當資訊需求者欲掌握事件之全盤細節時，其往往藉由不同管道蒐集大量之相關資訊並加以閱讀，以進一步於此些相關資訊中獲取該事件較全面且周延之訊息。然而，現今之資訊需求者往往可於短時間內自網際網路或報章雜誌等多元搜尋管道中獲取大量的相關文件，卻無法於短時間、有限時間內完整掌握所獲取文件之內容脈絡。為解決此問題，本研究乃發展一套「文件內容重要度之判定及視覺化」模式。當蒐集欲了解之事件相關資訊後，資訊需求者可利用行動裝置將所獲取之各相關資訊文件以掃描方式輸入本研究所提出的模式中，以藉由此模式獲取文件各文句之重要度及各文句重要度所對應之重要度標註顏色。最後，資訊需求者即可透過此些文件各文句掃描後之即時影像上所標註之文句重要度顏色快速辨別文件中較具重要性之文句，以有效率地了解文件之內容脈絡。

When the information requester wants to grasp the overall details of the event, it often collects amount of relevant information from different place such as newspaper, magazing…etc. and than read it all. However, today's information requester can often obtain a large number of related documents from different place in a short period of time, but cannot read it all within a limited time. In order to solve this problem, this study develops a set of "A Model for Determination and Visualization the Importance of Document Content" and a corresponding system. After collecting relevant information which want to known, the information requester can use the mobile device to scan all the relevant information files into the mode. After model obtain the importance of each sentence of the document and get the marked with a different color, information requester can quickly identify the important part of the document by the different mark color which will be show on the real-time image scanned video. In this way, information requester can understand the context of the file more efficiently.

摘要...I
ABSTRACT...II
目錄...III
圖目錄...IV
表目錄...VII
第一章 、緒論...9
1.1研究動機與目的...9
1.2研究流程...11
1.3研究定位...14
第二章 、文獻回顧...16
2.1文件特徵擷取...16
2.2文件內容重要度解析...21
第三章 、文件內容重要度之判定及視覺化模式...29
3.1影響文句重要度因子辨析...30
3.2文句之重要度衡量模型...35
3.3文句重要度評比階段...46
第四章 、績效驗證與分析...64
4.1績效驗證規劃...64
4.2驗證結果分析...68
第五章、結論與未來展望...96
5.1 論文總結...96
5.2未來展望...98
參考文獻...100
附錄A、模式驗證資料...102
附錄B、模式於第二階段績效驗證結果...222

1.Chankook, P. and Taesaok, Y., 2017, “Prospect of Korean nuclear policy change through text mining,” International Science Confrence, pp. 72-78.
2.Chen, Y.-J., 2000, “Scalable Summarization for Chinese Text,” Department of Computer Science, National Tsing Hua University, Hsinchu, Taiwan, pp. 13-51.
3.Gupta, P. and Shankar, V., 2011, “Summarizing Text by Ranking Text Units According to Shallow Linguistic Features,” 13th International Conference on Advanced Communication Technology, pp. 13-16.
4.Zhang, L.-K. and Wang, H.-F., 2012, “Research on Sentence Extraction in Text Summarization,” Journal Of Chinese Information Processing, Vol. 26, pp. 97-102.
5.Ferreira, R., Cabrala, L.-D.-S., Freitas, F., Lins, R.-D., 2014, “A multi-document summarization system based on statistics and linguistic treatment,” Expert Systems with Applications, Vol. 41, pp. 5780-5787.
6.Cao, Z., Wei, F., Dong, L., Li, S. and Zhou, M., 2015, “Ranking with Recursive Neural Networks and Its Application to Multi-Document Summarization,” Twenty-ninth AAAI conference on artificial intelligence, pp. 2153-2159.
7.Chen, Z.-M., 2010, “Intelligent location-based mobile news service system with automatic news summarization,” Expert Systems with Applications, pp. 6651-6662.
8.Li, J., Fan, Q., Zhan, K., 2007, “Keyword Extraction Based on tf/idf for Chinese News Document,” Wuhan University Journal of Natural Sciences, Vol. 12, pp. 917-921.
9.Kavila, S.-D., Rajesh, B., Vyshnavi, N., Sushma, K.-M., “Automatic Key Term Extraction from Research Article using Hybrid Approach,” International Journal of Computer Applications, Vol. 166, pp. 17-21
10.Han, Y.-F., Xu, X.-Y., Li, B.-C., Zhu, W.-B., Chen, G., 2012, “Web News Multi-document Summarization Based on Event Extraction,” Journal Of Chinese Information Processing, Vol. 26, pp. 58-67.
11.Luhn, H. P., 1958, “The Automatic Creation of Literature Abstracts,” IBM Journal of Research Development, Vol. 2, pp. 159-165.
12.Baxemdal, P. B., 1958, “Machine- Made Index for Technical Literature-An Experiment,” IBM Journal of Research Development, Vol. 2, pp. 354-361.
13.Chen, H.-J. and Lin, C.-J., 2000, “A multilingual news summarizer,” COLING '00 Proceedings of the 18th conference on Computational linguistics, Vol. 1, pp. 159-165.