Master of Science / Department of Computing and Information Sciences / Daniel A. Andresen / In this project I have built a web application "Recipe Search Engine Using Yummly API". This application is central information hub for the kitchen--connecting consumers with recipe ideas, ingredient lists, and cooking instructions. It will serve best for the people who uses digital tools to plan their cooking, these days almost everyone does. The various features available for users in this application are as following. Users can search for their favorite dishes. The search results contain information about ingredients list, total time needed for cooking, user's rating and cooking directions. Basic search filters are provided to filter out the search results like Breakfast, Lunch and Dinner recipes. The order of displayed results can be sorted according to ratings, total time required to prepare the dish. User can create an account and build their own favorite recipe collection by liking the recipes displayed. The liked recipes are stored into user’s account and user can view, add and delete those recipes anytime from his recipe collection. Users can use their social networking platform Facebook account credentials to log into this application or create a new account in this application. The application will communicate with the Yummly API to consume data from it. The Yummly API is largest recipe information aggregator with over one million recipes data.

「民以食為天」，飲食與我們的生活息息相關。近年來由於食安風暴肆虐，自行烹煮的需求隨之高漲。然而在家自行烹煮時常會面臨不知道該烹煮什麼料理的問題，因此有便利的食譜搜尋系統對烹煮的人而言將是相當方便的。然而使用搜尋系統時，由於我們只知道想用某些特定食材進行烹煮，而不知道哪些食譜含有特定食材，因此在以少數食材進行查詢時不免會得到過多的食譜結果而難以快速找到喜好的食譜。我們建立了一個食譜搜尋的自動完成機制，並依照該機制實做出了食譜搜尋引擎。使用者使用系統進行搜尋時，我們將會依照使用者輸入的食材尋找適合搭配的食材推薦給使用者，幫助使用者在查詢時使用更完整的Query讓搜尋系統可以找到更少更精準的食譜，幫助使用者更快的找到喜歡的食譜。然而只推薦搭配性食材，可能會推薦出與Query中的食材是替代關係的食材，也就是通常不會一起出現的食材，因此我們也進行了替代性食材的研究。給定由兩個食材組成的食材配對，我們研究如何自動的判斷替代性食材。我們將問題轉化成分類問題來解決，並使用One-Class Classification的技術解決分類問題中的Imbalanced Problem。我們使用f1-score觀看One-Class Classification與傳統分類器的比較。經實驗測試，One Class Classification與傳統分類器相比，One Class Classification較能協助我們解決Imbalanced Problem。