JP5084341B2 - Document analysis processing apparatus, image processing apparatus, document analysis processing program, document analysis processing method - Google Patents

Description

  The present invention relates to a document analysis processing device, an image processing device, a document analysis processing program, and a document analysis processing method, and in particular, a document analysis processing device, an image processing device, which analyzes a document attribute and performs processing according to the attribute. The present invention relates to a document analysis processing program and a document analysis processing method.

  With the recent development of natural language processing technology and improvement of computer processing capability, conventional document analysis processing devices extract documents with similar contents from a large amount of stored documents and classify them based on the degree of similarity It is possible.

  The following methods are known as methods for determining whether documents are similar. First, in the conventional determination method, a target document is decomposed into elements each having a character string, a word, or a phrase as a unit, and a feature amount is calculated based on the combination of the elements. In the conventional determination method, the similarity of the feature amounts is obtained for all the combinations of documents, and if the similarity is equal to or higher than a certain level, it is regarded as similar.

  Various methods have been devised for calculating feature amounts. For example, in the conventional calculation method, after the target document is decomposed into elements in units of character strings, words, and phrases, the element weights are calculated based on the appearance frequency of each element in the document set and the appearance frequency in the target document. Ask. In the conventional calculation method, the feature amount is expressed by each element and a vector constituted by its weight.

  The similarity is calculated by obtaining the inner product of the vectors. In the conventional classification method based on similarity, the average value of feature quantities (vectors) of document groups defined as having the same classification is calculated, and the similarity between the feature quantity (vector) of the target document and the average vector is calculated. If is more than a certain value, the target document is determined to be in that category. Patent Document 1 describes an example of a technique related to searching for similar documents as described above.

  In addition, each organization such as a company is required to prevent information leakage of trade secrets and personal information handled. Patent Document 2 has a security policy (policy) in each organization that controls access so that only authorized persons can access confidential information, or encrypts confidential information and allows only authorized persons to view it. An example of a technique for preventing this is described.

As described above, in the conventional document analysis processing apparatus, when accessing a document, the above-mentioned technique for searching for similar documents is used, the attribute of the document is estimated from the contents of the document, and the access to the document violates the policy. I was able to monitor it.
JP 2000-148770 A JP 2006-185153 A

  However, in a conventional document analysis processing apparatus that monitors access to a document according to a policy, the user may feel inconvenience during operation as follows. For example, a policy is a principle operation such as “... is prohibited in principle, and is executed with the permission of the manager in charge when it must be executed”. Therefore, the conventional document analysis processing apparatus to which other than the principle can not be applied is often notorious for the following cases, such as inconvenience and inflexibility.

  The first case is an example in which the estimation result of the similar search / classification is logically correct, but the estimation attribute is not applied to the document in operation. A conventional document analysis processing apparatus to which other than the principle is not applicable is not flexible with respect to the first case.

  For example, it is assumed that there is a system for performing classification management by category classification “product A, product B, product C” of a developed product unit. For product A, it is stipulated that only development personnel can refer to it as a countermeasure against leakage to competitors, and “top-secret” treatment is stipulated for non-developer personnel.

  If the pamphlet document of product A is classified, it is determined as “product A” in the conventional document analysis processing apparatus to which other than the principle cannot be applied. However, since pamphlet documents are open to many people, they do not want to treat them as “top secret”. In addition, the draft document of the functional specification of product A is treated as “confidential” except for those involved in development, but if there is a description that can be referred to by other product B teams, it would be possible to refer to other product B teams.

  In such a case, if a similar document is classified after the next time, a result different from the previous determination result is requested. As one method for solving such a situation, in a conventional document analysis processing apparatus in which access control is performed, such as “secret”, “secret”, “private secret”, etc., an ID is assigned to a document regardless of content, and exception control is handled. Permits etc. are also devised as a function. However, the conventional document analysis processing apparatus using a permit is based on the ID assigned to the document, and cannot be applied to a document in which no ID is assigned to the document.

  The second case is an example in which the feedback / relearning function type of an e-mail spam filter is used as a method for changing the determination result of the similar search / classification. In the conventional document analysis processing apparatus using the feedback / relearning function type of the spam filter, the learning database itself is relearned, and the feature amount of the learning database is changed. Learning the correct answer if it is not correct leads to a decrease in the accuracy of the learning database.

