JP2014021552A - Task document processing system and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To use the meta data of files for files and folders stored in a file server, and to achieve the presentation of a folder tree which is unnecessary for a user and the deletion of a folder with necessary files left.SOLUTION: Deletion possibility probability is calculated on the basis of past deleted file data, and as for files just under a file to be processed, whether or not a folder is deletable is determined from the deviation of the deletion possibility probability, and as for the folder the most of files of which can be determined to be unnecessary, only the necessary files are automatically excluded from the target of deletion, and the deletion possibility probability is calculated from the undermost layer of a folder tree to a route folder so that a deletable folder tree can be detected, and deletion candidate folder trees are displayed in the form of a list, and also the deletion candidate folder trees are displayed in a tree structure, and the presentation and confirmation of files other than the target of deletion is performed, and files other than the files excluded from the target of deletion are deleted in a batch from a file server.

Description

  The present invention relates to a business document processing system and a program, for example, a technique for managing a large amount of file data stored in a file server.

  In recent years, file servers that are highly convenient for storing and sharing electronic data have become widespread in organizations such as companies. However, the number of files in the file server tends to increase explosively, and there is an increasing demand for technology for supporting and promoting the deletion and archiving of unnecessary files. For example, digital data is explosively increasing (Non-patent Document 1), the structure data stored in the RDB (Relational Database) is only 20% of the whole in the company, most are document files, etc. It has been investigated and reported that it is an unstructured document (Non-patent Document 2).

  In addition, software (Non-Patent Documents 3 to 5) that supports deletion of unnecessary files has been proposed. With these technologies, a function that presents files that are considered unnecessary using conditions such as the last update date is old, the last access date is old, there are duplicate files, or there are files with similar file names have. Some of these files have an automatic archiving / deleting function, and others have an archiving / deleting function upon confirmation / approval of the user who uses the file.

J. F. Gantz, et al. The Diverse and Exploding Digital Universe, An IDC White Paper, http://japan.emc.com/collateral/analyst-reports/diverse-exploding-digital-universe.pdf S. Grimes, "Unstructured Data and the 80 Percent Rule", Clarabridge Bridgepoints, http://clarabridge.com/default.aspx?tabid=137&ModuleID=635&ArticleID=551 NEC, Document / Content Management Support System, InfoFrame PROCENTER / C, http://www.nec.co.jp/pfsoft/procenter/index.html Justsystem, GDMS, http://just-enterprise.com/product/gdms/ Hitachi Solutions, Ltd., MEANS file server slimming solution, http://hitachisoft.jp/products/means/slimserver/

  However, since the existing technique indicates deletion candidates in units of files, if the number of deletion candidate files is enormous, the number of steps for the user to check and delete becomes enormous. In general, files and folders stored in the same folder are related to each other, and files and folders of the same category are managed in a hierarchical structure. For this reason, if only some files are deleted, other files of the same category may remain even though they are unnecessary.

  Therefore, when deleting files and folders efficiently, it is most efficient to delete them in categories that are no longer necessary, that is, in folder tree units.

  However, in order to delete a file in folder tree units, the user must trace the folder tree and search for a folder tree that is no longer needed, and confirm whether or not a large number of folders can be deleted. There is a problem of not becoming. Furthermore, it is troublesome to check whether or not the necessary file or folder is included in the folder tree determined to be unnecessary, and to move the necessary file to another folder.

  The present invention has been made in view of such a situation, and provides a technique for efficiently realizing presentation of a folder tree and deletion of a folder unnecessary for a user regarding files and folders stored in a file server. To do.

  In order to solve the above-mentioned problem, the business document processing system according to the present invention executes a machine learning process on information of a deleted file, and calculates a deletion possibility probability of each processing target file. Calculate the bias of the calculated deletion probability of each file and folder directly under each processing folder and the processing target folder, and determine whether the processing target folder can be deleted based on the bias of the deletion probability Based on the analysis process, information on whether or not the folder can be deleted, and information on the parent-child relationship in each folder, a deletion candidate folder tree acquisition process for acquiring deletion candidate folder tree information indicating a folder tree that can be deleted, and deletion candidate folder tree information And a support information output process for outputting the support information.

  According to the present invention, it is possible to present a folder tree that is highly likely to be unnecessary for the user, and the user can efficiently delete unnecessary folders.

  Further features related to the present invention will become apparent from the description of the present specification and the accompanying drawings. The embodiments of the present invention can be achieved and realized by elements and combinations of various elements and the following detailed description and appended claims.

  It should be understood that the description herein is merely exemplary and is not intended to limit the scope of the claims or the application of the invention in any way.

It is a functional block diagram which shows the example of schematic structure of the business document processing apparatus (business document processing system) by embodiment of this invention. It is a figure which shows the data structure example of file information and folder information. It is a figure which shows the data structure of deleted file information, a deletion candidate folder tree, a deletion confirmation folder tree, a non-target folder tree, and a non-target file. It is a flowchart for demonstrating the whole outline | summary of the process performed in the business document processing apparatus. It is a flowchart for demonstrating the detail of a deletion file analysis process part (109). It is a flowchart for demonstrating the detail of a deletion folder analysis process part (110). It is a flowchart for demonstrating the detail of a deletion candidate folder tree registration process part (111). It is a flowchart for demonstrating the detail of a deletion candidate display process part (112). It is a figure which shows the example of a screen displayed by the deletion candidate display process part. It is a figure which shows the example of a screen displayed by the deletion candidate display process part. It is a figure which shows the example of a screen displayed by the deletion candidate display process part.

  DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. 1 to 11 are diagrams illustrating an embodiment of the present invention. In these drawings, parts denoted by the same reference numerals represent the same items, and the basic configuration and operation are the same. In addition, in embodiment of this invention, the apparatus, method, etc. which are used are examples, and of course, this invention is not limited to these.

  Furthermore, as will be described later, the embodiment of the present invention may be implemented by software running on a general-purpose computer, or may be implemented by dedicated hardware or a combination of software and hardware.

  In the following description, each information of the present invention will be described in a “table” format. However, the information does not necessarily have to be expressed in a data structure by a table, such as a data structure such as a list, a DB, a queue, or the like. It may be expressed as Therefore, “table”, “list”, “DB”, “queue”, etc. may be simply referred to as “information” to indicate that they do not depend on the data structure.

  In addition, when explaining the contents of each information, the expressions “identification information”, “identifier”, “name”, “name”, “ID” can be used, and these can be replaced with each other. It is.

