國立臺灣師範大學圖書資訊學研究所卜小蝶陳思穎 2015-07-032015-07-032007-12-011990-9128http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/73070本文由使用者角度，分別採用實驗、觀察、問卷、訪談及檢索過程記錄分析等方法，就檢索效率、檢索效益及滿意度等範疇，評估比較分群與排序條列式搜尋引擎之檢索表現。整體而言，排序條列式搜尋引擎能在較快的時間內、找到品質相當且數量較多的相關網頁，使用者對檢索結果的滿意度也較高。而分群搜尋引擎則具有突顯重要概念、提供多元思考及降低資訊超載等特性。此外，使用者對於個別群集的接受度較整體群集架構來得高。分群搜尋引擎提供了嶄新的檢索經驗，但仍有不少問題需要進一步分析探討，本文最後則提供一些觀察與建議。This study provides user evaluation of clustering and ranked list search engines by comparing their search efficiency, search effectiveness, and users' satisfaction level. The methods include experiment, observation, questionnaire, interview, and search log analysis. The results show that the ranked list search engine performs better in users' search speed, relevant pages retrieved, and the satisfaction level. On the other hand, the clustering search engine has the merits of demonstrating important concepts, providing clues for diverse thinking, and helping to reduce the information overload. Meanwhile, users' acceptance of individual clusters is higher than that of the cluster scheme. The paper also provides some discussions and suggestions on improving the clustering search engines.網路分群搜尋引擎網路使用者研究搜尋引擎評估Web clustering search engineWeb user studiesSearch engine evaluation