  The present invention has been made in view of the above points, and an object of the present invention is to provide a document analysis processing device, an image processing device, a document analysis processing program, and a document analysis processing method that are flexible and can prevent deterioration in accuracy. .

In order to solve the above-described problems, the present invention is a document analysis processing apparatus that analyzes document attributes and performs processing according to the attributes, and associates the characteristics of an exception-handled document with exception-handled exception attributes. The exception attribute analyzing means for analyzing the exception attribute of the document based on the specified exception attribute storing means, and the estimated attribute storage means in which the characteristics of the document not handled as an exception and the estimated attribute not handled as an exception are associated with each other. An estimated attribute analyzing unit that analyzes the estimated attribute of the document, and based on a result of excluding the exception attribute of the document analyzed by the exception attribute analyzing unit from the estimated attribute of the document analyzed by the estimated attribute analyzing unit, a document attribute analysis means for the attribute analysis, a receiving unit for receiving the action information for the document and the document, Accession for the attribute and the document of the document Policy determination based on policy storage means in which the application information is associated with processing according to the document attribute and action information for the document, and based on the policy determination result, the document attribute and the document have a, and policy determination means for performing a process corresponding to the action information for the document, the exceptional attributes storage means, incidental conditions for exception handling is correspondence to exception attributes features and exception handling of the document to be exempt It is characterized by being.

  In addition, what applied the component, expression, or arbitrary combination of the component of this invention to a method, an apparatus, a system, a computer program, a recording medium, a data structure, etc. is also effective as an aspect of this invention.

  According to the present invention, it is possible to provide a document analysis processing device, an image processing device, a document analysis processing program, and a document analysis processing method that are flexible and can prevent a decrease in accuracy.

  Next, the best mode for carrying out the present invention will be described based on the following embodiments with reference to the drawings.

  FIG. 1 is a block diagram of an embodiment of a system according to the present invention. The system shown in FIG. 1 includes a document attribute learning / analysis server 1, a multifunction machine 2, a document learning cooperation program 3, and an exception learning cooperation program 4. The document attribute learning / analysis server 1 is an example of a document analysis processing apparatus. The multifunction device 2 is an example of an image processing apparatus.

  The document attribute learning / analysis server 1 in FIG. 1 includes a document attribute analysis program 11, a document attribute learning program 12, an exception learning program 13, an analysis result DB 14, a policy DB 15, an attribute feature base DB 16, and an exception information DB 17.

  Document attribute learning is performed by the document learning cooperation program 3 that is a learning client and the document attribute learning program 12 that is a learning server. The document attribute learning program 12 registers the document attribute learning result in the attribute feature base DB 16. Exception learning is performed by the exception learning cooperation program 4 as a learning client and the exception learning program 13 as a learning server. The exception learning program 13 registers the result of exception learning in the exception information DB 17.

  The document attribute analysis program 11 receives the analysis target document 5 from the multifunction machine 2 and analyzes the attribute of the document as described later. Then, the document attribute analysis program 11 performs policy determination using the policy DB 15 as described later. When a policy violation is detected, the document attribute analysis program 11 warns the administrator, for example. Finally, the document attribute analysis program 11 registers the result in the analysis result DB 14.

  The document attribute learning / analysis server 1 is realized by a hardware configuration as shown in FIG. FIG. 2 is a hardware configuration diagram of an embodiment of the document attribute learning / analysis server.

  The document attribute learning / analysis server 1 includes an input device 21, an output device 22, a drive device 23, an auxiliary storage device 24, a main storage device 25, an arithmetic processing device 26, and an interface device 27 that are mutually connected by a bus B. Is done.

  The input device 21 includes a keyboard and a mouse, and is used for inputting various signals. The output device 22 includes a display device and is used to display various windows, data, and the like. The interface device 27 includes a modem, a LAN card, and the like, and is used to connect to a network such as the Internet or a LAN.

  The document analysis processing program according to the present invention is at least a part of various programs that control the document attribute learning / analysis server 1. The document analysis processing program is provided, for example, by distributing the recording medium 28 or downloading from the network. The recording medium 28 on which the document analysis processing program is recorded is a recording medium on which information is optically, electrically or magnetically recorded, such as a CD-ROM, flexible disk, magneto-optical disk, ROM, flash memory, etc. Various types of recording media such as a semiconductor memory for electrically recording information can be used.

