JP2006099478A - Document classification device and document classification method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a document classification device capable of interactively adjusting a classification result in a document classification system using batch processing. <P>SOLUTION: This device comprises a comparison object document information storage means storing comparison object document information obtained by relating information of a comparative object document with the field of the comparison object document; a word weight information storage part storing words and the word weights thereof; a batch processing control part comparing the comparison object information with a classification key document to extract a commonly used word which is commonly used in the classification key document and the comparison object document, and generating common word information obtained by relating the commonly used word with the use frequency of the commonly used word and the word weight of the commonly used word read from the word weight information storage part; and an interactive processing control part determining the similarities of a plurality of comparison object documents with the classification key document from the common word information, specifying a field based on a comparison object document with high similarity determined, and adjusting the specified field based on an instruction from an input device. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a document classification apparatus and a document classification method for classifying document fields.

  Conventionally, there is a document classification system that identifies a field to which an input document belongs based on a plurality of document information stored in advance in a database. In such a document classification system, first, a comparison target document similar to an input document whose classification is to be specified is extracted from a plurality of comparison target documents whose fields are specified in advance and stored in the database. Thereafter, there is a method for specifying a field to which an input document belongs based on a field previously associated with the extracted comparison target document (for example, Patent Document 1).

Further, in order to increase the efficiency of the classification process, a document classification system generally performs a classification process for a large number of input documents at once. Also, in such a document classification system, the input document is often calculated for similarity with a large number of comparison target documents. Furthermore, in a document classification system that requires high accuracy, the result of classification by a computer is manually checked.
JP 2001-1555025 A

  In the conventional document classification system that classifies a large amount of input documents as described above, a large amount of documents can be processed efficiently at one time, but the classification result is fixed when the classification processing is performed collectively. The As a result, the problem of interactivity may arise.

  Specifically, when the operator confirms the result obtained by the classification process, if the result is not the target result, the interactive operation for adjusting the parameters again in the classification process and executing the classification again cannot be realized.

  In addition, since it is difficult to implement interactive operations in a document classification system that classifies using the batch processing method, a system that presents the user with information that provides the basis for classification such as characteristic words that have greatly influenced the classification result. Also did not exist.

  The present invention has been made to solve the above problems, and provides a document classification apparatus and a document classification method capable of interactively adjusting a classification result in a document classification method using batch processing. With the goal.

  According to the document classification device according to the feature of the present invention, a document classification device for classifying a classification key document to be classified into a field to which the document belongs, information on a comparison target document to be compared with the classification key document, A comparison target document information storage unit storing comparison target document information associated with a field of the comparison target document, a word, and a word weight serving as an index indicating characteristics of the field of the document including the word The word weight information storage section, the classification key document is compared with the comparison target document information, and a common use word that is a word commonly used in the classification key document and the comparison target document is extracted, and at least these common use A batch processing control unit that generates common word information in which a word, the number of times of use of a commonly used word, and the word weight of the commonly used word read from the word weight information storage unit are associated The similarity between a plurality of comparison target documents and the classification key document is obtained from the common word information, the field is specified based on the obtained comparison target document having a high similarity, and further, based on an instruction from the input device And a dialogue processing control unit for adjusting the specified field.

  According to the present invention having the above-described configuration, it is possible to provide a document classification device that interactively adjusts classification results.

  According to the present invention, it is possible to interactively adjust the classification result in the document classification method using batch processing.

[First embodiment]
The similar document classification device 1 according to the first exemplary embodiment of the present invention will be described below with reference to the drawings.

[Similar document classification device]
FIG. 1 is a block diagram of a similar document classification apparatus 1 according to the first embodiment of the present invention.

  The similar document classification device 1 shown in FIG. 1 includes a comparison target document information storage unit 11, a word weight information storage unit 12, a batch processing control unit 13, and a dialogue processing control unit 14.

  The comparison target document information storage unit 11 stores information on a comparison target document to be compared with a classification key document that is a target for classifying the field to which the document belongs, and comparison target document information associated with the field of the comparison target document. Yes.

