AmpSum: Adaptive multiple-product summarization towards improving recommendation captions
Explainable recommendation seeks to provide not only high-quality recommendations but also intuitive explanations. Our objective is not on generating accurate recommendations per se, but on producing user-friendly explanations through recommendation captions. Importantly, the focus of existing work has been predominantly on explaining a single item recommendation. In e-commerce websites, product recommendations are usually organized into “widgets”, each given a name, as a recommendation caption, to describe the products within. These recommendation captions are usually generic in nature and inadequate to reveal the purpose of recommendation, in part because they may be manually crafted, making it difficult to attach meaningful and informative names at scale. We propose an Adaptive Multiple-Product Summarization framework (AmpSum) that automatically and adaptively generates widget captions based on different recommended products. The multiplicity of products to be summarized in a widget caption is particularly novel. The lack of well-developed labels motivates us to design a weakly supervised learning approach with distant supervision to bootstrap the model learning from pseudo labels, and then fine-tune the model with a small amount of manual labels. To validate the efficacy of this method, we conduct extensive experiments on several product categories of Amazon data. The results demonstrate that our proposed framework consistently outperforms state-of-the-art baselines over 9.47-29.14% on ROUGE and 27.31% on METEOR. With case studies, we illustrate how AmpSum could adaptively generate summarization based on different product recommendations.