透過您的圖書館登入
IP:3.134.81.206
  • 期刊
  • OpenAccess

以模糊推論建構營養膳食諮詢系統中配菜功能

Functions for Choosing Menu Using Fuzzy Inference in Diet-Consultation System

摘要


現有膳食諮詢系統,大都著重於營養條件的符合,往往忽略了用餐者的味覺需求。由於口味的喜好因人而異,符合營養條件的電腦菜單,往往很難滿足每個人的口感。基於模糊推論接近人類思考模式,若依此建立擁有口味選擇能力的專家系統,則可使諮詢系統的使用更加人性化。本研究針對食物的油膩感,由菜單的油脂量與含脂食材比兩個變數值,經模糊推論得到五個等級。此推論與100位大專同學對相同菜單的油膩感問卷結果比對,認知相似程度達百分之八十五。此外,因選擇了需要的口感後配菜組合數目相對減少,每次查詢時間可由原先的平均56.47分鐘降為2.31分鐘,即節省了96%運算時間。歸納以上結果,口味選擇功能的加入,不但仍滿足了原有營養諮詢的功能,更可以有效篩選出合乎諮詢者口味要求的菜單組合並減少使用時的等待時間。因此,新的諮詢系統將更符合個別使用者的需求。在未來若能加以推廣使用,對國人的飲食均衡習慣將有非常大的助益。

並列摘要


Most diet-consultation systems were focused on satisfying nutrients status, not the delicious taste. As we knew, it was difficult to make definition for ”palatableness” because everyone was different. The food satisfied the nutrition request was not always palatable for everyone. The fuzzy inference which liked the thinking in the human brain could be easy to solve obscure problem. Base on this opinion, if a ”taste option” added to the diet-consultation system by a fuzzy expert system, it would make system more humanization. For this objective, a reasoning system based on fuzzy logic to choose menu was built. The level of amount of fat and weight percentage of containing oil material described five linguistic variables were used to reason oily level of food and separated the oily food to five levels. According to questionnaire of oily taste in same menu, it was found up to 85% similarity to result from computer calculation. Moreover, the average time of calculation reduced from 56.47 minutes to 2.31 minutes saving above the 96% calculating time due to the composition of dishes in menu was decreased after taste option applied in our system. Result show that with the taste option without changing nutritional status, the consulting system not only chose the menus for consumers' need but also reduced the searching time. Therefore, new consulting system could fit personal need much better. If the system was widely implemented in the future, balance of our eating manner would be more helpful than before.

被引用紀錄


林煥智(2007)。以知識本體為基礎的個人化營養配方諮詢系統〔碩士論文,中原大學〕。華藝線上圖書館。https://doi.org/10.6840/cycu200700679

延伸閱讀