘方法於圖書資訊學領域之運用

Adoption of Data Mining Methods in the Discipline of Library and Information Science

本文探索2009至2018年，圖書資訊學領域研究運用資料探勘方法的趨勢。本研究自Scopus資料庫分別蒐集資料探勘領域和圖書資訊學領域之書目紀錄，並根據基於規則（rule-based）的文字分析法，建構資料探勘方法術語字典;藉由此字典，調查近期圖書資訊學研究中常見之各種資料探勘方法。研究結果發現，圖書資訊學領域運用多元資料探勘法，如大數據、機器學習、文字探勘、資訊檢索以及降維（dimension reduction）等；同時，本研究發現近期流行之機器學習技法（machine learning techniques）的確也被運用於圖書資訊學研究。

The purpose of this paper is to explore the recent trends of data mining method adoption in the library and information science (LIS) discipline. Bibliographic records from the data mining and LIS fields were collected respectively from the Scopus database. A dictionary of data mining method terms was constructed based on a rule-based textual analysis. Using the dictionary, this study investigated a range of prevalent data mining methods utilized in recent LIS studies. The findings of this study reveal different areas of data mining methods employed in LIS, such as big data, machine learning, text mining, information retrieval, and dimension reduction. The study also confirms the recent popularity of machine learning techniques in LIS research.