Needs-based collaborative filtering for embedded insurance recommendation on e-commerce platforms
Year: 2025
Type: article
Abstract: Distribution channels such as bancassurance, brokers, agents, direct online sales, and insurance aggregators have been key to ensuring premium growth for both life and non-life insurers. However in recent years, an emerging channel known as embedded insurance has started to provide insurers with a brand-new growth driver. In this paper, we first present an introduction to embedded insurance – what it is and how it will shape insurance distribution in the industry. We then introduce a framework to classify embedded insurance recommendation system. Finally, we propose a novel insurance recommendation system using supervised learning algorithms that can be applied to e-commerce platforms. This needs-based collaborative filtering technique recommends one of three insurance products that would be most appropriate for each buyer on the Olist e-commerce platform based on order-level data. Our work is relevant for actuaries in this field interested in the pricing of embedded insurance risk as well as insurers seeking to improve insurance penetration on such platforms.
Source: British Actuarial Journal
Authors: Zhan Liang Chan, Niharika Anthwal, Xin Yung Lee
Institution: Institute and Faculty of Actuaries
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid:
OpenAlex ID: W4411836238