...
The USPTO has used OCR software to convert images of approximately 166,000 US patents issued between 1970 and 1976 missing from the current text file. It has also converted the US Patent backfile from 1970 to 1790, which is approximately 3.9 million additional documents. Work has been completed to load the converted text into the USPTO search engine, BRS/Search, for access via the search clients EAST and WEST. The load of the U.S. Patent OCR database is being implemented in two segments. One segment covers the time period 1790 to 1919, while the other segment covers the time period 1920 to 1971. As of January 2002, examiner access was provided through both EAST and WEST search clients to the 1920 to 1970 data. USPTO embellished these text records by obtaining and processing an electronic source of titles and inventor names. Providing access to the segment for the time period 1790 to 1919 has not been planned.
The USPTO’s Data Maintenance Branch and staff perform the data loading and maintenance of both text and image data for the following domestic databases:
Patent Image Retrieval System (PIRS), Patent Images on the Web (PIW), Application Image Retrieval System (AIRS), Application Images on the Web (AIW), Bibliographic Retrieval Service (BRS) Patent Grant and Application Text Database, Publication Site for Issued and Published Sequences (PSIPS), Patent Application Location and Monitoring (PALM - Tape Creation Process), Patent Application Services and Security (PASS Grants and Application - Tape Creation Process), Electronic Filing System (EFS - Tape Creation Process), CD-Rom Reference Library System and the Trademark Image Capture and Retrieval System.
ATRPatents2007Updatingcharts.doc - USPTO Image and Text Repository Growth Charts
...