通常在考慮決定一項事物時,會受到許多層面、因素的影響,所以推薦系統則是一項能夠在複雜的資訊中,幫助使用者做出分析需求、推薦的系統。而我們研究是將焦點放在醫療輔助上,依國內行政院衛生署的統計,2009年的主要死因死亡率,糖尿病名列十大死因之一,因此對於糖尿病發展一個有效的治療是非常重要的。由於糖尿病患者眾多,避免專科醫師不足或是患者無法給予專科醫師看診的情況下,希望藉由此推薦系統來幫助家庭醫師在選擇藥物上做出合適的輔助。此研究的實驗希望藉由患者知識本體的建構建立患者項目,接著再利用多準則決策的方法,其重要性的熵量度運算將其資料的用藥程度做出計算,接著與藥品知識本體做出結合排序並推薦出最適用的藥品類型,最後呈現醫師較適用的藥類推薦。
Generally speaking, we have to consider the effect of many factors when making a decision. The recommended system can help user to analyze needs and give suggestions with complex information. Our research focuses on the assistance of medical suggestion for doctors. According to the statistics of Department of Health, diabetes was one of the top ten causes of death in Taiwan in 2009. Therefore, developing an effective treatment for diabetes is very important. The fact is that the amount of diabetes specialists is not enough and not all diabetic patients can be treated by specialists. We present this recommender system to help treatment. The purpose is to develop a recommender system to assist the general practitioner (GP) to make more appropriate decision in selecting anti-diabetic medicines. First, we built up the ontology of diabetic knowledge, and then multiples criteria decision making method (MCDM) was applied to compute for medication. Entropy was used to compute data of patients’ profile. With medicine knowledge ontology, the results of calculation will list appropriate medications. Finally, the system will show the preference of medications to doctors.