  When the recording medium 28 on which the document analysis processing program is recorded is set in the drive device 23, the document analysis processing program is installed from the recording medium 28 into the auxiliary storage device 24 via the drive device 23. The document analysis processing program downloaded from the network is installed in the auxiliary storage device 24 via the interface device 27.

  The auxiliary storage device 24 stores the installed document analysis processing program and also stores necessary files, data, and the like. The main storage device 25 reads and stores the document analysis processing program from the auxiliary storage device 24 at the time of activation. The arithmetic processing unit 26 implements various processes as described later in accordance with the document analysis processing program stored in the main storage device 25.

  The document analysis processing program according to the present invention includes a document attribute analysis program 11, a document attribute learning program 12, and an exception learning program 13. The document attribute learning process is realized by the document learning cooperation program 3 and the document attribute learning program 12.

  The document attribute learning program 12 receives a document attribute learning request from the document learning cooperation program 3, and performs registration processing for registering the document attribute learning result in the attribute feature base DB 16 based on the received document. The document learning cooperation program 3 and the document attribute learning program 12 monitor a predetermined folder, and when a confidential document is stored in the folder, the document attribute learning of the confidential document is performed and the document attribute learning result is registered. It can be realized using existing technology.

  The exception learning process is realized by the exception learning cooperation program 4 and the exception learning program 13. FIG. 3 is a flowchart showing the procedure of the exception learning process. In step S1, the exception learning cooperation program 4 receives, for example, a document that is to be treated specially as an exception and an attribute (exception attribute) of the document that is desired to be treated as an exception from a user such as an administrator.

  FIG. 4 is an image diagram of an example of an exception learning screen in which a document to be specially handled as an exception and attributes of the document to be handled as an exception are input. The exception learning screen 40 of FIG. 4 represents an example in which a document to be handled as an exception is input as “target file” and an attribute of the document to be handled as an exception is input as “attribute”. The “target file” can also be input using a file management screen displayed by pressing the “reference” button 41.

  In step S2, the exception learning cooperation program 4 transmits a request for learning exception information to the exception learning program 13 based on the document that the user wants to handle as an exception and the attribute of the document that the user wants to handle as an exception. The request for learning exception information includes a document input by the user that is to be handled as an exception and the attributes of the document that are to be handled as an exception as a learning document and an exception attribute.

  Furthermore, when a context (ancillary condition) necessary for applying the exception attribute is input from the user, the exception information learning request further includes the context. For example, in the context, specify the person to whom the exception attribute is applied (exception is permitted only when Mr. XXX copies), or specify the location where the exception attribute is applied (copy in the xxx room) Etc.).

  In step S3, the exception learning program 13 accepts an exception information learning request from the exception learning cooperation program 4. In step S4, the exception learning program 13 performs exception information registration processing based on the received exception information learning request.

  In the exception information registration process, text information that is the source of the full-text search is extracted from the learning document included in the accepted exception information learning request. When the learning document is an image such as a scanned document, text information is extracted by OCR processing.

  In step S5, the exception learning program 13 calculates a feature quantity for full-text search from the text information, and links the specified exception attribute included in the received exception information learning request together with the feature quantity to the exception information DB 17. Register with

  The exception learning process shown in the flowchart of FIG. 3 can be realized by using existing technology. For example, the processing for calculating the feature amount for full text search from text information can be realized by applying the prior art. The feature quantity for full-text search can be represented by an n-dimensional vector of the appearance frequency and weight of each element in text information decomposed into elements of a combination of character string, word or clause.

  When the feature quantity for full-text search is represented by an n-dimensional vector, the similarity between documents can be calculated as follows. For the calculation of the similarity, for example, a method of calculating by an inner product or cosine between n-dimensional vectors as described in JP 2000-148770 A can be used. If the similarity exceeds the threshold, it is determined that the two documents are similar.

  FIG. 5 is an image diagram of records registered in the exception information DB. The record in FIG. 5 includes feature data of an exception handling document (exception document), an exception document attribute (exception attribute), a context 1 that identifies a person (user) to apply the exception attribute, and a place to apply the exception attribute. And a context 2 that specifies The feature data of the exception document is a feature amount for full-text search calculated from text information. Exception attributes, contexts 1 and 2 are included in the received request for learning exception information.

