Changes

The answer to connecting the Patent database to the Assignment database lies somewhere in using the information in the table DOCUMENT_INFO to get connect to a patent_id from the Patent database PATENT table for each assignment in the Assignment database ASSIGNMENT table. On further investigation into "The patent_id will be stored USPTO Patent Assignment Dataset: Descriptions and Analysis, the field they called "grant_doc_num" in DOCUMENT_INFO is the patent number (the description of DOCUMENT_INFO table once it above has been found, but it is unclear currently how exactly to use the information in DOCUMENT_INFO altered to accomplish reflect this.)

According the The paper on "The USPTO Patent Assignment Dataset: Descriptions and Analysis" mentions in the text of the paper that there will be errors in patent numbers number stored in DOCUMENT_INFO under grant_doc_num so a separate table constructed from DOCUMENT_INFO called DOCUMENT_INFO_ADMIN if the patent was assigned a patent numberconstructed to determine how prevalent errors were. The They queried the patent number is "from for the appno_doc_num from administrative data." The "appno_doc_num" is the Application Document USPTO number from DOCUMENT_INFO.On further investigation into the paper, it became clear that "administrative data" refers to internal USPTO data that was available to the authors of the paper, but that we might not have. However, in 99% of cases the grant_doc_num and the queried patent number based on the appno_doc_num were the same, so we can probably rely on grant_doc Also in the paper, they mentioned the Assignments on the Web for Patents (AOTW-P), a searchable database of individual USPTO assignment records keyed on reel-frame identification, patent number, and assignor or assignee name (https://assignment.uspto.gov/patent/index.html#/patent/search). Obviously it would not be possible to individually use this tool to query all the patent numbers, but if it would be possible to write a script to somehow query each patent number using the rf_id and parse the response, this could be very useful