How authors evaluate the novelty of their articles: A comparative analysis of highlights in research articles

Due to the growing capacities of research articles and journals, the tasks of selecting interesting articles and quickly identifying interesting sections of articles have become primary challenges faced by researchers. Therefore, the notion of Highlights, a novel introductory section included in academic publications, has been proposed to directly emphasise the novelty and value of research articles to improve article retrieval and knowledge dissemination. In this article, we developed a classification schema featuring five categories to analyse the content and explore the features of sentences contained in the Highlights sections of articles. Subsequently, we conducted an experiment by using the fields of Library and Information Science (LIS) and Computer Science (CS) as examples to statistically analyse domain differences in the arrangement of Highlights sections. The experiment focused on both the sentence level and the article level and emphasised differences in research paradigms and principles of evaluation. In particular, we discovered that LIS is relatively ‘result-heavy’, while CS is ‘method-heavy’; furthermore, in self-evaluated contributions, LIS authors concentrated more on academic contributions and applications, while CS authors preferred to demonstrate the value of their articles by comparing their research with previous studies.