  Further, the document attribute analysis process is realized by the document attribute analysis program 11, the analysis result DB 14, the policy DB 15, the attribute feature base DB 16, and the exception information DB 17. FIG. 6 is a block diagram of an embodiment of a document attribute analysis program. The document attribute analysis program 11 shown in FIG. 6 includes a document analysis request receiving unit 61, a document analysis determination processing unit 62, an attribute feature base determination unit 63, an exception information determination unit 64, and a policy determination unit 65.

  FIG. 7 is a flowchart showing the procedure of document attribute analysis processing. In step S11, the document analysis request receiving unit 61 receives the analysis target document 5 and action information (for example, who did what) from the outside of the multifunction device 2 or the like via the network, and analyzes the analysis target document 5 and action information. It transmits to the determination process part 62.

  In step S 12, the document analysis determination processing unit 62 requests exception determination from the exception information determination unit 64. The exception information determination unit 64 requested to determine the exception performs a similar document search for searching for an exception document that is substantially the same as the feature data of the analysis target document 5 from the feature data of the exception document registered in the exception information DB 17. Then, the exception information determination unit 64 extracts the exception attribute associated with the searched exception document from the exception information DB 17 (result 1).

  Note that the exception information determination unit 64 searches the exception data that is almost the same as the feature data of the analysis target document 5 from the feature data of the exception document registered in the exception information DB 17, and the similarity threshold is shown in FIG. Set higher than usual. FIG. 8 is a graph showing an example of a similarity threshold used for searching for almost identical exception documents.

  In the graph of the threshold value, the similarity value and distribution differ depending on the document search engine used. Therefore, when a similar document search is performed, a recommended value for the threshold is determined through an experiment in advance. It is desirable to determine the threshold value for the similar document search using sample data (learning document) close to the purpose at the evaluation stage.

  In exception determination, I want to extract similar documents that are nearly identical. Therefore, in the exception determination, a similarity degree normalized so as to be 100% when a question is asked with the same document (question document) as learned is used.

  However, if the learning document or question document is a scanned image obtained from the multifunction device 2, it is difficult to obtain the exact same image or OCR result every time, so even if the same paper image is asked, 100% cannot be obtained. . Therefore, when a similar document search is performed, the similarity threshold is set to a high value (2) considering some difference results. The value (1) represents a normal threshold value considering the balance between the misperception that the correct answer is regarded as an error and the misperception that the error is regarded as a correct answer.

  In step S13, the document analysis determination processing unit 62 requests the attribute feature base determination unit 63 to estimate the attribute. The attribute feature base determination unit 63 requested to estimate the attribute estimates the analysis target document 5 by normal content analysis (similar document search & document classification) based on the document attribute learning result registered in the attribute feature base DB 16. Extract attributes (result 2).

  In step S14, the document analysis determination processing unit 62 performs comprehensive determination on the analysis target document 5 based on the exception attribute of the result 1 and the estimated attribute of the result 2. FIG. 9 is a structural image diagram showing the exception attribute of result 1, the estimated attribute of result 2, and the overall result. As shown in FIG. 9, the structure of the exception attribute of Result 1 is composed of a plurality of exception attributes and reliability. The structure of the estimated attribute of the result 2 includes a plurality of estimated attributes and reliability. The overall result of the analysis target document 5 is obtained by removing the attribute list representing the structure of the exceptional attribute of the result 1 from the attribute list representing the structure of the estimated attribute of the result 2.

  FIG. 10 is a processing image diagram showing the exception attribute of result 1, the estimated attribute of result 2, and the overall result. In the example of FIG. 10, the exception attribute “category A” of the result 1 is excluded from the estimated attributes “category A, category B” of the result 2, and an overall result “category B” is obtained.

  After the comprehensive result is obtained, the policy determination unit 65 proceeds to step S15, and performs policy determination using the policy DB 15 based on the comprehensive result obtained in step S14 and the action information received in step S11. In the policy DB 15, a policy as shown in FIG. 11 is set.

  FIG. 11 is a configuration diagram of an example of a policy set in the policy DB. The policy in FIG. 11 is “When a document of CATEGORY_A is scanned, a warning mail is sent to the administrator”. In the policy DB 15, the attribute “category A” of the document and the action information “scan” for the document are associated with the process “send a warning mail to the administrator”.

  If a policy violation is detected as a result of the policy determination in step S15, the policy determination unit 65 proceeds from step S16 to step S17, and performs duty processing such as warning mail and warning log in accordance with the policy set in the policy DB 15, Proceed to step S18.

