CLIRDec 8, 2021

Zero-Shot Recommendation as Language Modeling

arXiv:2112.04184v164 citations
Originality Incremental advance
AI Analysis

This addresses the need for recommendation systems that do not require structured training data, offering a novel approach for users in domains like movies or products, though it is incremental as it builds on existing language models.

The paper tackles the problem of recommendation systems by proposing a framework that uses pretrained language models on unstructured text to rank items, achieving competitive performance compared to standard matrix factorization methods.

Recommendation is the task of ranking items (e.g. movies or products) according to individual user needs. Current systems rely on collaborative filtering and content-based techniques, which both require structured training data. We propose a framework for recommendation with off-the-shelf pretrained language models (LM) that only used unstructured text corpora as training data. If a user $u$ liked \textit{Matrix} and \textit{Inception}, we construct a textual prompt, e.g. \textit{"Movies like Matrix, Inception, ${<}m{>}$"} to estimate the affinity between $u$ and $m$ with LM likelihood. We motivate our idea with a corpus analysis, evaluate several prompt structures, and we compare LM-based recommendation with standard matrix factorization trained on different data regimes. The code for our experiments is publicly available (https://colab.research.google.com/drive/1f1mlZ-FGaLGdo5rPzxf3vemKllbh2esT?usp=sharing).

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes