Skip to content

Scripts used to make and evaluate OpenAlex's concept tagging model

License

Notifications You must be signed in to change notification settings

ourresearch/openalex-concept-tagging

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

36 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

openalex-concept-tagging

This repository contains all of the code for getting the OpenAlex concept tagger up and running. Go into the model iteration directory (V1 or V2) to find a more detailed explanation of how to use this repository. To learn more about concepts in OpenAlex, check out the docs.

Model Iterations

  • V1 (no longer used)
  • V2 (no longer used)
  • V3 (currently used)

Both a V1 and a V2 model were created but as of right now, the V3 model is being used in OpenAlex. Initially, abstract data was not available for the model so we went with a V1 model that only looked at paper titles and a few other features. Once paper abstract data became available, a V2 model was created and we saw a substantial increase in performance. In order to meet the needs for some of our users, a V3 model was created which used the same base tagging model that was developed for V2 but added additional logic for assigning parent concepts so that all concepts would have a path to the top of our concept tree/graph. For more information, please read the information at the top of the V3 directory.

Model Development

You can find an explanation of the modeling and deployment process at the following link: OpenAlex: End-to-End Process for Concept Tagging

Concepts

Input can be tagged with one or more of about 65,000 concepts, listed here. Concepts are part of a hierarchical tree, with levels 0 (e.g., Mathematics) through 5 (e.g., Generalized inverse Gaussian distribution).

About

Scripts used to make and evaluate OpenAlex's concept tagging model

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •