意見領袖在評論網的功用：以Yelp社交網路為例

Are Opinion Leaders Satisfied? Social Network on Yelp

李承恩(Cheng-En Li)

指導教授 ： 謝志昇

國立臺灣大學/社會科學院/經濟學研究所/碩士(2021年)

https://doi.org/10.6342/NTU202101964

摘要

Yelp 作為世界用戶數量數一數二的用戶評論網，社群中的意見領袖對於公司行號、用戶及平台本身應該有所助益。本研究藉該評論網之公開資料集，分析其菁英會員在評論網路中，對於餐廳、一般用戶及平台的影響，並提供洞見。本研究提供兩種菁英會員影響他人的可能機制：口耳相傳與評論廣播。實證結果顯示菁英會員在兩個機制中皆不影響其他用戶的評分，且菁英會員可能促進其他用戶撰寫評論。對餐廳來說，招攬菁英會員撰寫評論無助於增加餐廳評分，但可能可以提昇名氣。對 Yelp 來說，引進菁英會員制度不會影響其他人的評分，並可能促進社群活躍程度。

關鍵字

社會網路 ； 意見領袖 ； 因果圖模型 ； 有向無環圖 ； 口耳相傳

並列摘要

Yelp ranks as one of the most visited business review sites in the world. On Yelp, prolific and high-quality reviewers, referred to as elites, may significantly affect the businesses, other users, and the platform itself. By investigating the potential impacts of these elite users, this research provides insights into the various roles played by these option leaders on this platform.
We propose two mechanisms explaining the elite's effect on others: word-of-mouth (WOM) and broadcasting. When elite users directly connect to other users (through friendship links) on Yelp and influence their behaviors, this is called WOM. When elite users do not directly link to other non-elite users but influence their behaviors through reviews and tips, this mechanism is called broadcasting.
The results suggest that no matter through WOM or broadcasting, elites do not affect other users' rating behavior. However, having more elite friends or being exposed to more elites' reviews are associated with a greater tendency towards writing more reviews. For restaurants, soliciting elites in order to improve ratings is futile. Nevertheless, it does help to gain higher popularity. As for Yelp, the elite program does not distort the rating system but increases community engagement.

並列關鍵字

social network ； opinion leader ； causal graph ； directed acyclic graph ； word-of-mouth (WOM)

參考文獻

Agha, Leila, and David Molitor.2018. “The Local Influence of Pioneer Investigators on Technology Adoption: Evidence from New Cancer Drugs.”The Review of Economics and Statistics100(1): 29–44.https://doi.org/10.1162/REST_a_00670.

Bellemare, Marc F, Jeffrey R Bloem, and Noah Wexler. 2020. “The paper of how: Estimating treatment effects using the frontdoor criterion.” Working Paper, University of Minnesota. http://marcfbellemare.com/wordpress/wp-content/uploads/2020/09/BellemareBloemWexlerFDCSeptember2020.pdf.

Carman, Ashley. 2018. “Why do you leave restaurant reviews?.” The Verge, April. https://www.theverge.com/2018/4/10/17215784/whyd-you-push-that-button-yelp-tripadvisor-reviews-restaurant.

Casaló, Luis V, Carlos Flavián, and Sergio Ibáñez Sánchez. 2020. “Influencers on Instagram: Antecedents and consequences of opinion leadership.” Journal of Business Research 117: 510–519. https://doi.org/10.1016/j.jbusres.2018.07.005.

Chernozhukov, Victor, Hiroyuki Kasahara, and Paul Schrimpf.2021. “Causal impact of masks, policies, behavior on early covid-19 pandemic in the U.S..” Journal of Econometrics 220(1): 23–62. https://doi.org/10.1016/j.jeconom.2020.09.003.