  The word weight information storage unit 12 stores a word and a word weight serving as an index indicating characteristics of the field of the document including the word.

  The collective processing control unit 13 compares the classification key document with the comparison target document information, extracts common usage words that are commonly used in the classification key document and the comparison target document, and at least these common usage words And common word information in which the number of times of use of the commonly used word is associated with the word weight of the commonly used word read from the word weight information database.

  The dialogue processing control unit 14 obtains a similarity with a plurality of comparison target documents from the common word information, obtains a classification result specifying the field based on the field of the comparison target document having a high similarity, and further classifies the classification. Adjustment means for adjusting the result is provided.

  FIG. 2 shows an example of comparison target document information stored in the comparison target document information storage unit 11. The comparison target document is a document to be compared with a classification key document that is a target for classifying a field. The comparison target document information is generated based on the comparison target document.

  Specifically, the comparison target document information shown in FIG. 2 includes information such as “title”, “field”, “use word”, and “use count” of a plurality of comparison target documents. According to the example shown in FIG. 2, the title of the comparison target document 1 is “reduction in database update processing time”, and the field is “database update”. Further, the words used in the comparison target document 1 and the number of times of use thereof are 2 for “large scale”, 5 for “database”, 8 for “update processing”, and 3 for “time”. Times, “shortening” represents 2 times.

  FIG. 3 shows an example of word weight information stored in the word weight information storage unit 12. In the word weight information shown in FIG. 3, for example, the word weight of “automatic classification” is “8.5”, and the word weight of “database” is “4.3”.

  For this “word weight”, for example, the reciprocal of the frequency of use of the used word in all the comparison target documents in the comparison target document information storage unit 11 is used. This is because a more frequently used word is a general word and is considered to be a word that does not represent the characteristics of the document. Conversely, words that are used less frequently are considered more characteristic. In this embodiment, it is assumed that word weight information as shown in FIG. 3 is created and stored in advance as word weight information used for classification.

  As shown in FIG. 4, the similar document classification device 1 according to the preferred embodiment of the present invention includes a central processing control device 101, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, and an input / output interface 109. Are connected via the bus 110. An input device 104, a display device 105, a communication control device 106, a storage device 107, and a removable disk 108 are connected to the input / output interface 109.

  The central processing control device 101 reads and executes a boot program for starting the similar document classification device 1 from the ROM 102 based on an input signal from the input device 104, and further reads an operating system stored in the storage device 107. Further, the central processing control device 101 controls various devices based on input signals from the input device 104, the communication control device 106, etc., and reads programs and data stored in the RAM 103, the storage device 107, etc., into the RAM 103. A processing device that loads and implements a series of processing described later, such as data calculation or processing, based on a program command read from the RAM 103.

  The input device 104 includes input devices such as a keyboard and a mouse through which an operator inputs various operations. The input device 104 generates an input signal based on the operation of the operator, and inputs via the input / output interface 109 and the bus 110. It is transmitted to the central processing control apparatus 101. The display device 105 is a CRT (Cathode Ray Tube) display, a liquid crystal display, or the like. The display device 105 receives an output signal to be displayed on the display device 105 from the central processing control device 101 via the bus 110 and the input / output interface 109. This is a device that displays the processing results of the processing control device 101. The communication control device 106 is a device such as a LAN card or a modem, and is a device that connects the similar document classification device 1 to a communication network such as the Internet or a LAN. Data transmitted / received to / from the communication network via the communication control device 106 is transmitted / received to / from the central processing control device 101 via the input / output interface 109 and the bus 110 as an input signal or an output signal.

