pyResearchInsights—An open‐source Python package for scientific text analysis

Abstract

Abstract With an increasing number of scientific articles published each year, there is a need to synthesize and obtain insights across ever‐growing volumes of literature. Here, we present pyResearchInsights, a novel open‐source automated content analysis package that can be used to analyze scientific abstracts within a natural language processing framework. The package collects abstracts from scientific repositories, identifies topics of research discussed in these abstracts, and presents interactive concept maps to visualize these research topics. To showcase the utilities of this package, we present two examples, specific to the field of ecology and conservation biology. First, we demonstrate the end‐to‐end functionality of the package by presenting topics of research discussed in 1,131 abstracts pertaining to birds of the Tropical Andes. Our results suggest that a large proportion of avian research in this biodiversity hotspot pertains to species distributions, climate change, and plant ecology. Second, we retrieved and analyzed 22,561 abstracts across eight journals in the field of conservation biology to identify twelve global topics of conservation research. Our analysis shows that conservation policy and landscape ecology are focal topics of research. We further examined how these conservation‐associated research topics varied across five biodiversity hotspots. Lastly, we compared the utilities of this package with existing tools that carry out automated content analysis, and we show that our open‐source package has wider functionality and provides end‐to‐end utilities that seldom exist across other tools.

Keywords

• automated content analysis
• exploratory analysis
• natural language processing