UNICON: A unified framework for behavior-based consumer segmentation in e-commerce

Manuel Dibak,Vladimir Vlasov,Nour Karessli,Darya Dedik,Egor Malykh,Jacek Wasilewski,Ton Torres,Ana Peleteiro Ramallo

2023-09-18

Abstract:Data-driven personalization is a key practice in fashion e-commerce, improving the way businesses serve their consumers needs with more relevant content. While hyper-personalization offers highly targeted experiences to each consumer, it requires a significant amount of private data to create an individualized journey. To alleviate this, group-based personalization provides a moderate level of personalization built on broader common preferences of a consumer segment, while still being able to personalize the results. We introduce UNICON, a unified deep learning consumer segmentation framework that leverages rich consumer behavior data to learn long-term latent representations and utilizes them to extract two pivotal types of segmentation catering various personalization use-cases: lookalike, expanding a predefined target seed segment with consumers of similar behavior, and data-driven, revealing non-obvious consumer segments with similar affinities. We demonstrate through extensive experimentation our framework effectiveness in fashion to identify lookalike Designer audience and data-driven style segments. Furthermore, we present experiments that showcase how segment information can be incorporated in a hybrid recommender system combining hyper and group-based personalization to exploit the advantages of both alternatives and provide improvements on consumer experience.

Information Retrieval,Machine Learning

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve

This paper aims to address the issue of consumer segmentation in e-commerce, particularly in the field of personalized recommendations for fashion e-commerce. Specifically, the authors propose a unified framework called UNICON (Unified framework for behavior-based CONsumer segmentation) that utilizes consumer interaction behavior data to achieve two types of consumer segmentation:

1. **Lookalike Segmentation**: Given a predefined target seed user group, expand this group through a machine learning model to find users whose behavior is similar to the seed group but do not yet fully meet the business rules.
2. **Data-Driven Segmentation**: Discover non-obvious user groups based on consumer behavior data, where these groups have similar interest preferences.

### Main Contributions
- Propose a unified long-term consumer representation method based on a multi-head self-attention Transformer model.
- Utilize implicit embeddings to extract two types of consumer segmentation and apply them to real data to identify lookalike designer audiences and data-driven style segments.
- Demonstrate how to integrate segmentation information into a hybrid recommendation system, combining hyper-personalization and group-based personalized recommendations to enhance user experience.

The effectiveness of the framework is validated through experiments, showcasing its impact in practical applications.