Can Pretrained Language Models Generate Persuasive, Faithful, and Informative Ad Text for Product Descriptions?

When: Wed 9 September 2022, at 1:00 pm (GMT+10)

Speaker: Mr Fajri Koto (U)

Host: Dr Ash Rahimi

Zoom: https://uqz.zoom.us/j/89362232168

Abstract:

For any e-commerce service, persuasive, faithful, and informative product descriptions can attract shoppers and improve sales. While not all sellers are capable of providing such interesting descriptions, a language generation system can be a source of such descriptions at scale, and potentially assist sellers to improve their product descriptions. Most previous work has addressed this task based on statistical approaches (Wang et al., 2017), limited attributes such as titles (Chen et al., 2019; Chan et al., 2020), and focused on only one product type (Wang et al., 2017; Munigala et al., 2018; Hong et al., 2021). In this work, we jointly train image features and 10 text attributes across 23 diverse product types, with two different target text types with different writing styles: bullet points and paragraph descriptions. Our findings suggest that multimodal training with modern pretrained language models can generate fluent and persuasive advertisements, but are less faithful and informative, especially out of domain.

Bio:

Fajri Koto is an incoming postdoctoral researcher at MBZUAI. Previously, he did PhD at NLP Lab, with Professor Timothy Baldwin and Dr. Jey Han Lau at the University of Melbourne, with research focused on discourse analysis and summarization systems. During his PhD, Fajri published his work in top-tier international conferences (e.g. ACL, EMNLP, NAACL, COLING, EACL, AACL) and journals (e.g. JAIR). He also did an internship as an applied scientist at Amazon under the supervision of Prof. Chunhua Shen and Prof. Anton Van De Hengel.