從社群平台中找尋高效益的網紅組合

Finding High-Utility Influencers on Social Media Platforms

陳映樵(Ying-Ciao Chen)

指導教授 ： 李瑞庭

國立臺灣大學/管理學院/資訊管理學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202101394

摘要

越來越多的公司行號在社群平台上創立自身的品牌帳號，或與具有影響力的網紅合作來宣傳品牌形象或產品，如何為公司選擇有影響力的網紅已成為一個重要的研究議題。但若合作的網紅具有類似的喜好，他們會有較多的共同粉絲，為了讓預算的效益最大化，所選擇的網紅，彼此間最好具備低相似度。因此，本研究提出一個架構，透過考慮網紅的影響力與相似度，從目標群體中找出高效益的網紅組合，使得組合中的網紅具高影響力且低相似度。首先，我們利用每位網紅所發佈、分享和喜歡的貼文來計算網紅間的相似度；接著，我們利用Cobb-Douglas生產函數和PageRank演算法建立影響力模型，計算每位網紅的影響力，所建立的影響力模型可調整各項參數，以滿足公司行號的需求；最後，我們提出一個方法，考慮網紅的影響力與彼此間的相似度，挑選出高效益的網紅組合。實驗結果顯示，我們的方法與最優解方法相近，並且優於其它比較方法。本研究不僅可以幫助公司選擇高效益的網紅組合，將預算運用效率最大化，還可以提供不同的影響力模型與不同的選擇方法來滿足公司行號的需求。

關鍵字

網紅影響力 ； 影響力最大化 ； Cobb-Douglas生產函數 ； 貪婪演算法 ； 社群平台

並列摘要

More and more companies spread brand awareness by creating their own business accounts or cooperating with social media influencers. How to select influencers for businesses to implement an effective marketing strategy has become an important issue. If the selected influencers have similar profiles, they may share many followers on a social media platform. To maximize the budget efficiency, it is better to employ the influencers with least similarity. Therefore, it is desirable and essential to help businesses select high-utility influencers, where the selected influencers have high influence but least similarity between each other. In this study, we propose a framework to select high-utility influencers from a target group by considering influencers’ influence and similarity. First, we compute the similarity between influencers by using the posts posted, shared and liked by each influencer. Second, based on the Cobb-Douglas production function and the PageRank algorithm, we develop an influence model to derive the influence for each influencer for fulfilling businesses’ various needs. Finally, we propose a novel method to select high-utility influencers by considering both influence and similarity between influencers. The experiment results show that our proposed framework is comparable to the optimal method and outperforms the comparing methods. Our framework can not only help businesses select high-utility influencers for maximizing budget efficiency, but also provide different selection methods with various influence models to meet their needs such as increasing product exposure, promoting brand awareness and images, etc.

並列關鍵字

influencer influence ； influence maximization ； Cobb-Douglas production function ； greedy algorithm ； social media platforms

參考文獻

Ali ME, Eusuf SS, Abdullah K, Choudhury FM, Culpepper JS, Sellis T (2018) The maximum trajectory coverage query in spatial databases. Proceedings of the VLDB Endowment 12(3):197–209.

Aral S (2011) Commentary—Identifying social influence: A comment on opinion leadership and social contagion in new product diffusion. Marketing Science 30(2):217–223.

Aral S, Walker D (2011) Creating social contagion through viral product design: A randomized trial of peer influence in networks. Management Science 57(9):1623–1639.

Aslay C, Matakos A, Galbrun E, Gionis A (2018) Maximizing the diversity of exposure in a social network. Proceedings of the IEEE International Conference on Data Mining. 863–868.

Chae I, Stephen AT, Bart Y, Yao D (2017) Spillover effects in seeded word-of-mouth marketing campaigns. Marketing Science 36(1):89–104.