Detecting and Tracking The Real-time Hot Topics: A Study on Computational Neuroscience

TL;DR

This paper presents a real-time method for detecting and tracking hot topics in computational neuroscience by analyzing weekly updated article usage data, revealing key technologies and their dynamic trends.

Contribution

It improves existing topic detection methods by utilizing real-time usage data to dynamically monitor hot topics in a specific scientific field.

Findings

Hot topics include 'fmri', 'eeg', 'erp'

Topics are tracked dynamically over time

Real-time data enables timely monitoring of trends

Abstract

In this study, following the idea of our previous paper (Wang, et al., 2013a), we improve the method to detect and track hot topics in a specific field by using the real-time article usage data. With the "usage count" data provided by Web of Science, we take the field of computational neuroscience as an example to make analysis. About 10 thousand articles in the field of Computational Neuroscience are queried in Web of Science, when the records, including the usage count data of each paper, have been harvested and updated weekly from October 19, 2015 to March 21, 2016. The hot topics are defined by the most frequently used keywords aggregated from the articles. The analysis reveals that hot topics in Computational Neuroscience are related to the key technologies, like "fmri", "eeg", "erp", etc. Furthermore, using the weekly updated data, we track the dynamical changes of the topics. The characteristic of immediacy of usage data makes it possible to track the "heat" of hot topics timely and dynamically.