  The storage device 107 is a semiconductor storage device, a magnetic disk device, or the like, and stores programs and data executed by the central processing control device 101. The removable disk 108 is an optical disk or a flexible disk, and signals read / written by the disk drive are transmitted / received to / from the central processing control apparatus 101 via the input / output interface 109 and the bus 110. The storage device 107 of the similar document classification device 1 according to the embodiment of the present invention stores a similar document classification program and a comparison target document information storage unit 11 and a word weight information storage unit 12. Also, the similar document classification program is read and executed by the central processing control device 101 of the similar document classification device 1, whereby the batch processing control unit 13 and the interactive processing control unit 14 are implemented.

  The similar document classification device 1 according to the best embodiment of the present invention may be realized by a single computer or a plurality of computers that can communicate with each other. For example, the configuration for performing batch processing and the configuration for performing interactive processing may be on the same computer system or on different computer systems connected via a network or the like. . Further, the batch processing control unit 13 and the dialogue processing control unit 14 may each be realized by one computer, or may be realized by a plurality of computers.

  As shown in FIG. 5, the batch processing control unit 13 in the similar document classification device 1 according to the embodiment of the present invention includes a control unit 200 and a memory unit 250.

  The control unit 200 includes an initialization unit 201, an input unit 202, a word weight reading unit 203, a classification key document information generation unit 204, a comparison target document information reading unit 205, a common word information generation unit 206, and a data output unit 207.

  The memory unit 250 includes a word weight information buffer unit 251, a classification key document information buffer unit 252, a comparison target document information buffer unit 253, and a common word information buffer unit 254.

  The word weight information buffer unit 251 stores word weight information in which each used word and its weight are associated with each other for a used word that is a word used in the comparison target document.

  The classification key document information buffer unit 252 stores classification key document information generated from a classification key document to be classified.

  The comparison target document information buffer unit 253 stores comparison target document information to be compared with the classification key document information.

  The common word information buffer unit 254 includes a common use word that is a word commonly used in the classification key document and the comparison target document, the number of times the common use word is used in the document, and the word weight of the common use word. The common word information associated with is stored.

  The initialization unit 201 initializes the buffer units 251 to 254 of the memory unit 250.

  The input unit 202 inputs a classification key document and an operation instruction from the input device.

  The word weight reading unit 203 reads word weight information from the word weight information storage unit 12 into the word weight information buffer unit 251.

  The classification key document information generation unit 204 decomposes the input classification key document into words. Further, the classification key document information generation unit 204 generates classification key document information including each decomposed word and the number of times the word is used, and stores it in the classification key document information buffer unit 252.

  The comparison target document information reading unit 205 reads the comparison target document information from the comparison target document information storage unit 11 into the comparison target document information buffer unit 253.

  The common word information generation unit 206 reads out the classification key document information stored in the classification key document information buffer unit 252 and the comparison target document information stored in the comparison target document information buffer unit 253. After that, the common word information generation unit 206 extracts a common word that is used in common in the classification key document and the comparison target document, and uses the common usage word in the document and the word weight of the common word. Is generated and stored in the common word information buffer unit 254.

  The data output unit 207 outputs the common word information stored in the common word information buffer unit 254 to the dialogue processing control unit 14.

  As shown in FIG. 6, the dialogue processing control unit 14 in the similar document classification device according to the embodiment of the present invention includes a control unit 300 and a memory unit 350.

  The control unit 300 includes an initialization unit 301, an input unit 302, a data input unit 303, a comparison target document similarity calculation unit 304, a field-specific similarity accumulation unit 305, a field identification unit 306, and a word weight adjustment unit 307.

  The memory unit 350 includes a common word information buffer unit 351, a comparison target document similarity buffer unit 352, and a field-specific similarity integrated value buffer unit 353.

  The common word information buffer unit 351 stores common word information.

  The comparison target document similarity buffer unit 352 stores, as the comparison target document similarity, the similarity using the commonly used word for each comparison target document obtained for the classification key document.

  The field-specific similarity integrated value buffer unit 353 stores field-specific similarity integrated values obtained by adding the comparison target document similarities for each field to which the comparison target documents belong.

  The initialization unit 301 initializes the buffer units 351 to 353 of the memory unit 350.

