資料探勘於 Facebook 直播商業模式發展之研究

The Study of Data Mining Approach Implements on Facebook Live Streaming Business Model development

潘品茗(Pimchanok Puttong)

指導教授 ： 廖述賢

淡江大學/商管學院/經營管理全英語碩士學位學程/碩士(2019年)

https://doi.org/10.6846/TKU.2019.00916

摘要

Facebook Live已迅速在流媒體世界中流行起來。儘管它是一個新進入市場的人，但近80％的線上視頻觀眾已經在平臺上觀看。最重要的是，Facebook Live受眾群體非常投入：他們觀看直播視頻的時間比預先錄製的視頻長三倍，並以10倍的速度對其進行評論。在每週觀看直播的消費者中，Facebook Live上的觀看量超過任何其他頻道。本研究探討了用戶體驗，以瞭解用戶的行為，包括他們與其他用戶的互動。這項研究還達到了我們的最終目標，即提出新的服務和功能，以滿足具有不同特徵和消費習慣的使用者，提供有價值的商業模式。通過來自1,082名受訪者的資料庫的結構化集合，我們首先應用K-means聚類將Facebook Live用戶分組為（1）k-pop男孩，（2）學生博主，以及（3）Fanfiction;然後實現關聯分析以生成解釋三種模式之間關係的關聯規則。最後，研究結果告訴管理者獲取用戶注意力的重要性，並在管理隱含部分中揭示Facebook Live上購物活動的可能性。

關鍵字

社群媒體 ； 社群互動 ； 資料探勘 ； Facebook Live Streaming ； 商業模式

並列摘要

Facebook Live has quickly become popular in the streaming media world. Nearly 80 percent of online- video audiences are already watching on the platform, despite it being a newer entrant to the market. Most importantly, Facebook Live audiences are highly engaged: They watch live videos three times longer than pre-recorded ones and comment on them at 10 times the rate. And among consumers who view live-streams on a weekly basis, more watch on Facebook Live than any other channel. This study examines user experiences to understand users’ behavior, including their interaction with other users. This study also reaches our final goal to suggest new services and functions that cater to users with different characteristics and consuming habits, delivering valuable business models. By a structured collection of the database from 1,082 respondents, we first apply K-means clustering to group Facebook Live users into (1) k-pop boys, (2) Student bloggers, and (3) Fanfiction; association analysis is then implemented to generate association rules that explain the relationship among three patterns. Finally, study results inform managers of the importance of gaining users attention and reveal the possibility for shopping activity on Facebook Live in the managerial implication section.

並列關鍵字

Social media ； Social interaction ； Data mining ； Facebook Live Streaming ； Business model

參考文獻

Reference

Abdul K. A., N. & Sebastian M. P. (2009). Improving the Accuracy and Efficiency of the k-means Clustering Algorithm. Proceedings of the World Congress on Engineering Vol I, London, U.K.

Ahamed, S. B. M. & Hareesha, K. S. (2012). Dynamic Clustering of Data with Modified K-Means Algorithm. Department of Computer Science & Engineering, Manipal Institute of Technology, Manipal. International Conference on Information and Computer Networks (ICICN 2012)

Animal, Z, Edwards, K., Kumar, Pilades, P., Rdz, J., & Agios Nikolaos. (2018, November 16). Facebook Video: The Guide Marketers Are Looking For. Retrieved from https://adespresso.com/blog/facebook- video-marketing/

Apostolopoulos, J., Tan, W., & Wee, S. (2002). Video Streaming: Concepts, Algorithms, and Systems. Mobile and Media Systems Laboratory. HP Laboratories Palo Alto.