透過您的圖書館登入
IP:3.21.76.0
  • 學位論文

量販店產品推薦系統之設計

A Design of Product Recommendation System for Hypermarket

指導教授 : 周永山

摘要


在當今的生活中,推薦系統已經是個隨處可見的資訊過濾方法,不管是YouTube、Netflix等影音串流平台的影片推薦、或是Google瀏覽器的廣告推薦,都是推薦系統的應用,但現實生活中的實體量販店卻較少看見類似的商品推薦系統,因此本論文設計一套量販產品銷售推薦系統,以便對每位顧客進行個人專屬的商品推薦。 本系統包含下列三個子系統,(1)人臉辨識系統、(2)消費習慣遠端資料庫、(3)手持商品推薦系統。本系統由於需要對顧客進行個人專屬的商品推薦,因此需辨識顧客身分,在考慮便利性等情況下,採用人臉辨識作為身份辨識的方法。資料庫是為了儲存每位顧客的相關資料及消費紀錄,藉著分析資料庫的資料來推測顧客的喜好,如此才能夠決定推薦何種商品給顧客。手持商品推薦系統的部分使用即時通訊平台Line開發的Line Bot聊天機器人來傳送推薦商品的訊息。選擇使用Line Bot是因為在台灣的網路使用者中,有高達95.7%的人有使用Line,而透過Line Bot便可以在不需要額外安裝其他App的情況下,直接透過Line傳送推薦商品的訊息給顧客,非常快速且便利。

並列摘要


A recommendation system nowadays has been an information filtering method that can be seen in many places, for examples the video or movie recommendations of YouTube, Netflix and other video streaming platforms, or the advertisement recommendation of the Google browser. They all use recommendation systems. However, in real-life shopping malls, product recommendation systems are rarely seen. Therefore, this thesis designs a hypermarket sales recommendation system to make individual product recommendations for each customer. The system consists of three subsystems: (1) face recognition system, (2) customers’ habits database, and (3) handheld product recommendation system. Since this system needs to make individual product recommendations for each customer, it needs to identify the customer's identity. Considering the convenience and other factors, face recognition is adopted as the identification method. The database is to store the relevant information and consumption records of each customer. By analyzing the data in the database, the customer preferences can be understood which facilitates the choice of the products for making recommendations to the customers. The part of the handheld product recommendation system is to use the Line Bot developed by the instant messaging platform Line to send the recommended product message. The reason for using Line Bot is that among the Internet users in Taiwan, 95.7% are using Line according to the data. The system can directly send product recommendation message to customers through Line without the need to install other apps. It’s fast and convenient.

參考文獻


[1] 經濟部統計處,產業經濟統計簡訊,https://www.moea.gov.tw/Mns/dos/bulletin/Bulletin.aspx?kind=9 html=1 menu_id=18808 bull_id=9429
[2] 柯榮宗,行動商務自動化推薦系統之研究-以台糖量販店為例,國立臺南大學數位學習科技學系碩士學位論文,中華民國101年1月。
[3] Infosys調查報告,https://www.infosys.com/newsroom/press-releases/Documents/genome-research-report.pdf
[4]資策會產業情報研究所,零售電商消費者調查,https://mic.iii.org.tw/news.aspx?id=622
[5] DATAREPORTAL,《Digital 2022: Taiwan》,

延伸閱讀