ParsRec: Meta-Learning Recommendations for Bibliographic Reference Parsing

TL;DR

This paper introduces ParsRec, a meta-learning system that recommends the most suitable bibliographic reference parser for each reference string, significantly improving extraction accuracy over individual parsers.

Contribution

ParsRec is a novel meta-learning approach that recommends the best parser(s) for reference strings, enhancing parsing accuracy by learning at the field level.

Findings

2.6% increase in F1 score over the best parser

20.2% reduction in false positive rate

18.9% reduction in false negative rate

Abstract

Bibliographic reference parsers extract metadata (e.g. author names, title, year) from bibliographic reference strings. No reference parser consistently gives the best results in every scenario. For instance, one tool may be best in extracting titles, and another tool in extracting author names. In this paper, we address the problem of reference parsing from a recommender-systems perspective. We propose ParsRec, a meta-learning approach that recommends the potentially best parser(s) for a given reference string. We evaluate ParsRec on 105k references from chemistry. We propose two approaches to meta-learning recommendations. The first approach learns the best parser for an entire reference string. The second approach learns the best parser for each field of a reference string. The second approach achieved a 2.6% increase in F1 (0.909 vs. 0.886, p < 0.001) over the best single parser (GROBID), reducing the false positive rate by 20.2% (0.075 vs. 0.094), and the false negative rate by 18.9% (0.107 vs. 0.132).