  In the following, each process in the embodiment of the present invention will be described using “program” as a subject (operation subject). However, a program is executed by a processor and a process determined by a memory and a communication port (communication control device). Since it is performed while being used, the description may be made with the processor as the subject. Further, the processing disclosed with the program as the subject may be processing performed by a computer such as a management server or an information processing apparatus. Part or all of the program may be realized by dedicated hardware, or may be modularized. Various programs may be installed in each computer by a program distribution server or a storage medium.

<Configuration of business document processing apparatus>
FIG. 1 is a functional block diagram showing an internal schematic configuration of a business document processing apparatus (also referred to as a document processing system) 1 according to an embodiment of the present invention. In FIG. 1, the business document processing apparatus 1 is shown as a single computer, but is configured as a computer system in which each component is remotely arranged and connected via a network. You may make it.

  The business document processing apparatus 1 includes a file information DB 100, a folder information DB 101, a deleted file information DB 102, a central processing apparatus 103 that performs necessary arithmetic processing and control processing, and processing necessary for the central processing apparatus 103. A program memory 104 for storing a program, a data memory 105 for storing data necessary for processing in the central processing unit 103, a display device 106 for displaying data, and a menu for the displayed data are selected. And an input device 107 for performing operations such as these.

  The central processing unit 103 includes a machine learning processing unit 108, a deletion file analysis processing unit 109, a deletion folder analysis processing unit 110, a deletion candidate folder tree registration processing unit 111, and a deletion candidate display processing unit 112. Yes. In this embodiment, since it is configured by a computer, a machine learning processing unit 108, a deletion file analysis processing unit 109, a deletion folder analysis processing unit 110, a deletion candidate folder tree registration processing unit 111, and a deletion candidate display processing unit 112 is realized as a part of the function of the program executed on the computer. Therefore, each processing unit can be read as each processing program. Note that these programs are stored in the program memory 104.

  The data memory 105 includes file information 113, folder information 114, deleted file information 115, a deletion candidate folder tree 116, a deletion confirmation folder tree 117, and a non-deletion target file 118.

<Data structure>
Hereinafter, the data structure of each piece of information included in the data memory 105 will be described with reference to FIGS.

(I) FIG. 2 is a diagram showing the data structure of the file information 113 and the folder information 114 included in the data memory 105. For example, FIG. 2 shows information about files and folders on the file server.

  File information 113 includes file ID 200, file name 201, parent folder ID 202, extension 203, size 204, update date 205, creation date 206, access date 207, text 208, owner 209. The deletion possibility probability 210 is included as a configuration item, and for example, these pieces of information are held in the form of an array. The file ID 200 stores a value uniquely assigned to the file. The file name 201 indicates an identification name such as a file name. The parent folder information 202 is information indicating the absolute path of the folder in which the target file is stored. The extension 203 is information indicating the storage format of the target file. The size 204 is information indicating the capacity of the target file, and is held in units of bytes. The update date 205 is information at the time when the writing process for the target file was last performed. The creation date 206 is information at the time when the target file is created. The access date 207 is information at the time when the access process for the target file was last performed. A text 208 indicates information on the text when the target file is a file storing text information. The owner 209 is information indicating the ID of the user who owns the target file. The deletion possibility 210 is information indicating the probability that the target file can be deleted.

  The folder information 114 includes a folder ID 211, a folder name 212, a storage file ID 213, a storage folder ID 214, a hierarchy 215, a deletion probability 216, and a deletion flag 217 as configuration items. For example, in the form of an array I have this information. The folder ID 211 stores a value uniquely assigned to the folder. The folder name 212 indicates an identification name such as a folder name. The storage file ID 213 indicates the file ID of a file stored immediately below the target folder. A storage folder ID 214 indicates a folder ID of a folder stored immediately below the target folder. A hierarchy 215 indicates a hierarchy in which the target folder is located from the uppermost folder (root folder) in the file server. For example, when the item holds “4”, the folder belongs to a position that can be reached by following the child folder four times from the root folder. The deletion possibility probability 216 is information indicating the probability that the target folder can be deleted. The deletion flag 217 indicates whether the target folder can be deleted by “true” and “false”. In the initial stage, “false” indicating that the target folder can be deleted or is not known is not stored.

(Ii) FIG. 3 is a diagram showing the data structure of the deleted file information 115, the deletion candidate folder tree 116, the deletion confirmation folder tree 117, and the non-deletion target file 118 included in the data memory 105.

  The deleted file information 115 includes a file ID 300, a file name 301, a parent folder ID 302, an extension 303, a size 304, an update date 305, a creation date 306, an access date 307, a deletion date 308, The text 309 and the owner 310 are included as configuration items, and for example, information about files deleted in the file server in the past is held in an array format. The file ID 300 stores a value uniquely assigned to the file. The file name 301 indicates an identification name such as a file name. The parent folder ID 302 is information indicating the absolute path of the folder in which the target file is stored. The extension 303 is information indicating the storage format of the target file. The size 304 is information indicating the capacity of the target file, and is held in units of bytes. The update date 305 is information at the time when the writing process for the target file was last performed. The creation date 306 is information at the time when the target file is created. The access date 307 is information at the time when the access process for the target file was last performed. The deletion date 308 is information when the target file is deleted from the file server. The text 309 indicates information on the text when the target file is a file storing text information. The owner 310 is information indicating the ID of the user who owns the target file. These pieces of information are information used for machine learning in the analysis of the file deletion probability described later.

  The deletion candidate folder tree 116 includes a deletion candidate folder ID 311 as configuration information. The folder ID 311 holds the folder ID 211 of the folder information 114, which is the root folder of the folder tree that is a deletion candidate, in an array format.

  The deletion confirmation folder tree 117 includes a folder ID 312 as a deletion confirmation target as configuration information. The folder ID 312 holds the folder ID 211 of the folder information 114, which is the root folder of the folder tree to be deleted, in an array format.

  The non-deletion target file 118 has a folder ID 313 as configuration information. The folder ID 313 is determined to be necessary in the folder tree to be deleted, and the file ID 200 of the file information 113 to be deleted is stored in an array format.

<Outline of processing in business document processing apparatus>
Next, processing performed in the business document processing apparatus 1 of the present embodiment will be described. FIG. 4 is a flowchart schematically showing the overall processing flow performed in the business document processing apparatus 1. This flowchart shows a process of determining whether a folder and a folder tree can be deleted from the file server file deletion probability and displaying the result. Further, the main processing unit in the flowchart is a central processing unit (also simply referred to as a processor) 103.

