Modelling and Classifying the Components of a Literature Review Journal: Engineering Applications of Artificial Intelligence

Francisco Bolaños, Angelo Salatino, Francesco Osborne, Enrico Motta · 2025

Read Modelling and Classifying the Components of a Literature Review Journal: Engineering Applications of Artificial Intelligence by Francisco Bolaños, Angelo Salatino, Francesco Osborne, Enrico Motta (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-04337.txt.

Description

Francisco Bolaños, Angelo Salatino, Francesco Osborne, and Enrico Motta classify the rhetorical roles of sentences in scientific papers in a 2025 paper. Tagging sentences as gap, result, limitation, or method extension helps machines read literature and could feed automatic review writers. A clean schema and a hard benchmark were both missing.

Their benchmark joins seven hundred expert-labeled sentences with over two thousand machine-labeled ones across fields. Thirty-seven large language models were tested, zero-shot and fine-tuned. Fine-tuning on good data pushed scores above ninety-six percent, with small open models rivaling giants. Adding semi-synthetic examples lifted small encoders further. Reliable automatic annotation now looks within reach.

Catalog record

CREATORFrancisco Bolaños, Angelo Salatino, Francesco Osborne, Enrico Motta
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE85 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-04337

Browse the full library · About the Tower of Babel