  If no policy violation is detected as a result of the policy determination in step S15, the policy determination unit 65 proceeds from step S16 to step S18. In step S18, the document analysis determination processing unit 62 registers the comprehensive result in the analysis result DB 14.

  Note that the system according to the present invention is not limited to the configuration shown in FIG. 1, but may have the configurations shown in FIGS. FIG. 12 is a block diagram of another embodiment of the system according to the present invention. The system shown in FIG. 12 has a configuration in which various functions are incorporated in the multi-function device 2 instead of a server-client cooperation type.

  The system shown in FIG. 12 does not cooperate with the document learning cooperation program 3 and the exception learning cooperation program 4, but takes a form of capturing a document by monitoring the learning source file server 121 from the multifunction peripheral 2 side.

  The MFP 2 includes a document attribute analysis program 11, a document attribute learning program 12, an exception learning program 13, an analysis result DB 14, a policy DB 15, an attribute feature base DB 16, an exception information DB 17, an application 120 such as a copy, a scanner, and a facsimile. It is.

  The document attribute analysis program 11 acquires the analysis target document 5 from the application 120 of its own device, and analyzes the attribute of the document as described above. Then, the document attribute analysis program 11 performs policy determination using the policy DB 15 as described above. When a policy violation is detected, the document attribute analysis program 11 issues a warning to the administrator using, for example, an operation panel. Finally, the document attribute analysis program 11 registers the result in the analysis result DB 14.

  FIG. 13 is a block diagram of another embodiment of the system according to the present invention. The system shown in FIG. 13 has a configuration in which various functions are incorporated in the multifunction machine 2, as in the system shown in FIG. 12, but learning is performed in cooperation with the document learning cooperation program 3.

  The system according to the present invention prepares an exception information DB 17 that is independent from the attribute feature base DB 16 and learns exception information independently, thereby reducing flexibility without reducing the accuracy of the attribute feature base DB 16. Can be increased.

  In addition, in order to specify an exception document by similar document search, by setting a similarity threshold higher than usual in order to reduce ambiguity, it is possible to search for almost identical exception documents. By realizing retrieval of almost identical exception documents, in the system according to the present invention, the document content itself becomes an element for specifying the document.

  The present invention is not limited to the specifically disclosed embodiments, and various modifications and changes can be made without departing from the scope of the claims.

1 is a configuration diagram of an embodiment of a system according to the present invention. FIG. It is a hardware block diagram of one Example of a document attribute learning / analysis server. It is a flowchart showing the procedure of exception learning processing. It is an image figure of an example of the exception learning screen into which the document which wants to handle exception specially and the attribute of the document which wants to handle exception are input. It is an image figure of the record registered into exception information DB. It is a block diagram of one Example of a document attribute analysis program. It is a flowchart showing the procedure of document attribute analysis processing. It is an example of the graph showing the threshold value of similarity used for searching for almost identical exception documents. It is a structure image figure showing the exception attribute of result 1, the presumed attribute of result 2, and the comprehensive result. It is a processing image figure showing the exception attribute of result 1, the presumed attribute of result 2, and the comprehensive result. It is a block diagram of an example of the policy set to policy DB. It is a block diagram of the other Example of the system by this invention. It is a block diagram of the other Example of the system by this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Document attribute learning / analysis server 2 MFP 3 Document learning cooperation program 4 Exception learning cooperation program 5 Document to be analyzed 11 Document attribute analysis program 12 Document attribute learning program 13 Exception learning program 14 Analysis result DB
15 Policy DB
16 Attribute feature base DB
17 Exception information DB
21 Input Device 22 Output Device 23 Drive Device 24 Auxiliary Storage Device 25 Main Storage Device 26 Arithmetic Processing Device 27 Interface Device 61 Document Analysis Request Accepting Unit 62 Document Analysis Determination Processing Unit 63 Attribute Feature Base Determination Unit 64 Exception Information Determination Unit 65 Policy Determination Department 120 Application 121 File server

Claims (5)

