Author
Listed:
- Park, Jinhee
- Ahn, Hyeongjin
- Kim, Dongjae
- Park, Eunil
Abstract
With the notable growth of the Internet, a number of platforms have emerged and attracted an enormous number of users. Based on the impact of these platforms, some ‘influencers’ are highlighted. These influencers wield significant power, shaping consumer behavior. This influence spawned the concept of influencer marketing, where companies leverage these personalities to advertise their products. YouTube stands out as a prominent platform in this trend. However, considering the limited number of influencers and their concepts, the majority of companies, which hope to conduct their marketing campaigns with influencers face challenges in identifying suitable influencers for their campaigns. With this trend, we introduce GNN-IR, a graph neural network for influencer recommendation, based on the connections between companies and influencers of YouTube, one of the largest content platforms. In developing GNN-IR, we adopted a data-driven methodology utilizing a meticulously curated dataset collected in-house. Our dataset comprises a total of 25,174 relationship entries between advertisers and influencers, involving 1,886 distinct advertisers and 3,812 unique YouTube influencers. It encompasses diverse data modalities, including images, text, and assorted metadata. The data was sourced from two primary platforms: YouTube and ugwanggi. Ugwanggi provided valuable insights into the relationships between advertisers and influencers via their information. Meanwhile, YouTube offered more comprehensive and detailed influencer-centric information. We employed PyTorch Geometric to construct a bipartite graph representing interconnected data. Our recommendation system operates via link prediction, suggesting the Top-k influencers to advertisers based on the calculated connection probability between nodes. To assess GNN-IR's performance, we employed a range of evaluation metrics. For link prediction, we measured Accuracy, Precision, Recall, and F1-score. Additionally, in the recommendation phase, we evaluated Precision@k, Recall@k, and F1-score@k. Using GNN-IR and incorporating profile images from YouTube, keyword features, metadata, and sentiment gleaned from YouTube comments, we achieved Precision levels of 96.51% at k=1 and 93.68% at k=10. Based on the experimental results, several implications and limitations are presented. The collected dataset is publicly available at https://github.com/dxlabskku/GNN-IR.git.
Suggested Citation
Park, Jinhee & Ahn, Hyeongjin & Kim, Dongjae & Park, Eunil, 2024.
"GNN-IR: Examining graph neural networks for influencer recommendations in social media marketing,"
Journal of Retailing and Consumer Services, Elsevier, vol. 78(C).
Handle:
RePEc:eee:joreco:v:78:y:2024:i:c:s0969698924000018
DOI: 10.1016/j.jretconser.2024.103705
Download full text from publisher
As the access to this document is restricted, you may want to search for a different version of it.
Corrections
All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:eee:joreco:v:78:y:2024:i:c:s0969698924000018. See general information about how to correct material in RePEc.
If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.
We have no bibliographic references for this item. You can help adding them by using this form .
If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: Catherine Liu (email available below). General contact details of provider: https://www.journals.elsevier.com/journal-of-retailing-and-consumer-services .
Please note that corrections may take a couple of weeks to filter through
the various RePEc services.