  The input unit 302 inputs an operation instruction from the input device.

  The data input unit 303 inputs the common word information output from the data output unit 207 of the batch processing control unit 13 and stores it in the common word information buffer unit 351.

  The comparison target document similarity calculation unit 304 reads the common word information stored in the common word information buffer unit 351, and calculates the comparison target document similarity that is the similarity for each comparison target document of each common word information. In addition, the comparison target document similarity calculation unit 304 stores the calculated comparison target document similarity in the comparison target document similarity buffer unit 352.

  In this embodiment, the method for calculating the comparison target document similarity is obtained by multiplying the sum of the number of used words commonly used in the two documents of the classification key document and the comparison target document by multiplying the word weight. A description will be given using an example in which the degree of similarity is used. The method for calculating the similarity is not limited to the above method, and may be obtained by another calculation method.

  The field-specific similarity accumulation unit 305 reads the comparison object document similarity stored in the comparison object document similarity buffer unit 352 and accumulates the comparison object document similarity for each field. This integrated value is a field-specific similarity integrated value. The field-specific similarity accumulation unit 305 stores the field-specific similarity accumulation value obtained for each field in the field-specific similarity accumulation value buffer unit 353.

  The field specifying unit 306 associates the integrated value of the similarity stored in the field-specific similarity integrated value buffer unit 353 with the similarity calculation result stored in the comparison target document similarity buffer unit 352 that is the origin. In addition, the data is output to an output device such as a display device connected to the similar document classification device.

  The word weight adjustment unit 307 causes the comparison target document similarity buffer unit 352 to store a value input by the user for adjusting the word weight from an input device such as a connected keyboard.

  Next, processing of the similar document classification device 1 according to the embodiment of the present invention will be described with reference to FIGS. 7 and 8. FIG. 7 is a flowchart for explaining processing in the batch processing control unit 13. FIG. 8 is a flowchart for explaining processing in the dialogue processing control unit 14.

  In this embodiment, the classification process is divided into two processes, a batch process that is a first process and an interactive process that is a second process.

[batch processing]
The batch process is a process for comparing the classification key document that is the target of classifying the field with a plurality of pieces of comparison target document information stored in the comparison target document information storage unit 11.

  First, as shown in the flowchart of FIG. 7, the initialization unit 201 initializes the buffer units 251 to 254 of the memory unit 250 (S001). Thereafter, the word weight reading unit 203 reads word weight information from the word weight information storage unit 12 into the word weight information buffer unit 251 (S002).

  Subsequently, when the classification key document is input, the classification key document information generation unit 204 decomposes the classification key document into words. The classification key document information generating unit 204 generates classification key document information including each decomposed word and the number of times each word is used, and stores the generated classification key document information in the classification key document information buffer unit 252 ( S003).

  FIG. 9 shows an example of an input classification key document. FIG. 9 shows the classification key document 1, and the latter half is omitted. In this way, a plurality of classification key documents are input.

  FIG. 10 shows an example of classification key document information generated from “classification key document 1”. For example, in the classification key document information, “used word” used in the classification key document is associated with “use count” that is the number of times the used word is used in the target classification key document. Information.

  The process of step S003 is performed on all target classification key documents (S004). For example, when 2000 documents are input as the classification key document, the process of step S003 is repeated 2000 times.

  Next, the comparison target document information reading unit 205 reads the comparison target document information from the comparison target document information storage unit 11 into the comparison target document information buffer unit 253 (S005).

  Subsequently, the common word information generation unit 206 extracts words commonly used in the classification key document and the comparison target document, and uses the extracted common use words in the classification key document and the comparison target document. The common word information is generated together with the total number of times of the search, and stored in the common word information buffer unit 254 for each search target document (S006).

