基於個人健康紀錄之線上健康保健知識建議系統

PHR-Based On-Line Health Care Knowledge Suggestion System

林俊成(Lin Chun-Cheng)

指導教授 ： 林志豪

亞洲大學/資訊學院/資訊多媒體應用學系碩士班/碩士(2010年)

摘要

隨著國民知識水準提升，大眾對於醫療品質也逐漸重視。因此，能夠主動管理個人健康紀錄的意識抬頭，以了解個人健康狀況，進而更投入於個人的醫療計畫。因此，關於個人健康紀錄的研究也逐漸受到重視。此外，隨著網際網路的發達，線上教學的應用也日益廣泛，許多不同領域的線上問答系統也陸續開發出來，社會大眾的詢問行為也逐漸轉移至網路上；因為語言是人類最自然的溝通方式，要讓電腦理解人類的語言實為複雜的工程，在問答系統的表現上，自然語言處理能力即是影響問答系統在回覆準確度的主因；此外，影響問答系統表現的因素還有回覆的時效性。
本研究利用本體論概念、自然語言處理的技術以及題庫的方式，並搭配個人健康紀錄系統，建構出以健康保健知識為資料來源的問答系統；系統利用Autotag斷詞系統針對問句進行斷詞，再利用語意樣板進行語意分析，最後透過系統的知識本體與個人健康紀錄進行適當的知識搜尋回覆，目的在於達到適性化的知識建議，並且即時的回覆建議相關知識。由實驗結果看出，本系統對於使用者的詢問可以有一定程度的理解，並且做到適性化的目標。預期此研究架構的設計，在研究上，能夠幫助自然語言的相關研究；在生活上，能夠讓使用者對於自身的健康照護及醫療計劃更加的重視與投入。

關鍵字

個人健康紀錄 ； 問答系統 ； 自然語言處理 ； 本體論 ； 適性化

並列摘要

Because people rises attention to medical quality day by day, the sense is risen that patients manage their personal health record (PHR) themselves, and the research in PHR is respected, because more people be participate in the medical plan. In addition, along with internet’s popularization, more on-line learning services are provided. A lot of on-line QA systems in different domains are developed, so it makes people’s queries via the internet more. Because language is the most naturally using to people’s intercommunication, it’s really complicated engineering that how to let computers understanding the language. The main factors in QA system’s accuracy not only the Natural Language Processing (NLP), but also the response effective.

We collate ontology techniques, NLP techniques, question bank and PHR system to build a healthcare knowledge suggestion system. The system employs Autotag system to intercept user’s asking sentences, semantic templates analyze lexical, and do search and reply the fitness knowledge for user via the knowledge ontology and PHR system at the final. The purpose lies in fitness knowledge suggestions and reply prompt. By the experimental calculation, our system can accomplish get some degree to understand user’s asking, and achieve the goals that reply fitness suggestions. In the research, expect this study can be the advancement of Chinese Natural Language Processing. In life, expect can allow users respect the personal health statue themselves and be more participate in the individual medical plan.

並列關鍵字

PHR ； QA System ； Natural Language Processing ； Ontology ； Fitness

參考文獻

[6] 莊宗南、龔榮源、陳俊龍，“以資料探勘技術建立病患就醫導引：以胃腸科病患為例”，台灣醫學資訊學會醫療資訊雜誌，第十五卷，第一期，2006/3，17-34頁。

[13] AHIMA e-HIM PHR Work Group, “Defining the Personal Health Record”, Journal of AHIMA/American Health Information Management Association, 76, 24-25, 2005.

[7] B. Hu and J. Bai, “Managing Dynamic Medical Data in a Distributed Mode,” Proceedings of the 20th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Vol. 20, No 3, 1998, pp. 1292-1294

[8] WHO, “Retrieved December 14,” 2006. http://www.who.int/

[10] T. A. Horan, “Prospective Personal Health Record use among different user groups: results of a multi-wave study,” Proceedings of the 41st Hawaii International Conference on System Sciences, 2008.