圖書與資訊學刊論文的高頻詞語抽取和分析

Abstract

本論文中嘗試利用自動詞語抽取的技術，對政治大學出版圖書與資訊學期刊的論文進行高頻詞與抽取與分析。在這個問題中，需要克服中文的自然語言處理問題，我們嘗試利用統計方法配合語言學知識來發展詞語抽取技術。利用這個技術，我們對近六年圖書與資訊學期刊的164篇論文的題名與摘要抽取出了233種詞語，在本論文中並對於這些詞語做了一些探討。本研究發現圖書與資訊學期刊論文的詞語以圖書館學的相關概念最多，其次資訊使用研究、資訊科技、圖書資訊學教育、圖書館利用教育和檔案學的相關詞語也相當多。

This paper employs some automatic diction extraction techniques to locate and analyze some key and high-frequency terms from articles published in the Bulletin of Library and Information Science of National Chengchi University. Faced with some difficulties in natural language processing of the Chinese language, we have developed techniques based on statistical information and linguistic heuristics. Two hundred and thirty-three types of terms are extracted from the titles and the abstracts of 164 pieces of articles published last six years. From the concepts carried in the extracted terms, we discover important subject topics in the journal. They include library science, information use studies, information technologies, education for library and information science, library user education, and archives.