  FIG. 11 shows an example of common word information. In this common word information, for each combination of a classification key document and a comparison target document, the “field name” and “use word” of the comparison target document and the use word are used in the classification key document and the comparison target document. The “number of uses” and the “word weight” of the used word are stored in association with each other. For example, in the example shown in FIG. 11, when the classification key document 1 is compared with a document whose title is “comparison target document 1” and whose title is “reduction in database update processing time”, the number of uses of the word “Large Scale” is five. , The usage count of the word “database” is 11 times, and the usage count of “time” is 5 times.

  The processing of the common word information generation unit 206 in step S006 is for outputting common word information, which is intermediate information necessary for the dialogue processing control unit 14 when the dialogue processing as the second processing is performed.

  In this embodiment, the common word information is obtained using a value obtained by adding the number of times of use and the word weight based on the number of common words. However, in addition to this, it is also possible to calculate the similarity using a vector space method.

The data output unit 207 transmits the common word information extracted in S006 to the dialogue processing control unit 14. (S007)
In the present embodiment, it is assumed that common word information with all comparison target document information is transmitted. Further, the transmission means from the batch processing control unit 13 to the dialogue processing control unit 14 may transfer data via a network or may send it via an offline medium such as a magnetic tape or a DVD-R.

  The above steps S001 to S007 are the common word information generation process using the batch process which is the first process.

  As described above, according to the collective processing described above, it is possible to perform processing with high processing efficiency by performing the processing until the common word information is generated collectively for a plurality of classification key documents.

  At that time, if the comparison target documents are enormous and the database is large, more results are obtained by the processing. In addition, there are cases where there are few common words with the comparison target document, or even if many common words are included, all of them are general words and have no characteristics.

  Therefore, the similarity is calculated for a result in which the number of common words is less than the set threshold value in the interactive process that is the second process following the batch process or a result that does not include a word having a word weight greater than the set threshold value. Filtering is performed according to conditions such as whether or not the obtained value satisfies a threshold value. As a result of filtering, it is possible to reduce transmission addition by selecting and transmitting only those having a high degree of similarity.

[Interactive processing]
Next, dialogue processing in the dialogue processing control unit 14 as the second processing will be described.

  First, the initialization unit 301 initializes the buffer units 351 to 353 of the memory unit 350 (S101).

  Next, the data input unit 303 receives the common word information transmitted from the batch processing control unit 13 and stores it in the common word information buffer unit 351 (S102). Here, it is assumed that the common word information stored in the common word information buffer unit 351 is the same as the common word information shown in FIG. 11 generated by the batch processing.

  Subsequently, the comparison target document similarity calculation unit 304 reads the common word information stored in the common word information buffer unit 351, calculates the similarity, and uses the calculated similarity as the comparison target document similarity. It is stored in the degree buffer unit 352 (S103). In order to calculate the comparison target document similarity, first, the product of the number of times of use and the word weight is calculated for each commonly used word. The total value of the products obtained for each commonly used word is set as each comparison target document similarity for each classification key document.

  For example, in the case of “comparison target document 1” shown in FIG. 11, the comparison target document similarity is 5 × 2.1 + 11 × 4.3 + 5 × 1.7 = 66.3.

  FIG. 12 shows an example of the comparison target document similarity stored in the comparison target document similarity buffer unit 352.

When the processes in steps S102 and S103 are performed for all the comparison target documents (S104), the field-specific similarity accumulation unit 305 determines the comparison target document similarity calculated for each comparison target document calculated in step S103 for each field. The result is integrated and stored in the field-specific similarity integrated value buffer unit 353 as the field-specific similarity integrated value. (S105)
When the similarity calculation result is in the state shown in FIG. 12, first, in the field of “database update”, the similarity 66.3 of “reduction of database update processing time” and the similarity 43.3 of “XML document database” are obtained. 5 is added, and the similarities of the documents classified as “database update” in the subsequent classification target documents are added to obtain an integrated similarity value for each field.

  FIG. 13 shows an example of integrated similarity values for each field obtained as a result of processing all the comparison target documents in step S105.

