Academic Article Recommendation Using Multiple Perspectives

TL;DR

This paper explores combining content-based filtering and graph-based methods for academic search, leveraging embeddings like Specter and ProNE to improve recommendation quality by capturing different aspects of scholarly literature.

Contribution

It identifies nine differences and potential synergies between CBF and GB, proposing hybrid approaches using Specter and ProNE embeddings for enhanced academic recommendations.

Findings

Specter effectively encodes abstracts for content-based filtering.

ProNE captures citation-based relationships among papers.

Hybrid methods show promise for improved recommendation accuracy.

Abstract

We argue that Content-based filtering (CBF) and Graph-based methods (GB) complement one another in Academic Search recommendations. The scientific literature can be viewed as a conversation between authors and the audience. CBF uses abstracts to infer authors' positions, and GB uses citations to infer responses from the audience. In this paper, we describe nine differences between CBF and GB, as well as synergistic opportunities for hybrid combinations. Two embeddings will be used to illustrate these opportunities: (1) Specter, a CBF method based on BERT-like deepnet encodings of abstracts, and (2) ProNE, a GB method based on spectral clustering of more than 200M papers and 2B citations from Semantic Scholar.