(第 14 期)   第七卷第二期   2013 年 6 月 1 日出刊

使用情緒分析於圖書館使用者滿意度評估之研究

A Study on Library Users’ Satisfaction Evaluation Using Sentimental Analysis

本文關鍵字:圖書館使用者、滿意度評估、情緒詞、情緒分析、Library users、Satisfaction evaluation、Sentimental words、Sentimental analysis

本文摘要

本研究採用情緒分析探究社群網路上(如噗浪等)之使用者對於圖書館的意見屬於正向或負向,並以圖形方式顯示評估結果,可及時提供圖書館業務調整與改善使用者滿意度。蒐集網路上圖書館的使用者意見進行人工標記,建立相關情緒分析辭典後,使用KNN、NB 和SVM 等分群工具探討圖書相關語料分類的效果外,提出導入程度詞、否定詞與情緒詞權重、及情緒極性與情緒類別等兩種情緒分析方法。研究結果顯示KNN 具有較佳的分類效果以及以情緒極性與情緒類別得到較佳的圖書館使用者滿意度之評估結果。進一步,也證實所有建置之情緒分析辭典的可靠性。 This paper employs the sentimental analysis to study the library users’ satisfactory evaluation, where users’ opinions were classified as positive, neutral and negative, respectively. To provide the immediate library service improvement, the graphic representation is further used to show evaluation results. After collecting the library users’ opinions from micro webs, i.e., Plurk, to be the corpus, the corpus was then pre-processed and annotated manually. Moreover, all the related dictionaries for sentimental analysis were constructed. accuracy each category, were employed to cluster the corpus into five categories. Furthermore, two new sentimental analyses for library satisfactory evaluation were proposed, which include introducing degree words, negatives and sentimental words weights, and introducing sentimental polarities and categories, respectively. Experimental results showed that both KNN and sentimental analysis introducing sentimental polarities and categories obtain better satisfactory evaluation results. Furthermore, all the constructed dictionaries for sentimental analysis were shown their feasibility as well.
本文附件:

本期同分類其他文章


本刊著作權屬於「中華民國圖書館學會」所有。
Powered By Vanilla Journal - 香草期刊系統 0.256 / 2006 - 2007 © Weizhong Yang