  In FIG. 4, first, the central processing unit 103 reads the information of the file information DB 100 into the file information 113 and the information of the folder information DB 101 into the folder information 114 in order to obtain the file and folder information of the file server. (Step 400). The timing for reading may be the timing for starting the deletion candidate display processing or the timing for updating the file information or folder information. However, at this stage of reading, none of the values of the deletion possibility probability 210 in the file information 113 and the deletion possibility probability 216 in the folder information 114 are retained. Since the deletion possibility probability 210 and the deletion possibility probability 216 hold the probability that the target file and folder can be deleted, values are set in the analysis of the file deletion possibility and the folder deletion possibility described later.

  Next, the central processing unit 103 uses the deleted file analysis processing unit 109 to calculate the probability of deletion of each file on the file server (step 401). Details of this processing will be described later with reference to FIG.

  When the file deletion probability is obtained, the central processing unit 103 uses the deletion folder analysis processing unit 110 to determine whether or not the folder can be deleted based on the file deletion probability (step 402). Details of this processing will be described in detail with reference to FIG.

  Further, the central processing unit 103 acquires a deletion candidate folder tree from the analysis result of whether or not the folder can be deleted (step 403). Details of this processing will be described in detail with reference to FIG.

  Finally, the central processing unit 103 uses the deletion candidate display processing unit 112 to display the deletion candidate folder tree on the screen (step 404). A screen display example obtained as a result of the processing will be described in detail with reference to FIGS.

<Details of File Deletability Probability Analysis Processing (Step 401)>
FIG. 5 is a flowchart for explaining details of processing by the deleted file analysis processing unit 109 in step 401 of FIG. In this process, a classification model is generated by machine learning from file information deleted in the past, and a deletion possibility probability is calculated for each file to be processed. The processing is realized by the central processing unit 103 executing the deleted file analysis processing unit (deleted file analysis processing program) 109.

  First, the central processing unit 103 acquires file information deleted in the past from the deleted file information DB 102 and stores it in the deleted file information 113 in the data memory 105 (step 501).

  Then, the central processing unit 103 performs machine learning by setting arbitrary file information in the data memory 105 as “file that could not be deleted” and deleted file information as “file that could be deleted”, and whether or not the file can be deleted. A classification model is generated (step 501). In the machine learning process, for example, “file name”, “parent folder name”, “extension”, “size”, “text”, “owner”, and “deleted from creation date, update date, access date” A classification model based on a decision tree in which “Each elapsed time until day” is set as an attribute and “Files that can be deleted” and “Files that cannot be deleted” are set as attribute values is generated. In addition, a machine learning algorithm such as k-nearest neighbor method, logistic regression, naive bayes, or the like may be used.

  Next, the central processing unit 103 initializes an index variable file_idx for calculating a deletion probability for each file to be processed (step 502).

  Then, the central processing unit 103 checks whether the file information 113 includes data greater than or equal to file_idx (step 503). If there is no target data, the process ends.

  If there is a file greater than or equal to file_idx in step 503, the central processing unit 103 calculates the probability of deletion for the file_idx-th file using the classification model generated by machine learning, and the result is the probability of deletion in the file information of the file. 210 (step 504). For example, the average value of the metadata (for example, creation date) of the deleted file is obtained, and the probability of deletion is obtained so that the probability increases if the distance from the average value is close.

  Finally, the central processing unit 103 increments file_idx (step step 505) and repeats the processing from step 503.

  With the above processing, the probability of deletion for each file on the file server is calculated. It should be noted that deletion candidates are presented to the user (displayed on the screen), and the accuracy of the deletion probability increases as the number of files actually deleted by the user accumulates.

<Details of Folder Deletability Analysis Process (Step 402)>
Next, details of the processing by the deletion folder analysis processing unit 110 in step 402 of FIG. 4 will be described. This process determines whether or not a folder can be deleted based on the probability of file deletion. The probability of deletion of a folder is a numerical value between 0 and 100. Usually, determination of whether or not a folder can be deleted is based on these sets, and an average value or minimum value of the probability of deletion is obtained, and whether or not deletion is possible based on a threshold value. It is conceivable to make a determination. However, if there is a bias in the probability of deletion, it may not be possible to determine whether a folder can be deleted based on the average value. For example, assume that there is one file with a probability of deletion of 10% and 10 files with 90% in a certain folder. Then, the average value of the probability of deletion becomes 82%, and it is determined that the file can be deleted. However, it can be said that this folder cannot be deleted because there is a file with a probability of deletion of 10% in this folder. In addition, when the probability of deletion within a folder is set to the minimum value of the target folder, it can be detected by the influence of a file with a low probability of deletion even though the folder tree has a large number of files that can be deleted. There is a possibility of disappearing. Alternatively, when a threshold of the probability of deletion that can not be deleted is set and files and folders below the threshold are excluded, it is determined as a folder that cannot be deleted, for example, a folder that can be deleted with a uniform distribution of probability of deletion, In addition, many files remain as files that cannot be deleted. Then, while the original purpose is the purpose of deleting in the unnecessary category, that is, the folder tree, the files of unknown use remaining after the deletion are scattered.

  Therefore, the deletion folder analysis processing unit 110 corresponds to the biased folder having the probability of deletion as described above, and determines whether the folder can be deleted using the Gini coefficient and the median. Determine. First, a Gini coefficient indicating the degree of bias regarding whether or not a file and folder stored directly under the folder can be deleted is calculated, and it is checked whether or not the probability that the file and folder can be deleted is more than a threshold. For the threshold, a minimum probability that the file can be deleted is set in advance, and a value above the threshold is defined as “can be deleted”, and a value below the threshold is defined as “cannot be deleted”. When it is found from the Gini coefficient that the deletion possibility probability of the files and folders immediately under the folder is biased, the median value of the deletion possibility probability is obtained and confirmed to be equal to or greater than the threshold value. If the median is found to be equal to or greater than the threshold, it can be seen that the probability of deletion of files and folders is biased toward “can be deleted”. In addition, in a folder that is determined to be deleted, if there is a file with a relatively extremely low probability of deletion within the folder, the file is actually displayed as a deletion candidate and displayed to the user during the deletion candidate display process described later. Make sure that it can be deleted. In this way, by using the Gini coefficient and the median value for determining whether or not a folder can be deleted, the deleteability probability value described above is relatively extremely low in a folder or folder in which the deleteability probability value is distributed in both extremes. It is possible to determine whether or not a folder can be deleted without being affected.