A document analysis processing device that analyzes document attributes and performs processing according to the attributes,
An exception attribute analyzing means for analyzing the exception attribute of the document based on an exception attribute storage means in which a feature of the document to be handled as an exception and an exception attribute to be handled as an exception are associated;
An estimated attribute analyzing means for analyzing the estimated attribute of the document based on an estimated attribute storage means in which a feature of a document not handled as an exception and an estimated attribute not handled as an exception are associated;
A document attribute analyzing unit that analyzes the attribute of the document based on a result obtained by removing the exception attribute of the document analyzed by the exception attribute analyzing unit from the estimated attribute of the document analyzed by the estimated attribute analyzing unit ;
Receiving means for receiving the document and action information for the document;
Policy determination is performed based on policy storage means in which the attribute of the document and action information for the document are associated with processing corresponding to the attribute of the document and action information for the document, and based on the result of the policy determination Policy determining means for performing processing according to the attribute of the document and action information for the document;
I have a,
The document analysis processing apparatus, wherein the exception attribute storage means associates an incidental condition for handling an exception with a feature of the document to be handled as an exception and an exception attribute to be handled as an exception . The document document analysis processing device according to claim 1 Symbol mounting characterized in that it is a document that has been processed by the image processing apparatus. An image processing apparatus having at least one of a plotter unit and a scanner unit that analyzes an attribute of a document and performs processing according to the attribute,
An exception attribute analyzing means for analyzing the exception attribute of the document based on an exception attribute storage means in which a feature of the document to be handled as an exception and an exception attribute to be handled as an exception are associated;
An estimated attribute analyzing means for analyzing the estimated attribute of the document based on an estimated attribute storage means in which a feature of a document not handled as an exception and an estimated attribute not handled as an exception are associated;
A document attribute analyzing unit that analyzes the attribute of the document based on a result obtained by removing the exception attribute of the document analyzed by the exception attribute analyzing unit from the estimated attribute of the document analyzed by the estimated attribute analyzing unit ;
Receiving means for receiving the document and action information for the document;
Policy determination is performed based on policy storage means in which the attribute of the document and action information for the document are associated with processing corresponding to the attribute of the document and action information for the document, and based on the result of the policy determination Policy determining means for performing processing according to the attribute of the document and action information for the document;
I have a,
The image processing apparatus according to claim 1, wherein the exception attribute storage means associates an incidental condition for handling an exception with a feature of the document to be handled as an exception and an exception attribute to be handled as an exception . A document analysis processing device that analyzes document attributes and performs processing according to the attributes.
Exceptions attribute analysis means to analyze the exception attributes of the document based on the exception attribute storage means and exception attributes associated to features and exempt of the document to be exempt,
Estimated attribute analysis means to analyze the estimated attributes of the document based on the estimated attribute storage means and estimated attributes associated not characterized and exception handling documents that do not exempt,
The basis of the estimated attributes of the document analyzed by estimating attribute analysis unit to the exception attribute analyzing means results exception attributes of the analyzed document by the document attribute analysis means to analyze the attributes of the document,
Receiving means for receiving the document and action information for the document;
Policy determination is performed based on policy storage means in which the attribute of the document and action information for the document are associated with processing corresponding to the attribute of the document and action information for the document, and based on the result of the policy determination Policy determining means for performing processing according to the attribute of the document and action information for the document
And then allowed to function,
In the exception attribute storage means, the incidental conditions for handling exceptions are associated with the characteristics of the documents to be handled as exceptions and the exception attributes to be handled as exceptions.
Document analysis processing program characterized by A document analysis processing method in a document analysis processing apparatus that analyzes an attribute of a document and performs processing according to the attribute,
An exception attribute analyzing step of analyzing the exception attribute of the document based on an exception attribute storage means in which a feature of the document to be handled as an exception and an exception attribute to be handled as an exception are associated;
An estimated attribute analysis step of analyzing the estimated attribute of the document based on estimated attribute storage means in which a feature of the document that is not handled as an exception and an estimated attribute that is not handled as an exception are associated;
A document attribute analysis step for analyzing the attribute of the document based on a result obtained by removing the exception attribute of the document analyzed by the exception attribute analysis step from the estimated attribute of the document analyzed by the estimated attribute analysis step ;
Receiving a document and action information for the document;
Policy determination is performed based on policy storage means in which the attribute of the document and action information for the document are associated with processing corresponding to the attribute of the document and action information for the document, and based on the result of the policy determination A policy determination step for performing processing according to the attribute of the document and action information for the document;
I have a,
The document analysis processing method, wherein the exception attribute storage means associates an incidental condition for handling an exception with a feature of the document to be handled as an exception and an exception attribute to be handled as an exception .

Images