Patent Application
20080162602
1. Field of the Invention
Systems and methods described herein relate generally to information retrieval and, more particularly, to the archiving user information for subsequent searching and retrieval.
2. Description of Related Art
Modern computer networks, and in particular, the Internet, have made large bodies of information widely and easily available. Internet search engines, for instance, index many millions of web documents that are linked to the Internet. A user connected to the Internet can enter a simple search query to quickly locate web documents relevant to the search query.
In addition to publicly available documents, such as websites and other online documents, recent endeavors have been made to facilitate the indexing and storing of user documents, such as word processing documents, emails, music, etc. Applications such as Google Desktop Search, Copernic Desktop Search, and Apple Computer, Inc.'s Safari typically crawl designated portions of a user's local storage and maintain an index of searchable documents identified therein. Unfortunately, conventional document indexing tools do not provide for storage or efficient indexing of non-text based documents.
According to one aspect, a method may include receiving a document image. The document image may be converted into a text document. Searchable information may be obtained relating to the text document. At least one searchable metadata element may be associated with the text document. The text document and the at least one searchable metadata element may be stored for subsequent retrieval based on the at least one searchable metadata element.
According to another aspect a system may include a document capture system configured to capture an image of a document and a processor system. The processor system may be configured to identify text contained within the image; generate a text document based on the identified text; obtain searchable information relating to the text document; associate at least one searchable metadata element with the text document; and transmit the text document and the at least one searchable metadata element to a database via a computer network for subsequent retrieval based on the at least one searchable metadata element.
According to yet another aspect, a method may include receiving an image document; identifying text contained within the image document; generating a text document based on the identified text; obtaining searchable information relating to the text document; associating at least one searchable metadata element with the text document based on the searchable information; and storing the text document and the at least one searchable metadata element in a database for subsequent retrieval based on the at least one searchable metadata element.
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate an embodiment of the invention and, together with the description, explain the invention. In the drawings,
The following detailed description of the invention refers to the accompanying drawings. The same reference numbers in different drawings may identify the same or similar elements. Also, the following detailed description does not limit the invention.
More and more types of documents are becoming searchable via search engines. For example, some documents, such as personal documents, financial documents, receipts, correspondence, etc. may be scanned and their text recognized via optical character recognition (OCR). Consistent with implementations described herein, it may be beneficial to enable archiving and searching of these documents in an efficient and simple manner.
Systems and methods consistent with embodiments described herein may facilitate capturing or retrieval of documents and assignment of relevant metadata information to the documents. The documents may be OCR'd or otherwise processed to generate a textual version of the captured document. The document and its associated metadata and text version may be stored in an online repository or server, such that the document information may be easily searchable or retrievable by a number of devices based on information included in the text version and the associated metadata.
As described in more detail below, processing system 120 may be configured to perform OCR on documents captured or otherwise retrieved by document capture system 110 to recognize text associated with the document. Processing system 120 may include a client entity, where an entity may be defined as a device, such as a personal computer, a wireless telephone, a personal digital assistant (PDA), a laptop, or another type of computation or communication device, a thread or process running on one of these devices, and/or an object executable by one of these devices. In other aspects, processing system 120 may include a server entity that gathers, processes, searches, and/or maintains documents. In such an aspect, a “thin client” device may be configured to interact with sever-based processing system 120, where processing of documents may be performed remotely to the client device.
In one implementation, OCR processing by processing system 120 may be performed on an entirety of each captured document, with no preconfigured metadata associated therewith. In an alternative implementation, OCR processing may be based on a template or preliminary configuration that may be either automatically selected by processing system 120 or selected and/or configured by a user. Templates may assign searchable metadata to sections of documents or may instruct processing system 120 to OCR only predetermined portions of documents.
Using the bank statement example from above, a bank provided OCR template may instruct processing system 120 as to what portions of the statement relate to what kinds of information. For example, a first portion of statement documents may include account information, while a second portion may include transaction information. The template may further indicate that only the transaction information portion of the statement should be OCR'd. By providing information about a document in advance of OCR or other processing of the document, information capturing may be performed more efficiently. In one exemplary implementation, templates may be stored or otherwise maintained on a template database 155 of template database server 150 and may be accessible via network 130. In another embodiment (not shown), template database server 150 and/or template database 155 may be local to processing system 120. Additional details relating to the above-described implementations are set forth in detail below.
