In work I was doing on #46, I ended up dealing a bit with the ORCID fetching process. I discovered a better way to summarize the ORCID API content using the pyOrcid package and made a couple of contributions to that project to further improve the result. I need to re-run all of the ORCID caching to store this new and improved structure and then process the information for useful person entity content.
Some records contain useful additional names. I will likely keep the name people have used in their USGS Staff Profiles as presumably preferred but will add other names I find to aliases.
Keywords in ORCID are essentially like expertise terms in USGS Profiles in that they are simple a list of terms. I need to look at these comprehensively between the two sources and decide how to handle them in the GeoKB. The best approach would be to combine topic modelling with full texts of creative works and other contributions and then get those concepts fully organized into the graph, but I need to move faster to get concepts stubbed out so they can be linked and then do further work on knowledge organization later.
I haven't decided if there is anything processable out of other content in ORCID records, but I'll store it for now and eventually get back to knowledge organization work.
In work I was doing on #46, I ended up dealing a bit with the ORCID fetching process. I discovered a better way to summarize the ORCID API content using the pyOrcid package and made a couple of contributions to that project to further improve the result. I need to re-run all of the ORCID caching to store this new and improved structure and then process the information for useful person entity content.