A Multi-Objective Learning to re-Rank Approach to Optimize Online Marketplaces for Multiple Stakeholders

Abstract

Multi-objective recommender systems address the difficult task of recommending items that are relevant to multiple, possibly conflicting, criteria. However these systems are most o en designed to address the objective of one single stakeholder, typically, in online commerce, the consumers whose input and purchasing decisions ultimately determine the success of the recommendation systems. In this work, we address the multi-objective, multi-stakeholder, recommendation problem involving one or more objective(s) per stakeholder. In addition to the consumer stakeholder, we also consider two other stakeholders; the suppliers who provide the goods and services for sale and the intermediary who is responsible for helping connect consumers to suppliers via its recommendation algorithms. We analyze the multi-objective, multi-stakeholder, problem from the point of view of the online marketplace intermediary whose objective is to maximize its commission through its recommender system. We define a multi-objective problem relating all our three stakeholders which we solve with a novel learning-to-rerank approach that makes use of a novel regularization function based on the Kendall tau correlation metric and its kernel version; given an initial ranking of item recommendations built for the consumer, we aim to re-rank it such that the new ranking is also optimized for the secondary objectives while staying close to the initial ranking. We evaluate our approach on a real-world dataset of hotel recommendations provided by Expedia where we show the effectiveness of our approach against a business-rules oriented baseline model.

Multi-objective optimization for long tail recommendation

“…Recently, user-oriented objectives such as user sentiment are considered for better recommendation [30], [31]. For a commercial RS, CTR (Click Through Rate) and GMV (Gross Merchandise Volume) are included in [32], [33] to gain higher profits.…”

Sample-Rank: Weak Multi-Objective Recommendations Using Rejection Sampling

“…Das, A. et al [7] also take a post processing optimization approach where they adjust output of the recommender system according to item profitability. In [6] Nguyen, P. et al combine multiple models tuned for each objective separately such that the new combined score is close to predictions from independent models.…”

Companion Proceedings of The 2019 World Wide Web Conference

“…The integration of the perspectives of multiple parties into recommendation generation and evaluation is the underlying goal of the new sub-field of multistakeholder recommendation [4,7,8,12]. The goal of a recommender system in a multistakeholder environment is, therefore, to generate recommendations taking all the stakeholder's needs and preferences into account.…”

“…In many real-world contexts, the system may gain some utility when recommending items, which could be in the form of a simple aggregate of the other stakeholders' utilities. In many e-commerce settings, the system will get a commission for each sale, and such benefits can be considered together with personalization [12]. Alternatively, the system may seek to tailor outcomes specifically to achieve particular objectives.…”