整合社群網絡之多準則餐廳推薦系統

近年來，隨著資訊交流的發達，人們可獲得越來越多的美食資訊，使得消費者越來越重視食物的品質、價位、以及餐廳服務態度和口碑；由於社群網絡的興起，消費者已習慣透過網路資訊和好友推薦以選擇更適合自己消費習慣的餐廳，社群所能擁有的資訊已多過於一些傳統的美食網站，但社群系統的訊息更替快速，使用者在搜尋舊有資料上有些難度。本研究的目的是希望建立一個以使用者為主（個人化）的餐廳推薦系統，結合使用者本身的餐廳偏好以及使用者周遭好友的經驗分享（由社群得知），幫助使用者過濾掉不感興趣或沒有習慣去的餐廳，以達到隨時隨地的有效推薦，透過三層的過濾模型所得的推薦結果顯示在視覺化地圖上，以簡化使用者餐廳選擇決策。經過一系列的驗證過程，我們比較了有社群經驗及沒有社群經驗的推薦結果，以及經驗多寡的影響；評估結果發現有社群經驗的推薦結果成功率及準確度較高，而經驗的多寡會使推薦的成功率及準確度上升。

關鍵字

資訊過濾；擷取技術；網路使用；社群網站；推薦系統

並列摘要

In recent years, people obtain more food information with developed information exchange system. Consumers pay more attention on the food quality, price, restaurant service attitude and reputation. Consumers have been habit to observe related information through Internet and recommendation to choose the satisfying restaurant for their own consumption habits. Through the rising of social networks, the amount of the related information possessed by a social network is much higher than the traditional delicacies web sites, but the information of the social network update much faster than traditional ones that is difficult to search the required message. The purpose of this study is to build-up a user-based (or personal) restaurant recommendation system, which combined the user restaurant preferences and share experiences around their friends (i.e., to learn from the social networks). The system helps filter out noninterest or non-habitual restaurant to achieve an effective recommendation anywhere. The recommendation results, which are fielded through the proposed three fielder levels model, are shown on the virtualization map to simply restaurant selective decision. After a series of validation process, we have compared the recommendation result by using experience of community and no experience of community, and the impact of the amount of experiences; we found that community experience has a higher success rate and accuracy, and the level of experience increases the success rate and accuracy.

並列關鍵字

information filtering and retrieval technology ； internet use ； social network ； recommendation system

參考文獻

資策會 FIND 網站(2014)。2013 年第3 季我國行動上網觀測。存取日期：2014 年9月15 日，取自：http://www.find.org.tw/find/home.aspx?page=many&id=382。

Google Scholar

Statista.com. Number of social network users worldwide from 2010 to 2018 (in billions).Retrieved September 15, 2014, From http://www.statista.com/statistics/278414/number-of-worldwide-social-networkusers/

Google Scholar

Statisticbrain.com, Facebook Statistics. Retrieved September 15, 2014, From http://www.statisticbrain.com/facebook-statistics/

Google Scholar

Jiang, G.,Ma, F.,Shang, J.,Chau, Patrick Y.K.(2014).Evolution of knowledge sharing behavior in social commerce: An agent-based computational approach.Information Sciences.278,250-266.