Abstract: A method of automatically extracting data from an electronic document containing a plurality of layout features through progressive refinement is provided. The method includes: analyzing each document to automatically extract images and text features wherein each document includes at least two features that are related to each other, and wherein said analyzing compares extracted features with a first search space of candidate features to try and recognize the extracted features; if one of the at least two related features is not recognized and at least one feature is recognized, selecting a second search space of candidate features in response thereto and in response to predefined rules about the relationship between the two features; and comparing the unrecognized feature with said selected second search space.

Abstract: In a document analysis system that receives and processes jobs from a plurality of users, in which each job may contain multiple electronic documents, to extract data from the electronic documents, a method of automatically extracting data from each received electronic document at least in part using data external to the electronic document but associated with the job containing the document is provided. The method includes: analyzing each electronic document in a job to automatically extract images and text features; and, if any of the images and text features extracted from the electronic document is not recognized, using data external to said document but associated with said job to identify the unrecognized feature, wherein the external source may be one of at least one other document in the job and a database having known values associated with the job.

Abstract: A method of grouping electronic document pages of a job that belong together is provided. The method includes: automatically analyzing images and text features extracted from each received electronic document page to associate the electronic document page with a corresponding document category; automatically identifying features extracted from the electronic document page that potentially indicate to which document group the electronic document page belongs; comparing the identified features with a set of group identifying features associated with corresponding document group, in which the set of group identifying features includes at least a set of page numbers and account numbers; and, if the identified features are found to include a set of a page number and an account number belonging to the set of group identifying features associated with the corresponding document group, grouping the electronic document page into the corresponding document group.

Abstract: In a document analysis system that receives and processes jobs from a plurality of users, in which each job may contain multiple electronic documents, to extract data from the electronic documents, a method of automatically correcting the extracted data using known constraints amongst semantics of extracted data elements is provided. The method includes: analyzing each electronic document in a job to automatically extract data; automatically analyzing the extracted data to identify incorrectly extracted data elements using rules defining constraints amongst semantics of extracted data elements; and automatically attempting to correct the incorrectly extracted data elements using the rules.

Abstract: A method of training a document analysis system to extract data from documents is provided. The method includes: automatically analyzing images and text features extracted from a document to associate the document with a corresponding document category; comparing the extracted text features with a set of text features associated with corresponding category of the document, in which the set of text features includes a set of characters, words, and phrases; if the extracted features are found to consist of the characters, words, and phrases belonging to the set of text features associated with the corresponding document category, storing the extracted text features as the data contained in the corresponding document; and, if the extracted text features are found to include at least one text feature that does not belong to the set of text features associated with the corresponding document category, submitting the unrecognized text features to a training phase.

Abstract: A method of automatically narrowing data search space and improving accuracy of data extraction using known constraints in a layout of extracted data elements for classified documented is provided.