  FIG. 6 is a flowchart for explaining the details of the processing by the deletion folder analysis processing unit 110 in step 402 of FIG. This process is realized by the central processing unit 103 executing the deletion folder analysis processing unit (deletion folder analysis processing program) 110.

  First, the central processing unit 103 initializes an index variable layer_idx for determining whether deletion is possible in order from the lowest folder to the root folder (step 600).

  The central processing unit 103 checks whether there is a folder hierarchy above the layer_idx layer (step 601). This is because it is not possible to calculate the deletion possibility probability of the upper level folder unless the deletion possibility probability is obtained from the folder of the lowest hierarchy. If there is no target data, the process ends.

  If there is a folder hierarchy above layer_idx in step 601, the central processing unit 103 initializes an index variable folder_idx for determining whether or not each folder in the folder hierarchy can be deleted (step 602).

  The central processing unit 103 checks whether there is data equal to or greater than folder_idx (step 603).

  If there is no target data, the central processing unit 103 decrements layer_idx in order to process the folder one level higher than the folder hierarchy because there is no folder for determining whether deletion is possible or not (step 611). The processing is repeated from step 601.

  On the other hand, if there is a file equal to or larger than folder_idx in step 603, the central processing unit 103 calculates the median value (step 604) and the Gini coefficient (step 605) in the probability of deletion of files and folders immediately under the folder_idxth folder. I do. Here, if there is only one file directly under the folder_idx-th folder, the Gini coefficient is set to 1. The folder deletion probability stores a median value regarding the deletion probability of files and folders directly under the folder.

  Then, the central processing unit 103 checks whether the Gini coefficient is equal to or greater than a threshold value (first threshold value) (step 606) and whether the median value is equal to or greater than a threshold value (second threshold value) (step 607). If both the Gini coefficient and the median are not greater than or equal to the threshold, the process ends.

  In step 606 and step 607, when both the Gini coefficient and the median value are equal to or greater than the corresponding threshold values, the central processing unit 103 determines that the difference between the median value and the erasure probability is the threshold value (the first value) immediately below the folder_idxth folder. The file IDs of the above files are registered in the non-deletion target file 118 in the data memory 105 (step 608). In this way, it is possible to remove a file that cannot be deleted from a folder including files that can be almost deleted as a whole, and to delete the folder. In addition, files that cannot be deleted and files that cannot be deleted, such as files that cannot be deleted and folders that cannot be deleted, are determined to be undeleteable by the Gini coefficient and the median value.

  Then, the central processing unit 103 sets the folder_idx-th folder deletion flag 217 to true, and recalculates the median value to determine the probability of deletion of the folder (step 609).

  Further, the central processing unit 103 increments folder_idx (step 610) and repeats the processing from step 603.

<Details of Delete Candidate Folder Tree Registration Process (Step 403)>
FIG. 7 is a flowchart for explaining details of the processing (step 403) by the deletion candidate folder tree registration processing unit 111 in the step of FIG. This processing is realized by the central processing unit 103 executing the deletion candidate folder tree registration processing unit (deletion candidate folder tree registration processing program) 111.

  In this process, the root folder of the folder tree to be deleted (deletion candidate folder tree) is acquired from the determination result of whether or not the folder can be deleted. Moreover, the range acquired as a folder tree includes a condition that the owner of all the files included in the folder tree is single. This is because when a user determines whether or not a folder tree can be deleted, the folder tree must be a folder managed only by a single user. When deleting a folder tree that stores a group of files owned by various users, it is often difficult for one user to determine whether or not deletion is possible. You must take the confirmation. Therefore, a folder tree is presented in which all the files have a single owner and the determination result of whether or not the folder can be deleted is deleteable. Some folders on the file server are shared by multiple users by storing files, but created in a controlled work flow such as quotations, acceptance notices, and checklists for internal controls Because it is a document file, it does not reach the large number of cases. Most of the other folders tend to be managed by a single user. For example, mail files, backup of source code and test data automatically generated during software development, and incidental collection in charge -Applicable to document files that describe created information. These files are managed and used for each user, and are stored according to the management policy of the user. For this reason, most of the folders in the file server store a group of files owned by a single user as a folder tree. In response to this trend, the process presents a folder tree in which only one user can determine whether or not a folder can be deleted. Therefore, the owner of all files is single and whether or not a folder can be deleted. A folder tree that can be deleted by the discrimination result is acquired.

  First, the central processing unit 103 initializes an index variable layer_idx for examining and acquiring the root folder of the deletion candidate folder tree in order from the root folder in the file server (step 700).

  The central processing unit 103 checks whether there is a folder hierarchy below the layer_idx layer (step 701). If there is no target data, that is, the index is one layer below the lowest layer in the file server, the process ends.

  If there is a folder hierarchy below layer_idx in step 701, the central processing unit 103 initializes an index variable folder_idx for checking whether there is a root folder of the deletion candidate folder tree for each folder in the folder hierarchy (step 702). ).

  Next, the central processing unit 103 checks whether there is a folder equal to or greater than folder_idx (step 703). If there is no target data, in order to process the next lower folder hierarchy, layer_idx is incremented (step 709), and the processing from step 701 is repeated.

  If there is a folder equal to or greater than folder_idx in step 703, the central processing unit 103 checks whether the owners of all files in the folder (from the folder to the end folder) are the same (step 704).

  When there are a plurality of owners, the central processing unit 103 determines that the folder tree having the folder as the root folder is a folder that cannot be determined by one user alone and cannot be deleted. Folder_idx is incremented (step 708), and the processing is repeated from step 703.

  If the owners are the same at step 704, the central processing unit 103 checks whether the deletion flag 217 of the folder is true (step 705).

  If the deletion flag 217 is false, the central processing unit 103 increments folder_idx to check the next folder without considering it as a deletion candidate because it is determined that the folder cannot be deleted (step 708). The processing is repeated from 703.

  If the deletion flag 217 is true in step 705, the central processing unit 103 checks whether the deletion flag 217 of the parent folder of the folder is false (step 706).

  When the deletion flag 217 of the parent folder is true, the folder is a folder in the folder hierarchy below the root folder in the deletion candidate folder tree, and the process includes the folder because the process is performed in order from the highest folder hierarchy. The root folder of the deletion candidate folder tree has already been acquired. Therefore, the central processing unit 103 increments folder_idx to skip the folder and examine the next folder (step 708), and repeats the processing from step 703.

