It is very regrettable that the quality of Semantic Scholar (SS) leaves a lot to be desired! Besides tons of duplicate publications, there also some serious issues concerning authors. Here is one glaring example! I focus only on machine learning & AI authors.
"Semantic Scholar is a free, AI-powered search and discovery tool for scientific literature, developed by the Allen Institute for Artificial Intelligence." (Google search result). Besides SS, there is only Google Scholar with a similar comprehensive approach. I am a very heavy user of SS.
This blog post is about following ML & AI researcher from China: Dacheng Tao. Here is his Google Scholar profile. He is a very prolific and frequent author, according to Google Scholar he published a total of 1909 papers, book chapters etc. since about 1988. I have been following his publications for several years now.
I personally know it is not easy to handle the Romanization of Chinese names. However, e.g. the first name Dacheng appears to be fairly rare among researchers in the field of ML & AI in my experience.
I suspect, SS has lots of duplicate papers and duplicate authors in its database. This is an example of it. Unfortunately, high quality is not the goal of SS. I have notified SS of many duplicate papers over the years. SS does not fix it nor do they ever respond to my notifications.
SS search result for the name Dacheng Tao. How many of these 10 authors is the same? My strong hunch is they could all be the same author, but I did not check it more carefully.
Following Dacheng Tao was not even listed: https://www.semanticscholar.org/author/Dacheng-Tao/2276069056. Semantic Scholar shows only 10 publication for this author. This author is actually the same as the one listed by Google Scholar and is also the same as the ones found by SS (see screen print above)
No comments:
Post a Comment