自動議題分類機制應用於愛滋病提問議題-以Yahoo知識家為例

Theme Classification for HIV/AIDS Medical Questions-A Case Study of Yahoo! Answers

陳信佑(Hsin-Yu Chen)

指導教授 ： 邱昭彰

元智大學/資訊學院/資訊管理學系/碩士(2011年)

https://doi.org/10.6838/YZU.2011.00029

摘要

近幾年在健康醫療服務(Health Care Service)之研究趨勢上，對於醫療諮詢者的需求，導向於建構實體和虛擬管道進行醫療諮詢之環境。隨著網際網路的發展，網路平台已成為尋求醫療諮詢與討論之熱門虛擬管道。從社會觀感來看，被貼上負面標籤的愛滋病議題上，提問者傾向選擇透過網路管道而逃開使用實體管道尋求相關資訊，其中尤其以尋求HIV/AIDS相關醫療諮詢更加明顯。
　　因此本研究針對愛滋病議題，開發Crawler程式以收集存在「Yahoo!奇摩知識家」之愛滋病提問文章，透過文件探勘技術(Text Mining)、人工智慧(Artificial Intelligence)等技術，提出一套愛滋病相關文章自動化議題分類機制，以有系統的彙整各項愛滋討論議題建立愛滋病提問之醫療回覆推薦系統資料庫，以提供相關單位及人員後續應用。
　　實驗結果發現同時考量平均字詞權重與出現位置區塊，予以加權計算，以及針對出現位置區塊進行適切的文章區塊劃分，皆能有效提升分類效果。本研究提出的權重處理方法之分類F-measure為76.32%。

關鍵字

文件探勘 ； 推薦系統 ； 議題分類 ； 基因演算法

並列摘要

Studies on health care service have been conducted to provide the medical consultation environments to the people who seek the medical suggestions in the real or the virtual way. With the fast development of the Internet, network platforms have become the popular place supplying with medical asking and discussion. Moreover, askers will choose virtual way instead of the real way to quest about the stigmatized HIV/AIDS issue.
If people take the unreliable and unverified medical advice from the Internet would cause some problems. For solving this issue, we collect the HIV/AIDS questions from “Yahoo! Answers” and use Text Mining and Artificial Intelligence to conduct a HIV/AIDS recommendation system. Askers can use the text form to describe their problems or the events they suffered, and the system will timely respond the dependably medical recommendation.
In the experiment, we found adjustively dividing article can improve the effectiveness of the classification. In addition, considering statistical and linguistic keywords weight can also enhance the accuracy. The F-measure of the best　 classification result shows 76.34%.

並列關鍵字

Text Mining ； Recommendation System ； Theme Classification ； Genetic Algorithm

參考文獻

21.Jong-Hun Kim, Jung-Hyun Lee, Jee-Song Park, Young-Ho Lee and Kee-Wook Rim,“Design of Diet Recommendation System for Healthcare Service Based on User Information,” 4th International Conference on Computer Sciences and Convergence Information Technology, No. 5367898, pp.516-518, 2009.

1.Bull, S. S., Mcfarlane, M., & King, D., “Barriers to STDs/HIV prevention on the Internet,” Health Education Research, Vol. 16, No. 6, pp. 661-671, 2001

2.Besnehard, Q., Marchessoux, C., Kimpe, T., “Generic and Optimized Framework for Multi-Content Analysis Based on Learning Approaches,” Proceedings of SPIE - The International Society for Optical Engineering 7540, Art. No. 75400V, 2010.

3.Cover, T., and Thomas, J.A., “Elements of Information Theory,” Wiley, 1991.

4.Creecy, H., Masand, M., Smith, J.,and Waltz, D., “Trading Mips and Memory for Knowledge Engineering,” Communications of the ACM, Vol. 35, No. 8, pp. 48-63, 1992.