文件內容擴增實境之視覺化模式

Visualization of Augmented Reality for Contextual Content

廖婕安(Chieh-An Liao)

指導教授 ： 侯建良

國立清華大學/工學院/工業工程與工程管理學系/碩士(2014年)

https://doi.org/10.6843/NTHU.2014.00284

摘要

當資訊需求者透過網際網路閱讀一份資料時，其往往因無法理解該資料之特定字句、或對於該資料之特定內容感到興趣而進一步搜尋該特定字句或內容之相關資料。然而，此些相關資料往往僅為特定說明網頁而非多元之參考資料，故此類特定說明網頁乃可能無法滿足資訊需求者之需求，而致使資訊需求者額外花費時間搜尋其欲釐清之特定字句或內容的相關資料。
為改善上述過程之效率及效能，本研究乃提出一套「文件內容之擴增實境與其視覺化模式」。此模式乃包含「現行補充資料內容解析」前置作業及「補充資料內容分析及特性呈現」方法論，而此方法論又包含「補充資料內容解析」、「補充資料面向解析」、「補充資料基本特性評估」、「補充資料特質視覺化」等四大階段。具體而言，此方法論乃先針對資訊需求者所欲閱讀之目標文件及對應之相關資料進行內容解析，並根據解析結果釐清目標文件所對應之補充資料；之後，此方法論乃以詞彙關聯性及向量空間模型等概念為基礎將此些補充資料予以分群，以推論此些補充資料所對應之主題、面向，再以類神經網路方法評估各補充資料內容之多項特性；最後，此方法論乃將各補充資料之面向及特性等資訊以視覺化方式呈現。
其次，本研究乃根據此模式開發對應之系統平台，以針對資訊需求者所欲閱讀之目標文件建構對應之擴增實境，進而呈現目標文件之補充資料及對應之關鍵資訊。之後，本研究乃以品質主題相關之資料測試所提出之模式及所開發之系統的績效表現，並得知此模式及系統能提升資訊需求者閱讀目標文件之效率，亦能使其更有效地瞭解目標文件之所有細項內容。

關鍵字

擴增實境 ； 詞彙關聯性 ； 向量空間模型 ； 類神經網路

並列摘要

As a person searches and reads documents of specific topic over the Internet, he/she might not understand or might be interested in some particular, confusing terms of the documents and might try to access some other references in order to clarify these confusing terms. However, there might be only few references linked to the documents via hyperlinks and these references usually provide limited ideas or concepts, which might cause the person to spend time on searching and filtering helpful references.
In order to improve the efficiency and effectiveness for studying and comprehending the contextual content, this research develops a model for construction and visualization of augmented reality for the contextual content. Firstly, this research analyzes a great number of target documents and identifies some reference documents corresponding to the target ones. Secondly, a model is developed for analyzing these reference documents and transforming the documents into structured ones. After that, the proposed model can be applied for analyzing the topics and evaluating the characteristics of each structured document by using the vector space model and the neural network method. Finally, the proposed model can visually display the topics and characteristics of reference documents with respect to the target one to the reader.
Based on the proposed model, this research develops a corresponding system to virtually construct the augmented reality for the target documents. Afterwards, this research designs some experiments to check the performance of the proposed model and constructed system. Consequently, the result of the experiments shows that the proposed model and system can effectively assist readers to understand the target document deeply and quickly.

並列關鍵字

augmented reality ； keyword correlation ； vector space model ； neural network

參考文獻

1. 陳妍妤，詹智凱，劉立，徐建業，2009，「以詞的關聯性為基礎之文件自動分類系統應用於網路癌症衛教資訊分類」，中華民國癌症醫學會雜誌，第25卷，第5期，頁350-356。

4. Miao, D., Duan, Q., Zhang, H. and Jiao, N., 2009, "Rough set based hybrid algorithm for text classification," Expert Systems with Applications, Vol. 36, No. 5, pp. 9168-9174.

5. Özel, S. A., 2011, "A web page classification system based on a genetic algorithm using tagged-terms as features," Expert Systems with Applications, Vol. 38, No. 4, pp. 3407-3415.

6. Li, C. H. and Park, S. C., 2009, "An efficient document classification model using an improved back propagation neural network and singular value decomposition," Expert Systems with Applications, Vol. 36, No. 2, pp. 3208-3215.

7. Saracoglu, R., Tutuncu, K. and Allahverdi, N., 2008, "A new approach on search for similar documents with multiple categories using fuzzy clustering," Expert Systems with Applications, Vol. 34, No. 4, pp. 2545-2554.