  Next, the field specifying unit 306 displays the contents of the comparison target document similarity buffer unit 352 on the display device that is an output device to which the field specifying results are connected in order from the field having the highest similarity value for each field. . This field identification result includes “field name”, “title” and “similarity” of the comparison target document corresponding to the field. The field specifying unit 306 also displays word weights used in calculating the similarity to each comparison target document. (S106) At this time, the word weight of each word is displayed in a rewritable state.

  FIG. 14 shows an example of the field identification result displayed in step S106. FIG. 15 shows an example of the word weight adjustment screen displayed in step S106. In the example shown in FIG. 15, the word weights are displayed by sorting in descending order of weight.

  The processes in steps S102 to S106 described above are limited in data to be referred to, and can be processed in a much shorter time than the time required for extracting the common word in the batch process.

Subsequently, the word weight adjustment unit 307 receives the adjustment value of the word weight displayed in step S106. (S107)
Specifically, the user refers to the contents of the displayed classification key document and the classification result, and when it is determined that the classification result is not correct, the user selects the word that affected the classification included in the classification result and the word weight. The classification result can be obtained again by referring to and adjusting the word weight.

  Specifically, among the displayed word weights, even though they do not show the characteristics of the field of the classification key document, when the weight is high, or conversely, Regardless, if the weight is low, the word weight can be adjusted.

  For example, the user determines that “database update” or “document search” is not appropriate as the field to which the classification key document shown in FIG. 9 belongs, and the cause is the word weight of “interactive” or “database”. Suppose that it is due to high. In this case, the word weight of “interactive” or “database” can be reset to a low value, for example, to 1.0.

  Subsequently, when the word weight adjustment unit 307 determines that the word weight has been changed (S108), the common word information buffer unit 351 is rewritten with the newly input word weight (S109). Thereafter, the field specifying process from step S103 is performed again.

  FIG. 16 shows an example of the similarity of the comparison target documents when the word weights of “interactive” and “database” are adjusted to 1.0 as in the above-described example. As a result of this adjustment, the scores for the “database update” and “document search” fields, which were not very suitable as the field to which the classification key document belongs, decreased, and the “document classification” field suitable as the classification destination increased. Result.

  In the present embodiment, the common word and its weight information are used as intermediate data for calculating the similarity, but in addition to these pieces of information, the score (plurality) is calculated by a plurality of methods on the batch processing side, An embodiment in which the field is specified by adjusting the blend ratio of the scores on the terminal side is also conceivable.

  The processing in steps S102 to S108 described above is repeated for all target classification key documents (S109). For example, if 2000 documents are input as the classification key document, the document is repeated 2000 times.

  According to the invention according to the first embodiment described above, by checking the basis of the similarity calculation result, for example, by adjusting the word weight and reducing the influence of words that are not suitable for classification, the similarity is calculated. It becomes possible to specify a new field. For this reason, it is possible to perform an interactive process in which the cause of the error in the classification result that cannot be realized by the conventional batch process is identified, the cause is removed, the classification process is re-executed, and the result is confirmed. It is also conceivable to accumulate word weight adjustment results as described above and use them to improve classification accuracy.

[Second Embodiment]
The similar document classification device 1a according to the second embodiment of the present invention will be described below with reference to the drawings. In the following description, description of the same points as those of the similar document classification device 1 according to the first embodiment will be omitted, and only different points will be described. Therefore, since the process in the batch process control unit 13 is the same, the description thereof is omitted, and only the process in the dialogue process control unit 14a is described.

  FIG. 17 shows the dialogue processing control unit 14a of the similar document classification device 1a according to the second embodiment of the present invention.

  The dialog processing control unit 14a illustrated in FIG. 17 is different from the dialog processing control unit 14 illustrated in FIG. 6 in that it does not include the word weight adjustment unit 307 but includes the field adjustment unit 308.

  The field adjustment unit 308 rewrites the common word information buffer unit 351 in accordance with an instruction input by a user from an input device such as a connected keyboard.

