護理經驗的知識擷取-以護理個案報告為例
Knowledge Acquisition from Nursing Experience-an
example of the nursing case report
中文摘要
本研究目的是希望從個案報告中擷取護理照護經驗的知識，給予未曾有過此一照
護經驗的護理人員的一個參考來源。透過視覺化來展現結構化的知識過程，利用
「文字探勘技術」發現照護經驗與知識之間的關聯性。
本研究收集
88 年至 95 年期間以「血液透析相關護理」為主題，經由護理學會
審查通過並公開發表於「護理雜誌」上之個案報告，共
28 篇做為知識擷取實作
的資料來源，結果產生
2681 個中英文詞，其中篩選出現頻率高於 33％的文詞
共
67 個，再經由洗腎室護理人員對「血液透析」相關的概念選出 17 個字詞為
「Ontology」元素，最後依排列組合共形成 136 個兩兩相關的聯結。
由單一因子變異數分析檢定發現：洗腎室護理員所評估元素與元素之間的關聯強
度並不等同於資料探勘的關聯強度，顯示臨床洗腎室的護理人員對「血液透析」
相關的認知概念與資料探勘所呈現的概念並不相同。將資料探勘的
Ontology 概
念結構以得分
37.4 分（機率值為 0.006）區分成高分及低分兩組，結果顯示洗
腎室護理人員與資料探勘的高分組概念一致，但非洗腎室人員與資料探勘的結果
不同，這表示資料探勘的核心知識與洗腎室人員相同。
經由資料探勘方式最後分析出兩個對洗腎病人重要的照護概念；分別為「護理經
驗_接受透析治療」
、
「護理問題_體液容積過量」
。期能藉由本研究之「文字探勘」
方式，呈現知識核心概念架構。
英文摘要
The purpose of this study is to acquire nursing knowledge and practical experience from case-study report which could provide a reference source for those non experienced nurses. The main process is, through demonstrating the structured knowledge visually, and using text mining technology, to discover the relationship between report knowledge and practical experience.
Using “hemodialysis” as topic, and referring to 28 articles which were published in “The Journal of Nursing” from 1999 to 2006 and certificated by Taiwan Nurses Association, we’ve found that from the result of 2681 terms, there’re 67 terms which have 33% higher frequency than all others, and 17 terms were identified from
those 67 terms, by nurses in Hemodialysis-Room, as an Ontology element which relating to hemodialysis concept. Finally, thru arranging those 17 terms, 136 terms were formed which linking to each other.
We used one way analysis of variance statistics, and found that the assessment of intensity relation amid elements from nurses in Hemodialysis-Room were not equal to the one of text mining and either the related concept about “hemodialysis”. We used the text mining’s score and set 37.4 (Probability value was 0.006) to divide into 2 groups, higher and lower. The result indicates that the concept is the same between nurses in Hemodialysis-Rooms and those who have higher score in text mining, and wasn’t if comparing to those non Hemodialysis-Rooms nurses. We conclude that the core knowledge of text mining was same as the one which nurses in Hemodialysis-Room have.
Two important nursing concepts were found during Data mining, for patients who experienced hemodialysis, they are “nursing experience - received dialysis
treatment" and "nursing diagnosis-excessive fluid volume". Basing on the result of this study, we suggest using the text mining method to demonstrate the core concept of the knowledge.
