Living Literature Reviews

TL;DR

This paper proposes a 'living literature review' model using Linked Data to continuously update and extend the relevance of literature reviews, demonstrated through a prototype and user study.

Contribution

It introduces a novel approach of using Linked Data for dynamic, updatable literature reviews, addressing the issue of rapid knowledge obsolescence.

Findings

The model is technically feasible.

Researchers accept the approach well.

Living reviews score higher than traditional ones.

Abstract

Literature reviews have long played a fundamental role in synthesizing the current state of a research field. However, in recent years, certain fields have evolved at such a rapid rate that literature reviews quickly lose their relevance as new work is published that renders them outdated. We should therefore rethink how to structure and publish such literature reviews with their highly valuable synthesized content. Here, we aim to determine if existing Linked Data technologies can be harnessed to prolong the relevance of literature reviews and whether researchers are comfortable with working with such a solution. We present here our approach of ``living literature reviews'' where the core information is represented as Linked Data which can be amended with new findings after the publication of the literature review. We present a prototype implementation, which we use for a case study where we expose potential users to a concrete literature review modeled with our approach. We observe that our model is technically feasible and is received well by researchers, with our ``living'' versions scoring higher than their traditional counterparts in our user study. In conclusion, we find that there are strong benefits to using a Linked Data solution to extend the effective lifetime of a literature review.