University of Illinois Urbana-Champaign

Problematic Citations: Detecting problematic citation patterns with the Open Citations Corpus

Schneider, Jodi

Permalink

https://hdl.handle.net/2142/100880

Description

Title
Problematic Citations: Detecting problematic citation patterns with the Open Citations Corpus
Author(s)
Schneider, Jodi
Contributor(s)
Ye, Di
Issue Date
2018-09-03
Keyword(s)
  • Biomedical literature
  • Citation of retracted research
  • Citation practices
  • OpenCitations Corpus
  • Problematic citation
  • Value of public domain licensing
  • Science of science
Abstract
The structure of citation networks provides evidence about how scientific information is diffused. Problematic citation patterns include the selective citation of positive findings, citation bias, as well as the continued citation of retracted literature (i.e. literature formally withdrawn due to error, fraud, or ethical problems). For instance, there is some evidence that positive results tend to receive more citations. The public domain licensing of the Open Citations Corpus makes it possible, in principle, to estimate the likelihood that any network of research papers suffers from problematic citation. To-date, problematic citation been documented ad-hoc, in several striking studies. In Alzheimer's disease research, biased citation, ignoring critical findings, was used to support successful U.S. NIH grant proposals (Greenberg 2009). Mistranslation of obesity research has been used to justify exertion game research (Marshall & Linehan 2017). Citation of fraudulent research about Chronic Obstructive Pulmonary Disease continued after its retraction (Fulton et al. 2015). The data resulting from such studies is of great use to my lab in replicating and determining how to generalize the detection of problematic citation patterns. Previously, the detection of problematic citation patterns has been a side effect of astute researchers, noticing suspicious findings while conducting systematic literature reviews. This talk will describe work-in-progress in my lab detecting problematic citation patterns using natural language processing, combined with network analysis on the Open Citations Corpus.
Type of Resource
other
Language
en
Permalink
http://hdl.handle.net/2142/100880

Owning Collections

Illinois Research and Scholarship (Open Collection) PRIMARY
This is the default collection for all research and scholarship developed by faculty, staff, or students at the University of Illinois at Urbana-Champaign