Document database server 140 may include a document database 145 configured to store the OCR'd text associated with a document as well as any metadata assigned to or associated with the captured document. In one implementation, an electronic copy of the captured document may be stored in document database 145 as well. As shown, in one implementation, document database server 140 may be connected to processing system 120 via network 130. However, in alternate implementations, document database server 140 and/or document database 145 may be stored locally with respect to processing system 120.
Document database server 140 may store a documents textual information and metadata information within a database record of document database 145. In one implementation, the records of document database 145 may be arranged to form a relational database, although any suitable database structure may be implemented in accordance with aspects described herein.
Network 130 may include a local area network (LAN), a wide area network (WAN), a telephone network, such as the Public Switched Telephone Network (PSTN), an intranet, the Internet, or a combination of networks. Processing system 120 and database servers 140 and 150 may connect to network 130 via wired, wireless, and/or optical connections.
Processor 220 may include a processor, microprocessor, or processing logic that may interpret and execute instructions. Main memory 230 may include a random access memory (RAM) or another type of dynamic storage device that may store information and instructions for execution by processor 220. ROM 240 may include a ROM device or another type of static storage device that may store static information and instructions for use by processor 220. Storage device 250 may include a magnetic and/or optical recording medium and its corresponding drive.
Input device 260 may include a mechanism that permits an operator to input information to system 110/120, such as a keyboard, a mouse, a pen, voice recognition and/or biometric mechanisms, etc. Output device 270 may include a mechanism that outputs information to the operator, including a display, a printer, a speaker, etc. Communication interface 280 may include any transceiver-like mechanism that enables system 110/120 to communicate with other devices and/or systems. For example, communication interface 280 may include mechanisms for communicating with another device or system via a network, such as network 130.
As will be described in detail below, system 110/120 may perform certain document processing-related operations. System 110/120 may perform these operations in response to processor 220 executing software instructions contained in a computer-readable medium, such as memory 230. A computer-readable medium may be defined as a physical or logical memory device and/or carrier wave.
The software instructions may be read into memory 230 from another computer-readable medium, such as data storage device 250, or from another device via communication interface 280. The software instructions contained in memory 230 may cause processor 220 to perform processes that will be described later. Alternatively, hardwired circuitry may be used in place of or in combination with software instructions to implement processes in various aspects of the invention. Thus, implementations of the invention are not limited to any specific combination of hardware circuitry and software.
More specifically, operating system 310 may include operating system software, such as the Microsoft Windows®, Unix, or Linux operating systems. OCR software 320 may include or use software (e.g., drivers) for interfacing with document capture system 110 to initiate capturing of document images by document capture system 110. Additionally, OCR software 320 may include software for converting an image of a captured document to a text version. As described briefly above, OCR software 320 may use a template retrieved from template database server 150 to facilitate efficient recognition of the document and assignment of metadata elements thereto.
In one implementation consistent with aspects described herein, OCR software 320 may determine an OCR confidence for a converted document that indicates or otherwise determines a likelihood that a document image has been accurately converted to a text version. In one embodiment, OCR software may initiate a rescan or recapture of a document image when the OCR confidence is below a predetermined level. In one implementation, the rescan or recapture may be performed at an increased resolution. In still a further implementation, OCR confidence may be generated for each area identified in a template, with rescan or recapture only being performed when the OCR confidence for predetermined areas are below the predetermined level. Alternatively, OCR confidence thresholds for different areas of a document may be different, depending on a relative importance of the information contained therein. This eliminates unnecessary delays caused by rescanning or recapturing data from unimportant or less important areas, while maintaining highly accurate conversions for more important areas.
Document management software 330 may include software for enabling a manual review of a text version of a document(s) output by OCR software 320. Document management software 330 may provide for the correction or editing of the text version, as well as the assignment of metadata elements to one or more portions of the text version. For example, continuing with the bank statement example described above, a statement date or date range and a bank or account name may be assigned to the document. Additionally, certain portions of the document may be assigned a “debit” metadata element, while additional portions of the document may be assigned a “credit” metadata element. Document management software 330 may provide for storage of the text version, its associated metadata elements, and/or its associated document image to document database server 140 for subsequent searching and retrieval. In one implementation, document management software 330 may include an image management application such as Google® Lighthouse™ or Picasa®.
