Modelling and Classifying the Components of a Literature Review
This work addresses the need for automated analysis of scientific literature to aid in generating high-quality literature reviews, representing an incremental advance in AI methods for text classification.
The paper tackles the problem of classifying sentences in scientific literature by rhetorical roles to support literature review generation, introducing a new annotation schema and evaluating 37 large language models (LLMs) on a novel benchmark, achieving over 96% F1 score with fine-tuning. It also shows that lightweight open-source models and semi-synthetic data can enhance performance.
Previous work has demonstrated that AI methods for analysing scientific literature benefit significantly from annotating sentences in papers according to their rhetorical roles, such as research gaps, results, limitations, extensions of existing methodologies, and others. Such representations also have the potential to support the development of a new generation of systems capable of producing high-quality literature reviews. However, achieving this goal requires the definition of a relevant annotation schema and effective strategies for large-scale annotation of the literature. This paper addresses these challenges by 1) introducing a novel annotation schema specifically designed to support literature review generation and 2) conducting a comprehensive evaluation of a wide range of state-of-the-art large language models (LLMs) in classifying rhetorical roles according to this schema. To this end, we also present Sci-Sentence, a novel multidisciplinary benchmark comprising 700 sentences manually annotated by domain experts and 2,240 sentences automatically labelled using LLMs. We evaluate 37 LLMs on this benchmark, spanning diverse model families and sizes, using both zero-shot learning and fine-tuning approaches. The experiments yield several novel insights that advance the state of the art in this challenging domain. First, the current generation of LLMs performs remarkably well on this task when fine-tuned on high-quality data, achieving performance levels above 96\% F1. Second, while large proprietary models like GPT-4o achieve the best results, some lightweight open-source alternatives also demonstrate excellent performance. Finally, enriching the training data with semi-synthetic examples generated by LLMs proves beneficial, enabling small encoders to achieve robust results and significantly enhancing the performance of several open decoder models.