DRIFT: A Toolkit for Diachronic Analysis of Scientific Literature

TL;DR

DRIFT is an open-source toolkit designed for diachronic analysis of scientific literature, enabling researchers to track research trends and semantic changes over time using various analysis methods.

Contribution

The paper introduces DRIFT, a comprehensive, easy-to-use toolkit for diachronic research analysis, combining existing methods with new techniques for tracking scientific literature evolution.

Findings

Demonstrated utility on arXiv cs.CL corpus

Effectively tracks research trend dynamics

Identifies semantic drift and trend shifts

Abstract

In this work, we present to the NLP community, and to the wider research community as a whole, an application for the diachronic analysis of research corpora. We open source an easy-to-use tool coined: DRIFT, which allows researchers to track research trends and development over the years. The analysis methods are collated from well-cited research works, with a few of our own methods added for good measure. Succinctly put, some of the analysis methods are: keyword extraction, word clouds, predicting declining/stagnant/growing trends using Productivity, tracking bi-grams using Acceleration plots, finding the Semantic Drift of words, tracking trends using similarity, etc. To demonstrate the utility and efficacy of our tool, we perform a case study on the cs.CL corpus of the arXiv repository and draw inferences from the analysis methods. The toolkit and the associated code are available here: https://github.com/rajaswa/DRIFT.