Assignment of metadata elements to a searchable text version of a document may facilitate more efficient retrieval of information contained in the document, using a combination of document data as well as one or more metadata elements. For example, a document including a particular transaction may be more easily retrieved in response to a user search for a specific payee in the text version as well as a date within the document's date range and a transaction type.
Processing may begin with the document capture system 110 capturing one or more images representing a document (act 510). As described above, one implementation may use conventional scanning techniques to capture images of the pages of the document. Alternatively, document images may be retrieved or captured from an electronic source accessible either locally or from remote resources accessible via network 130.
Once captured, OCR processing may be performed on the document images to generate a textual or searchable version of the document (act 515). OCR processing may involve an analysis of an image for recognizable text and characteristics of the text (e.g., font, size, formatting, etc.) included therein as well as information regarding where the text is located on the pages based on the images of the pages of the document.
In one implementation, OCR processing may be performed on an entirety of each document image. In another implementation, OCR processing may be performed on portions of the document images based on a template retrieved from template database server 150 or, alternatively, from local storage (e.g., data storage device 250). For example, in one implementation, a bank may provide a template from a web site hosted on server 150. In another example, a user may configure or save a template for subsequent use with similar types of documents. As described above, templates may indicate various areas in a type of document and may be used to establish or assign metadata elements to those areas or to the document as a whole. In another implementation consistent with aspects described herein, a template may instruct OCR processing to performing recognition to a certain confidence level.
Once a text version of a document has been generated, a confidence level for the conversion may be determined (act 520). It may then be determined whether the confidence level meets or exceeds a predetermined threshold level indicative of an accurate conversion (act 525). If the predetermined threshold has not been met (act 525—NO), the process may return to act 510 for recapture at a same or enhanced resolution. However, if the predetermined threshold has been met (act 525—YES), the generated text version may be presented to a user for manual review and/or editing (act 530). Any changes, additions, or deletions to the text version may be received (act 535). By providing for a manual review of the generated text version, users may efficiently correct OCR errors and may remove information from the text version that is considered sensitive or confidential.
Next, one or more metadata elements may be associated with or assigned to the text version to facilitate enhanced searching and/or retrieval of the text version (act 540). As described above, information not present in the text of the document, but representative of the document content may be added as metadata elements to either the entire document, or to designated portions of the text document. For example, using the bank statement example initially presented above, metadata elements such as “bank statement”, a document date or date range, account nickname, etc. may be assigned to the text version of the document. Additionally, metadata elements may be assigned to selected portions of the text version of the document. For example, credit transactions may be assigned a “credits” metadata element, while debit transactions in the bank statement may be assigned a “debits” metadata element. In this way, information relating to the OCR'd content may be associated with the text document.
Once desired metadata elements have been assigned or, if initially assigned by a template, removed or edited, the text version and its associated metadata elements may be stored in document database 145 on document database server 140 (act 545). In one exemplary implementation, document database server 140 may be a web server configured to maintain an online storage environment for the user's OCR's documents. In other implementations, users may also store the captured images in document database 145, thereby enabling subsequent retrieval of the actual image document along with its text version.
Systems and methods described herein may automatically identify metadata associated with a document and may create an association between the metadata and the image and/or text version of the document, making both the document content and its associated metadata available for searching and/or other processing.
The foregoing description of preferred embodiments of the present invention provides illustration and description, but is not intended to be exhaustive or to limit the invention to the precise form disclosed. Modifications and variations are possible in light of the above teachings or may be acquired from practice of the invention.
For example, while series of acts have been described with regard to
It will be apparent that aspects of the invention, as described above, may be implemented in many different forms of software, firmware, and hardware in the implementations illustrated in the figures. The actual software code or specialized control hardware used to implement aspects consistent with the principles of the invention is not limiting of the present invention. Thus, the operation and behavior of the aspects were described without reference to the specific software code—it being understood that one would be able to design software and control hardware to implement the aspects based on the description herein.
No element, act, or instruction used in the present application should be construed as critical or essential to the invention unless explicitly described as such. Also, as used herein, the article “a” is intended to include one or more items. Where only one item is intended, the term “one” or similar language is used. Further, the phrase “based on” is intended to mean “based, at least in part, on” unless explicitly stated otherwise.
