Login / Signup

The impact of affiliation naming proximity on the retrieval efficiency of Chinese universities-affiliated retractions in the Retraction Watch Database.

Shaoxiong Brian XuYunru ChenHuifang LiuEn XuGuangwei Hu
Published in: Accountability in research (2024)
The Retraction Watch Database (RWDB) is widely used to retrieve retraction data. However, its lack of affiliation normalization hinders the retrieval efficiency of retraction data for specific research-performing organizations. A query for a university name in the RWDB may yield retraction data entries for other universities with similar names, giving rise to the issue of affiliation naming proximity. This study assessed the impact of this issue on the retrieval efficiency of retraction records for 2,692 Chinese university names in English. The analysis revealed that the retrieval efficiency of retraction records for 206 Chinese university names can be influenced by 408 university names. As of 2022, the retrieval efficiency of retraction records for 96 Chinese university names was compromised by the involvement of 402 university names, resulting in an overall retraction inflation rate of 37.9% and an average rate of 45.0%. The findings highlight the importance of curating retraction data through affiliation-specific queries in the RWDB, adhering to the official English names of Chinese universities for scholarly publishing, and adopting the Research Organization Registry system for affiliation disambiguation. Given the significance of this issue concerning the English names of universities in non-English-speaking countries, the identified causes of the problem and proposed solutions can offer valuable insights for improving the retrieval of retraction records for non-Chinese universities in the RWDB.
Keyphrases
  • big data
  • machine learning
  • artificial intelligence
  • deep learning