lin-ao / enhancing_the_makg Goto Github PK

Hi Ao,

Thanks so much for this paper. The author name disambiguation is definitely helpful for future research. Particularly, I wonder if you are willing to share the 14.PaperAuthorAffiliations_new.txt file after disambiguation or the 13.all_positives.txt file? I couldn't find the authors' affiliation at time of the paper publication in the 24.PaperAuthorAffiliations_disambiguated.nt.bz2 file.

Thank you so much!

Hi Ao,
Very interesting paper, these enhancements such as author name disambiguation and paper classification will definitely benefit MAKG. In these enhancements, I am particularly interested in FOS classification, and we have verified that your FoS is more accurate than the original MAG FoS.
Currently, we try to use the enhanced FOS for our applications, but we found that only one FOS could be obtained for each MAKG paper, this puts us in a dilemma. We do not have sufficient computational resources to handle a huge dataset like MAKG. I am wondering if it is possible to access your pre-trained model (SciBERT) for FoS classification.