Automated document filing and processing methods and systems

Abstract

Systems, methods and computer program products for automatically ingesting and filing documents in a database having a plurality of file locations. An electronic file having one or more documents is received. For each document in the received file, text data is identified and used to generate a plurality of suggested file locations for the received documents. Machine learning systems may be used to enhance the accuracy of suggested file locations.

We claim: 
     
         1 . A method for automatic ingestion and filing of documents in a database having a plurality of file locations, the method comprising:
 receiving an electronic file including at least one document;   for each document in the at least one document
 identifying text data in the document; and 
 generating a plurality of suggested file locations for each respective document. 
   
     
     
         2 . The method of  claim 1 , wherein generating the plurality of suggested file locations comprises:
 processing the text data at a master node to generate a plurality of suggested file locations, wherein the master node is a machine learning node common to a plurality of users;   processing the text data at a client node to refine the plurality of suggested file locations for one of the plurality of users, wherein the client node is a machine learning node specific to the one of the plurality of users.   
     
     
         3 . The method of  claim 2 , wherein the plurality of suggested file locations generated at the master node comprises first or second level file locations in a hierarchy. 
     
     
         4 . The method of  claim 2 , wherein the plurality of suggested file locations generated at the master node comprises third or higher level file locations in a hierarchy. 
     
     
         5 . The method of  claim 1 , wherein generating the plurality of suggested file locations comprises:
 comparing the one or more document keywords to a corpus of stored keywords, the corpus of stored keywords previously generated based on a plurality of documents in the database, wherein each of the stored keywords in the corpus has at least one file location association identifying a file location associated therewith; and   generating a plurality of keyword scores based on the comparison of the one or more document keywords and the corpus.   
     
     
         6 . The method of  claim 1 , further comprising
 identifying a plurality of pages in the file;   determining a plurality of page markers for each page;   determining that the at least one document in the file comprises a plurality of distinct documents; and   assigning each page to one of the distinct documents by grouping the plurality of pages into the distinct documents by comparing the page markers for the plurality of pages.   
     
     
         7 . The method of  claim 6 , wherein the page markers comprise image-based page markers derived from a visual appearance of the page. 
     
     
         8 . The method of  claim 6 , wherein the page markers comprise text-based page markers derived from the text data in the document. 
     
     
         9 . The method of  claim 1 , further comprising:
 for each stored keyword in the corpus of stored keywords, determining a location-specific weighting for each file location association; and   generating the plurality of suggested file locations by weighting the plurality of keyword scores using the location-specific weightings.   
     
     
         10 . The method of  claim 9 , wherein:
 the database is arranged into a file directory having a plurality of folder levels with each file location in the plurality of file locations associated with a particular folder level, and   the location-specific weighting for each file location association is determined using the folder level of the file location corresponding to that file location association.   
     
     
         11 . The method of  claim 1 , further comprising, for each document in the at least one document:
 determining a keyword coefficient for each of the document keywords in the text data, each keyword coefficient indicating a measure of importance of the corresponding document keyword to the document; and   generating the plurality of keyword scores using the keyword coefficient.   
     
     
         12 . The method of  claim 11 , wherein the measure of importance of the corresponding document keyword to the document is determined by:
 identifying keyword text attributes for the document keyword, the keyword text attributes including at least one of a text size, a text location and a text format; and   determining the keyword coefficient for the document keyword in the text data based on the keyword text attributes.   
     
     
         13 . The method of  claim 1 , wherein identifying the text data comprises performing optical character recognition on the document to identify the text data. 
     
     
         14 . The method of  claim 1 , further comprising, determining a recommended file name for one of the received documents by:
 determining a keyword coefficient for each of the document keywords in the text data, each keyword coefficient indicating a measure of importance of the corresponding document keyword to the document; and   determining the recommended file name using the keyword coefficients of the document keywords.   
     
     
         15 . A computer program product for automatic ingestion and filing of documents in a database, the computer program product comprising a non-transitory computer readable storage medium and computer-executable instructions stored on the computer readable storage medium, the instructions for configuring a processor to:
 receive an electronic file including at least one document;   for each document in the at least one document
 identify text data in the document; and 
 generate a plurality of suggested file locations for each respective document.