  If the deletion flag 217 of the parent folder is false in step 706, the central processing unit 103 sets the folder ID of the folder as a deletion candidate in the data memory 105 because the folder is the root folder in the deletion candidate folder tree. Registration in the folder tree 116 (step 707).

  The central processing unit 103 increments folder_idx (step 708) and repeats the processing from step 703.

<Details of Delete Candidate Display Processing (Step 404)>
FIG. 8 is a flowchart for explaining details of the processing by the deletion candidate display processing unit 112 in step 404 of FIG. 4, and FIGS. 9, 10, and 11 are examples of screens displayed to the user in the course of the processing. It is. The processing is realized by the central processing unit 103 executing the deletion candidate display processing unit (deletion candidate display processing program) 112.

  In this process, in the deletion candidate folder tree 116 obtained from the process of step 403 in FIG. 4, a process is performed to display a folder tree having a file that is inferred that a part of the folder tree 116 cannot be deleted and a folder tree that is not. . By separating the folder trees having files inferred that they cannot be deleted, the user can determine whether or not the folder tree can be deleted with particular attention to these trees.

  First, the central processing unit 103 initializes an index delete_idx for acquiring one folder tree acquired as deletion candidates one by one (step 800).

  Next, the central processing unit 103 checks whether there is a delete_idxth folder tree (step 801).

  If it is determined in step 801 that there is a delete_idxth folder tree, the central processing unit 103 checks whether the delete_idxth folder tree stores a file included in the non-deletion target file 118 (step 802).

  If it is determined in step 802 that the delete_idxth folder tree stores a file included in the non-deleted file 118, the central processing unit 103 registers the delete_idxth folder tree in the deletion confirmation folder tree 117. Delete from the deletion candidate folder tree 116. If it is determined in step 802 that the delete_idxth folder tree does not store any file included in the non-deletion target file 118, the central processing unit 103 increments delete_idx (step 804). Repeat the process.

  On the other hand, if it is determined in step 801 that there is no delete_idx-th folder tree, the central processing unit 103 displays a deletion candidate folder tree 116 that is a folder tree inferred that all files can be deleted (step 805). The display process will be described later in detail with reference to FIG.

  The central processing unit 103 then waits for an instruction to end the screen operation by the user (step 806).

  When the central processing unit 103 receives an instruction to end the screen operation from the user's operation, the central processing unit 103 displays a deletion confirmation folder tree 117 as a deletion candidate folder tree including a file inferred that a part of the operation cannot be deleted (step 807). The display process will be described later in detail with reference to FIG.

  Further, the central processing unit 103 waits for an instruction to end the screen operation by the user (step 808). When receiving a screen operation end command from the user's operation, the process of the deletion candidate display processing unit 112 is ended.

<Delete candidate folder tree list display>
FIG. 9 is a diagram showing a screen example of a list display of the deletion candidate folder tree performed in step 805 of FIG.

  In the “deletion candidate folder tree list” in FIG. 9, first, the central processing unit 103 executes the deletion candidate display processing unit 112 to register folders registered in the deletion candidate folder tree 116 in the folder tree list 900 in FIG. 9. Displays the folder name.

  The central processing unit 103 adds the owner 209 of the folder tree to the owner list 901 and the total file size of the files to the file size total list 902 in the same row as the folder tree displayed in the folder tree list 900. The ratio of the file size to the total file size of all files in the file server is the file size ratio list 903, the total number of files is the total file number list 904, and the total number of files is the total number of all files in the file server. The ratio occupied by each is calculated and displayed in the file number ratio list 905. As a result, the user can confirm only the folder tree managed by himself / herself from the owner list 901, and efficiently reaches the folder tree that is likely to be deleted from a large number of folders. can do. In addition, in a company or the like, it is possible to display a folder tree that is owned by a retired person or a user who has moved to another department. It is possible to find and delete. Furthermore, since the deletion effect on the entire file server when each folder tree is deleted can be understood, the deletion efficiency can be estimated.

  When the radio button 906 is selected by the user, the central processing unit 103 causes the deletion candidate display processing unit 112 to hold the folder tree in the same row as the selected radio button as a selected state. Then, when the user presses “confirm selection tree” of the selection button 908, the central processing unit 103 causes the deletion candidate display processing 112 to execute “confirmation of folder tree” in FIG. Perform display processing with.

  When the user presses an end button 907, the central processing unit 103 receives a screen operation end command and closes this screen. When the user presses “Delete selected tree” on the delete button 909, the central processing unit 103 causes the deletion candidate display processing unit 112 to delete a file group excluding files not to be deleted in the selected folder tree. Delete from the file server. At this time, the file information 113 of the deleted file is additionally registered in the deleted file information DB. Since the deletion process can be realized by the existing technology (Non-Patent Document 5), the description is omitted.

<Delete confirmation folder tree list display>
FIG. 10 is a diagram showing a screen example of the deletion confirmation folder tree list display performed in step 807 of FIG.

  The “deletion confirmation folder tree list” in FIG. 10 has the same display form as the “deletion candidate folder tree list” in FIG. In this screen, the deletion candidate confirmation folder tree 117 is displayed. In addition to this display form, the central processing unit 103 causes the deletion candidate display processing unit 112 to calculate the total number of files corresponding to the non-deletion files 118 in the files and folders included in the displayed folder tree. The number of outside files is displayed in 1000. In this display, since the folder tree containing files that are inferred that some of them cannot be deleted is displayed, the number of files that cannot be deleted in each folder tree is displayed at the same time, so that the user can delete how many items in each folder tree. You can check whether a file that cannot be used is included. In particular, a folder tree with a large number of files that cannot be deleted tends to have related files stored nearby, so there may be other files that cannot be deleted. By performing the display, the user can determine whether the folder tree has priority and needs to be confirmed in detail from the number of files not to be deleted.

<Folder tree confirmation display>
FIG. 11 is a diagram showing an example of a folder tree confirmation display screen displayed when the “confirmation of selected tree” button in FIG. 10 is pressed.

  In “confirmation of folder tree” in FIG. 11, the central processing unit 103 uses the deletion candidate display processing unit 112 to display the folder tree in the selected state in the folder tree display area 1100 in a tree structure. The highest hierarchy follows the hierarchical structure of folders below the folder. In addition, the display form of each displayed folder differs depending on whether the file directly under the folder is included in the non-deletion target file 118. For example, a folder that stores files included in the non-deletion target file 118 is displayed in red on the tree. As a result, the user can easily identify in which folder on the folder tree the folder inferred that it cannot be deleted by machine learning, so that it is possible to efficiently confirm whether the file can be deleted.

