# On the Consistency of Average Embeddings for Item Recommendation

arXiv: https://arxiv.org/abs/2308.12767 · HTML: https://mlsanity.com/paper/2308.12767/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 11 of its 34 references that are in the map.
- First posted on arXiv 24 Aug 2023.

## Builds on (most cited first, 11 of 11)

- [Neural Collaborative Filtering](https://mlsanity.com/paper/1708.05031/) (2017)
- [Session-based Recommendations with Recurrent Neural Networks](https://mlsanity.com/paper/1511.06939/) (2015)
- [Billion-scale Commodity Embedding for E-commerce Recommendation in Alibaba](https://mlsanity.com/paper/1803.02349/) (2018)
- [PinnerSage: Multi-Modal User Embedding Framework for Recommendations at Pinterest](https://mlsanity.com/paper/2007.03634/) (2020)
- [Current Challenges and Visions in Music Recommender Systems Research](https://mlsanity.com/paper/1710.03208/) (2017)
- [Security and Privacy on Blockchain](https://mlsanity.com/paper/1903.07602/) (2019)
- [A Semi-Personalized System for User Cold Start Recommendation on Music Streaming Apps](https://mlsanity.com/paper/2106.03819/) (2021)
- [Collaborative Translational Metric Learning](https://mlsanity.com/paper/1906.01637/) (2019)
- [Session-aware Information Embedding for E-commerce Product Recommendation](https://mlsanity.com/paper/1707.05955/) (2017)
- [A Scalable Framework for Automatic Playlist Continuation on Music Streaming Services](https://mlsanity.com/paper/2304.09061/) (2023)
- [Hierarchical Latent Relation Modeling for Collaborative Metric Learning](https://mlsanity.com/paper/2108.04655/) (2021)

## Built on by (most cited first, 5 of 5)

- [Scaling User Modeling: Large-scale Online User Representations for Ads Personalization in Meta](https://mlsanity.com/paper/2311.09544/) (2023)
- [Interact2Vec -- An efficient neural network-based model for simultaneously learning users and items embeddings in recommender systems](https://mlsanity.com/paper/2506.22648/) (2025)
- [Spattack: Subgroup Poisoning Attacks on Federated Recommender Systems](https://mlsanity.com/paper/2507.06258/) (2025)
- [Struc-EMB: The Potential of Structure-Aware Encoding in Language Embeddings](https://mlsanity.com/paper/2510.08774/) (2025)
- [Constrained user-item allocation for e-commerce marketing campaigns](https://mlsanity.com/paper/2606.09623/) (2026)

## Abstract (arXiv)

A prevalent practice in recommender systems consists in averaging item embeddings to represent users or higher-level concepts in the same embedding space. This paper investigates the relevance of such a practice. For this purpose, we propose an expected precision score, designed to measure the consistency of an average embedding relative to the items used for its construction. We subsequently analyze the mathematical expression of this score in a theoretical setting with specific assumptions, as well as its empirical behavior on real-world data from music streaming services. Our results emphasize that real-world averages are less consistent for recommendation, which paves the way for future research to better align real-world embeddings with assumptions from our theoretical setting.