Search papers, labs, and topics across Lattice.
This paper introduces a novel approach for generating advertising headlines for e-commerce platforms by leveraging a Reinforcement Learning (RL) Policy gradient method applied to Transformer-based Masked Language Models. The significance of this work lies in its ability to produce high-quality, creative headlines at scale, addressing a common challenge faced by sellers in meeting the creative standards of advertising. Key results indicate that the proposed method not only surpasses existing Transformer and LSTM + RL approaches in overlap metrics and quality audits but also outperforms human-generated headlines in grammar and creativity.
Model-generated ad headlines beat human creativity and grammar, setting a new standard for e-commerce advertising.
For any E-commerce website it is a nontrivial problem to build enduring advertisements that attract shoppers. Many sellers often find it hard to pass the creative quality bar of the website, especially at a large scale. We thus propose a programmatic solution to generate product advertising headlines using retail content. We propose a state of the art application of Reinforcement Learning (RL) Policy gradient methods on Transformer (Vaswani et al., 2017) based Masked Language Models (Devlin et al., 2019). Our method creates the advertising headline by jointly conditioning on multiple products that a seller wishes to advertise. We demonstrate that our method outperforms existing Transformer and LSTM + RL methods in overlap metrics and quality audits. We also show that our model-generated headlines outperform human (advertiser) submitted headlines in terms of both grammar and creative quality as determined by audits.
Amazon Science