Last updated: 2022-02-04

paper-rec goal is to recommend users what scientific papers to read next based on their preferences. This is a test model used to explore Hugging Face Hub capabilities and identify requirements to enable support for recommendation task in the ecosystem.

2022-02-04

Recommender System model with support of a Language Model for feature extraction.

The overall idea for paper-rec test model is inspired by this work: NU:BRIEF – A Privacy-aware Newsletter Personalization Engine for Publishers.

However, for paper-rec, we use a different language model more suitable for longer text, namely Sentence Transformers: Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks, in particular: sentence-transformers/all-MiniLM-L6-v2.

The intended direct users are recommender systems' practitioners and enthusiasts that would like to experiment with the task of scientific paper recommendation.

The data used for this model corresponds to the RSS news feeds for arXiv updates accessed on 2022-02-04. In particular to the ones related to Machine Learning and AI:

N/A

The model is limited to the papers fetched on 2022-02-04, that is, those papers are the only ones it can recommend.