An intelligent marketing platform with influencer classification in social networking services

Collaborations with influencers in social networking services (SNS) for online commercial advertisement has gained significant momentum in recent years. However, the existing approaches suffer from a lack of precision in matching advertisements to specific target user groups (e.g., influencers). To address the issue, in this paper we present an intelligent marketing platform that classifies influencers based on a newly built SNS post text dataset and the comprehensive performance factors. The proposed platform fine-tunes a pretrained language model with the SNS post text dataset of influencers, which can achieve a better classification accuracy. To further improve the matching accuracy for marketing partners, we integrate multiple performance features of influencers into the proposed platform. Our results show that the proposed model can provide more than 90% classification accuracy for business partners in their product advertisements. Furthermore, we used a real-world business requirement to implement and evaluate the proposed platform with a web-based system, which provides an easy way for online commercial advertisement.