應用平行關聯演算法於中式速食連鎖餐廳之套餐設計

淡江大學機構典藏 > 商管學院 > 資訊管理學系暨研究所 > 學位論文 > Item 987654321/105544

jsp.display-item.identifier=請使用永久網址來引用或連結此文件: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/105544

题名:	應用平行關聯演算法於中式速食連鎖餐廳之套餐設計
其它题名:	Applying parallel association algorithms to value meal design for a Chinese fast food chain restaurant
作者:	蘇育群;Su, Yu-Chung
贡献者:	淡江大學資訊管理學系碩士班鄭啟斌;Cheng, Chi-Bin
关键词:	關聯規則;中式速食;最佳化問題;平行運算;Association rules;FP-growth;Chinese fast food;optimization problem;parallel computing
日期:	2015
上传时间:	2016-01-22 14:58:46 (UTC+8)
摘要:	隨著國人外食習慣逐漸增長，餐飲業發展也備受關注，而中式速食是餐飲業提升其經營效率的一個方向。本研究之個案公司即為一中式複合式快餐連鎖業者，該公司之策略為提升顧客消費單價與提升服務速度，並確立以提升套餐選購率做為行動方針。而該公司套餐是以固定的基底內容再加上特色主餐與主食類品項而成。欲實施此行動方針，必須確保固定的套餐基底內容是否符合顧客喜好，以及該基底內容價錢是否合乎其水準。因此，本研究利用個案公司所提供的POS資料中找尋到隱含的顧客喜好並透過關聯分析找出餐點品項間的關聯性。本研究亦將套餐基底設計視為最佳化問題，欲將套餐基底品項間的關聯最大化，並以求解所得之最佳套餐基底作為套餐設計之依據。本研究考量到POS資料隨營運成長將造成的效率問題，因此建置hadoop分散式運算平台，並採用平行FP-Growth 演算法作為關聯分析工具。透過資料清除、篩選、探勘、套入模型後本研究方法以個案公司之POS資料實證，結果顯示，透過本研究所提出的方法設計出求解所得之套餐基底內容，品項間的支持度較原有之套餐基底內容高，且價錢也在合理範圍內，顯示本研究方法於實務上的可行性。 As the growth of dining-out population in the recent years, the development of the food and beverage industry is getting attention, and Chinese fast-food restaurants are particularly considered as the way to improve the operating efficiency in the food and beverage industry. The case company of this study is a Chinese fast food chain restaurant. To enhance its operating efficiency, the company''s tactics are to encourage the expenditure by customer per transaction and to improve the service speed by serving more value meals (i.e. combo) to customers. The design of the company’s value meal is based on some fixed base items coupled with main dishes. To implement this operational policy, the company must confirm that the base items for value meals meet customer preferences, as well as appropriate prices. This study utilizes the POS data to find implicit information regarding customer preferences by the association analysis between individual items. The design of the value meal base is considered as an optimization problem where the objective is to maximize the overall associations in a value meal base. Considering the fast growth of POS data in the future, we adopt Hadoop as the computing platform, and use parallel FP-Growth algorithm for association analysis. Through data cleaning, filtering, and solving the optimization model, the empirical results confirm that the proposed approach is feasible. The results demonstrate that our designs of value meals render greater associations among items in a value meals than the current menu of the company can support.
显示于类别:	[資訊管理學系暨研究所] 學位論文

文件中的档案:

档案	描述	大小	格式	浏览次数
index.html		0Kb	HTML	121	检视/开启

在機構典藏中所有的数据项都受到原著作权保护.

TAIR相关文章

数据加载中.....