  Each folder in the folder tree displayed in the folder tree display area 1100 can be selected. FIG. 11 shows a state in which the customer material folder 1105 is selected.

  When the user selects a folder, the central processing unit 103 displays a list of files stored in the folder in the file list 1102. However, files included in the non-deletion target file 118 in the data memory 105 are not displayed in the file list 1102 but are displayed in the non-deletion target file list 1104. The file deletion check boxes 1101 and 1103 display differently depending on whether the files in the file list 1102 and the non-deletion target file list 1104 in the same row correspond to the files stored in the non-deletion target file 118. In the case of a file stored in the non-deletable file 118, the check box corresponding to the file is displayed in black indicating that it has been checked, and is displayed in white indicating that it has not been checked if it is not included.

  When the user checks the check box, the central processing unit 103 registers the file ID of the file corresponding to the check box in the non-deletion target file 118. On the other hand, when the user unchecks the check box, the central processing unit 103 can exclude the file ID of the file corresponding to the check box from the non-deletion target file 118.

  The user can determine whether the folder tree can be deleted collectively from the displayed folder tree, file list, and non-deletion target file list, and can confirm necessary files. At this time, the user can grasp whether or not the folder tree can be deleted from the folder name displayed in the folder tree or a part of the file list without viewing the entire file list.

  When the user presses the return button 1106, the central processing unit 103 closes the screen and redisplays the screen in FIG. 9 or FIG.

<Modification>
(I) In the embodiment of the present invention, the business document processing apparatus that detects an unnecessary deletion candidate folder tree for deleting a file in folder tree units has been described. The business document processing apparatus calculates the probability of deletion based on the past deletion file data, and determines whether or not the folder can be deleted from the bias of the deletion possibility with respect to the files immediately under the folder. Also, only necessary files are automatically excluded from deletion targets by using a bias of probability of deletion from a folder tree determined that most files are unnecessary. This detects a folder tree that can be deleted by calculating the probability of deletion of files from the lowest layer of the folder tree to the root folder and analyzing the hierarchical structure of the folder using the bias of the probability of deletion. In addition, when a folder tree is detected, a condition that the owner of the stored file is single is added, so the owner of the detected folder tree is managed by a single user. Therefore, when a file server user performs deletion using the screen of this device, the folder tree managed by the user can be reached without following the hierarchy of the folder, and the folder tree can be deleted efficiently. It will be possible to confirm the availability. Further, it is not necessary to confirm whether or not the folder tree can be deleted across a plurality of users. In the screen display of this apparatus, the user can confirm the effect and estimate of the file size and the number of files related to the deletion of each folder tree by the list display for the folder tree detected as deletable. In addition, since the tree structure of the deletion candidate folder tree is displayed, the user can confirm the folder name displayed in the tree structure, the file list to be deleted, and the file list that cannot be deleted together on one screen, It is easy to confirm whether or not the folder tree can be deleted. In addition, in the tree structure display, folders that contain files that cannot be deleted are displayed in a different color to confirm the files that can be deleted by machine learning, and to find files that cannot be deleted that cannot be inferred. effective. This is because files with high relevance tend to be managed together in the same folder. For example, files in the same folder as files that cannot be deleted with correct inference results are highly related and cannot be deleted as well. There is a possibility. About this, by highlighting the display form of the folder reflecting the inference result by changing the color or the like, it is possible to easily find the folder scattered in the file server. In addition, regarding the deletion candidate folder tree that can be deleted as a result of the confirmation, it is possible to delete all the folders that are not subject to deletion, and move the files that are not subject to deletion to a different storage location in advance. Can be omitted. In this way, the business document processing apparatus enables the user to efficiently delete files in folder tree units.

(Ii) Although the present embodiment has been described as an apparatus for detecting an unnecessary folder tree, files can be archived in units of folder trees, or files can be copied to file servers with low backup frequency in units of folder trees. It can also be applied to uses such as moving. For example, machine learning is performed on file information that is used infrequently, so that a folder tree that is used infrequently is detected, and a group of files in the folder tree is archived into a single file, or copied to a file server that is not frequently used for backup. Or move it.

  Further, in this specification, the processing for obtaining the probability of deletion in a file on the file server in the processing of step 401 in FIG. 4 has been described. As an additional function, the probability of deletion of all files and folders may be normalized at the end of step 401. In this case, it is effective when a value with a low overall probability of deleting files and folders obtained by machine learning is obtained. For example, if the maximum value of the probability of deletion that has been determined is 50%, all deletion candidate probabilities are multiplied by a coefficient that sets 50% to 100%. Then, even if the probability of deletion is low, it is possible to determine that deletion is possible if the probability is high when viewed relative to all files and folders.

(Iii) In the present embodiment, the processing of excluding files and folders that may not be deleted in the processing of step 402 in FIG. 4 has been described. Instead of deleting files whose deleteability probability difference from the median of the process is greater than or equal to the threshold value, files and folders with close deleteability probabilities are clustered to target groups with low deleteability probabilities. A method of making an external file and a non-target folder is also conceivable. In this case, it can be realized by using a clustering method such as a shortest distance method, a longest distance method, a group average method, a k-means method, or a Ward method as a method of dividing into groups.

(Iv) In the present embodiment, the process of determining whether each folder can be deleted in the process of step 402 in FIG. 4 has been described. Instead of the process of determining whether a folder can be deleted using the Gini coefficient of the process, an average value and a variance value of the deletion probability may be used. For example, when the average value is high and the variance value is low, it can be said that there is a high possibility that the folder can be deleted. In addition, when the average value is high and the variance value is high, it is highly likely that a file or folder with a low probability of deletion is included. Therefore, by setting a threshold, files and folders with a low probability of deletion are removed. Is considered possible.

(V) In the present embodiment, the process of determining whether or not each folder can be deleted in the process of step 402 in FIG. 4 has been described. Instead of processing to determine whether a folder can be deleted using the Gini coefficient of the processing, it is determined whether the folder can be deleted by taking the average of the absolute value of the difference of 50% with respect to the probability of deleting a file immediately under the folder It is also possible to do. For example, if there is one file with a 10% deletion probability, and 10 files with a 90% deletion probability, the value of the processing result is 40. The range of the value of the processing is 0 to 50, and the higher the value, the more likely the deletion possibility in the folder is biased toward 0%, 100%, or both sides. It can be seen that the concentration is in the vicinity of 50%. That is, if the processing result is close to 50 and the reference value of the probability of deletion is greater than or equal to the threshold value, it is determined that the folder can be deleted, and a small number of files that cannot be deleted can be excluded from deletion.

