Name
Title | ||
|---|---|---|
Who is repinning?: predicting a brand's user interactions using social media retrieval |
Abstract | ||
|---|---|---|
Despite the fact that firms spend heavily in marketing their brands across social media platforms, very little is understood about what media content, in a predictive manner, can generate high interaction rates among their prospects and customers. However, such understanding can significantly help brand marketers generate desired engagements with their target audience in marketing campaigns. In this paper, we study the problem of predicting a brand's user interactions on social media using the example of Pinterest, an emerging platform that has provided a large volume of brand as well as user data in the form of images. Specifically, we treat the prediction of a brand's user interactions, captured through "repinnings" on Pinterest, as the retrieval of relevant user-pinned images given a brand image. The prototype system that we build incorporates this basic principle, and is tested on a large-scale Pinterest dataset of more than one million images. We demonstrate that our system achieves significant lifts in recalling ground truth repinners of brand images for a variety of brands across several major industry categories. |
Year | DOI | Venue |
|---|---|---|
2013 | 10.1145/2501217.2501218 | MDM@KDD |
Field | DocType | Citations |
Content marketing,Brand image,Social media,Advertising,Ground truth,Target audience,User modeling,Engineering | Conference | 1 |
PageRank | References | Authors |
0.42 | 14 | 3 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Shantanu Singh | 1 | 31 | 5.06 |
| Yan Wang | 2 | 1 | 0.42 |
| Lei Ding | 3 | 1 | 0.42 |