How far does scientific community look back?

TL;DR

This paper investigates how the scientific community's usage of literature evolves over time by analyzing real-time download data, revealing patterns in article attention span and the impact of social media on rediscovering older papers.

Contribution

It introduces the concept of dynamic usage data and provides empirical analysis of real-time download patterns of scientific articles.

Findings

Recent papers have higher download rates than older ones.

Average lifetime of a downloaded paper is approximately 4.1 years.

Social media can quickly revive interest in older literature.

Abstract

How does the published scientific literature used by scientific community? Many previous studies make analysis on the static usage data. In this research, we propose the concept of dynamic usage data. Based on the platform of realtime.springer.com, we have been monitoring and recording the dynamic usage data of Scientometrics articles round the clock. Our analysis find that papers published in recent four years have many more downloads than papers published four years ago. According to our quantitative calculation, papers down-loaded on one day have an average lifetime of 4.1 years approximately. Classic papers are still being downloaded frequently even long after their publication. Additionally, we find that social media may reboot the attention of old scientific literature in a short time.