(Vi) In the present embodiment, the process of displaying a list of deletion candidate folder trees in FIG. 9 has been described. As an additional function, the files may be displayed after being sorted in descending order. This is because the number of files directly under the folder is uneven, and a small number of folders include a large number of files. That is, also in the folder tree, a folder tree including a folder having such a large number of files means that a large number of files are included in all files. Therefore, by displaying the folder tree at the top in the list display, the user can confirm the deletion in order of the folder tree with the highest deletion efficiency, and can estimate the work cost that can reach the target number of deleted files, for example. .

(Vii) In the present embodiment, the process of deleting files other than the non-deletion files in the deletion candidate folder tree from the file server when the delete button 909 is pressed in FIG. 9 has been described. As an additional function, when a file is deleted from the file server, the file not to be deleted may be moved to a designated folder, and the folder may be deleted together with the file to be deleted.

<Summary>
(I) In the present invention, machine learning processing is executed on the information of the deleted file to calculate each deletion possibility probability of the processing target file (file deletion possibility calculation processing). Further, the bias of the calculated deletion probability of each file and folder immediately below each of the processing target folders is calculated, and whether or not the processing target folder can be deleted is determined based on the bias of the deletion probability (folder deletion permission / non-permission). Analytical processing). Furthermore, deletion candidate folder tree information indicating a folder tree that can be deleted is acquired based on information on whether or not the folder can be deleted and parent-child information on each folder (deletion candidate folder tree acquisition process). Then, the deletion candidate folder tree information is output (displayed, printed, etc.) as support information (support information output process). By doing so, only the information on the files and folders that can be deleted is displayed, so that it is possible to dramatically reduce the man-hours required for the user's deletion permission confirmation work. It should be noted that by repeatedly executing this series of processing, the accuracy of the file deletion possibility is improved, and the information on the folder and file that can be finally deleted is also more accurate. Further, when obtaining the bias of the probability of deletion in the folder deletion possibility analysis process, for example, the median value of the deletion possibility and the Gini coefficient are used, or the absolute value of the difference between each of the deletion probability and the probability of 50% It is possible to use the average of. The file information includes at least a plurality of types of meta information among the file name of the file on the file server, the folder name of the parent folder, the file size, the extension, the text content, the creation date, the update date, the access date, and the owner information. Information can be used. Further, as the folder information, the folder name of the folder in the file server, the information of the file included in the folder, the information of the folder included in the folder, and the position information of the hierarchy in the folder tree of the file server are used. Can do. In addition, deleted file information includes the file names of multiple files deleted in the past, the folder name of the parent folder, file size, extension, text content, creation date, update date, access date, owner, and deletion At least a plurality of types of meta information can be used among the day information.

  Further, for each folder, it may be determined whether or not the folder stores only files having the same owner. In this case, a folder including only files determined to be the same owner by the folder owner determination is set as a candidate for deletion candidate folder tree acquisition processing. In this way, it is possible to provide information that allows one user to easily determine whether or not deletion is possible.

  Further, in the present invention, a folder tree that can be deleted is detected by performing in order from the lowest folder in the file server to the root folder. By doing so, it is possible to determine whether or not each folder tree can be deleted as well as a single folder and present a tree that can be deleted, thereby reducing the work burden on the user. As for detection of a deleteable folder tree, clustering may be performed based on the probability of deletion of files and folders directly under the processing target folder to detect the deleteable folder tree.

  In the support information output process, the root folder in the deleteable folder tree may be displayed as a list. At this time, the total file size of the file group in the deleteable folder tree and the ratio of the total file size to the entire file server, the total number of files, and the ratio of the total number of files to the entire file server are displayed in the list display. Anyway. Alternatively, the list display may be sorted in descending order with respect to the total file size or the total file. By providing such a display, the user can more easily execute a deletion / non-permission confirmation operation for a file or a folder. In addition, it is possible to search for and delete the folder tree that can be deleted, and furthermore, it is possible to estimate the deletion efficiency because the deletion effect on the entire file server when each folder tree is deleted can be understood.

  In addition, in the support information output process, a folder tree that can be deleted is displayed, and in response to a folder selection action from the displayed folder tree, a file included in the selected folder and a file that is not to be deleted May be displayed separately. Further, a folder including a file that is not to be deleted in the displayed folder tree may be displayed separately from other folders (for example, color-coded display). In this way, it is possible to make it easier to determine whether or not a user can be deleted.

(Ii) The present invention can also be realized by software program codes that implement the functions of the embodiments. In this case, a storage medium in which the program code is recorded is provided to the system or apparatus, and the computer (or CPU or MPU) of the system or apparatus reads the program code stored in the storage medium. In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiments, and the program code itself and the storage medium storing the program code constitute the present invention. As a storage medium for supplying such program code, for example, a flexible disk, CD-ROM, DVD-ROM, hard disk, optical disk, magneto-optical disk, CD-R, magnetic tape, nonvolatile memory card, ROM Etc. are used.

  Also, based on the instruction of the program code, an OS (operating system) running on the computer performs part or all of the actual processing, and the functions of the above-described embodiments are realized by the processing. May be. Further, after the program code read from the storage medium is written in the memory on the computer, the computer CPU or the like performs part or all of the actual processing based on the instruction of the program code. Thus, the functions of the above-described embodiments may be realized.

  Further, by distributing the program code of the software that realizes the functions of the embodiment via a network, it is stored in a storage means such as a hard disk or memory of a system or apparatus, or a storage medium such as a CD-RW or CD-R And the computer (or CPU or MPU) of the system or apparatus may read and execute the program code stored in the storage means or the storage medium when used.

  Finally, it should be understood that the processes and techniques described herein are not inherently related to any particular apparatus, and can be implemented by any suitable combination of components. In addition, various types of devices for general purpose can be used in accordance with the teachings described herein. It may be beneficial to build a dedicated device to perform the method steps described herein. Various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiments. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined. Although the present invention has been described with reference to specific examples, these are in all respects illustrative rather than restrictive. Those skilled in the art will appreciate that there are numerous combinations of hardware, software, and firmware that are suitable for implementing the present invention. For example, the described software can be implemented in a wide range of programs or script languages such as assembler, C / C ++, perl, shell, PHP, Java (registered trademark).

  Furthermore, in the above-described embodiment, control lines and information lines are those that are considered necessary for explanation, and not all control lines and information lines on the product are necessarily shown. All the components may be connected to each other.

  In addition, other implementations of the invention will be apparent to those skilled in the art from consideration of the specification and embodiments of the invention disclosed herein. Various aspects and / or components of the described embodiments can be used alone or in any combination in a computer system capable of managing data. The specification and specific examples are merely exemplary, and the scope and spirit of the invention are indicated in the following claims.

100 ... file information DB
101 ... Folder information DB
102 ... Deleted file information DB
103 ... Central processing unit 104 ... Program memory 105 ... Data memory 106 ... Display device 107 ... Input device 108 ... Machine learning processing unit 109 ... Deleted file analysis processing unit 110 Deletion folder analysis processing unit 111 ... Deletion candidate folder tree registration processing unit 112 ... Deletion candidate display processing unit 113 ... File information 114 ... Folder information 115 ... Deleted file information 116 ... Deletion candidate folder tree 117 ... Deletion confirmation folder tree 118 ... Non-deletable file

Claims (15)

A business document processing system that provides support information for organizing a plurality of files stored in a file server,
A storage device for storing information on the processing target file and information on the processing target folder, and information on the deleted file;
A processor for generating the support information and providing the generated support information,
The processor is
A file deletion probability calculation process for performing a machine learning process on the information of the deleted file and calculating a deletion possibility probability of each of the processing target files;
Folder deletion for calculating whether or not deletion is possible for the processing target folder based on the bias of the deletion possibility calculated by calculating the bias of the calculated deletion possibility of each file and folder immediately below each of the processing target folders Availability analysis processing,
A deletion candidate folder tree acquisition process for acquiring deletion candidate folder tree information indicating a folder tree that can be deleted based on the information on whether or not the folder can be deleted and parent-child relationship information in each folder;
Support information output processing for outputting the deletion candidate folder tree information as the support information;
A business document processing system characterized by executing In claim 1,
The processor reflects the information of the file actually deleted by the user based on the support information in the information of the deleted file, the file deletion probability calculation process, the folder deletion possibility analysis process, and the deletion candidate folder A business document processing system characterized in that tree acquisition processing is executed again. In claim 1,
The processor further executes, for each folder, a folder owner determination process for determining whether or not the folder stores only files having the same owner.
The business document processing system, wherein the processor includes a folder including only files determined to be the same owner by the folder owner determination process as a target of the deletion candidate folder tree acquisition process. In claim 2,
The processing target file information includes at least a plurality of types of file name, file name of parent folder, file size, extension, text content, creation date, update date, access date, and owner information in the file server. Including meta information,
The information of the processing target folder includes the folder name of the folder in the file server, the information of the file included in the folder, the information of the folder included in the folder, and the position information of the hierarchy in the folder tree of the file server. ,
The deleted file information includes file names of a plurality of previously deleted files, folder names of parent folders, file sizes, extensions, text contents, creation dates, update dates, access dates, owners, and deletion dates. Including at least multiple types of meta information
The processor performs the machine learning process using the meta information of the deleted file and the meta information of the file that the user has determined to be undeleteable in the file deletion probability calculation process. A business document processing system characterized by that. In claim 1,
The processor calculates a median and a Gini coefficient of the calculated probability of deletion of each file and folder immediately under each of the processing target folders in the folder deletion possibility analysis process, and the median and the Gini coefficient A business document processing system characterized in that a bias of deletion probability of files and folders in the processing target folder is calculated using, and whether or not the processing target folder can be deleted is determined based on the bias. In claim 2,
In the folder deletion feasibility analysis process, the processor determines whether the file and folder from a processing target folder in which the Gini coefficient of the file and folder is equal to or greater than a first threshold value and the median value of the file and folder is equal to or greater than a second threshold value. The file or folder whose difference between the median value of the file and the folder and the folder can be deleted is equal to or greater than a third threshold, the median value of the file and folder is calculated again for the processing target folder, and the processing target folder can be deleted. A business document processing system characterized by probability. In claim 1,
In the folder deletion feasibility analysis process, the processor uses an average of absolute values of differences between the calculated deletion probability and a probability of 50% to determine the deletion probability of files and folders in the processing target folder. A business document processing system characterized by obtaining a bias and determining whether or not the processing target folder can be deleted based on the bias. In claim 1,
The business document processing system, wherein the processor detects the folder tree that can be deleted by performing the folder deletion possibility analysis processing in the order of the lowest folder in the file server in order from the root folder. In claim 1,
The business document processing system, wherein the processor performs clustering based on a deletion probability of files and folders directly under the processing target folder and detects a deleteable folder tree. In claim 8,
The business document processing system, wherein the processor displays a list of root folders in the deleteable folder tree in the support information output processing. In claim 10,
The processor includes: a total file size of a file group in the folder tree that can be deleted in the list display; a ratio of the total file size to the entire file server; a total number of files; and a ratio of the total number of files to the entire file server Business document processing system characterized by displaying In claim 10,
The business document processing system, wherein the processor sorts the list display in descending order with respect to a total file size or a total file. In claim 8,
In the support information output process, the processor displays the folder tree that can be deleted, and responds to a folder selection action from the displayed folder tree and excludes the file included in the selected folder and the deletion target. A business document processing system that distinguishes and displays different files. In claim 13,
The business document processing system, wherein the processor displays a folder including a file that is excluded from the deletion target in the displayed folder tree in distinction from other folders. A program for operating a computer system as a business document processing system that provides support information for organizing a plurality of files stored in a file server,
The computer system includes a processor, a storage device that stores information on a processing target file and information on a processing target folder, and information on a deleted file.
The program is stored in the processor.
A file deletion probability calculation process for performing a machine learning process on the information of the deleted file and calculating a deletion possibility probability of each of the processing target files;
Folder deletion for calculating whether or not deletion is possible for the processing target folder based on the bias of the deletion possibility calculated by calculating the bias of the calculated deletion possibility of each file and folder immediately below each of the processing target folders Availability analysis processing,
A deletion candidate folder tree acquisition process for acquiring deletion candidate folder tree information indicating a folder tree that can be deleted based on the information on whether or not the folder can be deleted and parent-child relationship information in each folder;
Support information output processing for outputting the deletion candidate folder tree information as the support information;
A program characterized by executing

Images