  FIG. 18 is a flowchart for explaining processing in the dialogue processing control unit 14a of the similar document classification device 1a according to the second embodiment of the present invention. In FIG. 18, the same processes as those in the flowchart described with reference to FIG.

  When the field identification result and the word weight are displayed in step S106, the field adjustment unit 308 accepts the field adjustment (S207). When it is confirmed that the field has been adjusted (S208), the field adjustment unit 308 performs field adjustment (S209).

  In the processing illustrated in FIG. 18, as field adjustment, for example, rewriting of the field-specific similarity integrated value buffer unit 353 will be described as an example. Specifically, if the operator determines that the displayed field is clearly invalid in the displayed field identification result, the field-specific similarity is set for the field, which is deleted from the field identification result. It is conceivable to perform processing such as rewriting the integrated similarity value buffer unit 353.

  According to the present invention according to the second embodiment described above, while confirming the basis of the similarity calculation result, for example, the field of the comparison target document used as the comparison target is adjusted to reduce the influence of the field not suitable for classification. It is possible to specify a new field by calculating the similarity.

  It goes without saying that the present invention includes various embodiments not described herein. Therefore, the technical scope of the present invention is defined only by the matters described in the above description and the invention specific matters according to the obvious claims.

It is a block diagram explaining the similar document classification | category apparatus based on Example 1 of this invention. It is an example of the comparison object document information memorize | stored in the comparison object document information storage part which concerns on Example 1 of this invention. It is an example of the word weight information memorize | stored in the word weight information storage part which concerns on Example 1 of this invention. It is a figure explaining the similar document classification | category apparatus which concerns on Example 1 of this invention. It is a figure explaining the batch process control part in the similar document classification device based on Example 1 of this invention. It is a figure explaining the dialogue processing control part in the similar document classification device concerning Example 1 of the present invention. It is a flowchart explaining the process in the batch processing control part shown in FIG. It is a flowchart explaining the process in the dialogue process control part shown in FIG. It is an example of the classification | category key document input into the similar document classification | category apparatus based on Example 1 of this invention. It is an example of the classification key document information produced | generated in the similar document classification | category apparatus based on Example 1 of this invention. It is an example of the common word information produced | generated in the similar document classification | category apparatus based on Example 1 of this invention. It is an example of the comparison object document similarity memorize | stored in the similar document classification | category apparatus based on Example 1 of this invention. It is an example of the similarity integrated value according to field memorize | stored in the similar document classification | category apparatus which concerns on Example 1 of this invention. It is an example of the field identification result calculated | required in the similar document classification | category apparatus based on Example 1 of this invention. It is an example of the word weight adjustment screen displayed in the similar document classification | category apparatus based on Example 1 of this invention. It is an example of the similarity of the comparison object document displayed in the similar document classification | category apparatus based on Example 1 of this invention. It is a figure explaining the dialogue processing control part in the similar document classification | category apparatus based on Example 2 of this invention. It is a flowchart explaining the process in the dialogue process control part shown in FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1,1a ... Similar document classification | category apparatus 11 ... Comparison object document information database 12 ... Word weight information database 13 ... Collective processing control part 14, 14a ... Dialog processing control part 101 ... Central processing control apparatus 102 ... ROM
103 ... RAM
DESCRIPTION OF SYMBOLS 104 ... Input device 105 ... Display device 106 ... Communication control device 107 ... Memory | storage device 108 ... Removable disk 109 ... Input / output interface 110 ... Bus 200 ... Control part 201 ... Initialization part 202 ... Input part 203 ... Word weight reading part 204 ... Classification key document information generation unit 205 ... Comparison target document information reading unit 206 ... Common word information generation unit 207 ... Data output unit 250 ... Memory unit 251 ... Information buffer unit 252 ... Classification key document information buffer unit 253 ... Comparison target document information buffer Unit 254 ... common word information buffer unit 300 ... control unit 301 ... initialization unit 302 ... input unit 303 ... data input unit 304 ... comparison target document similarity calculation unit 305 ... field-specific similarity accumulation unit 306 ... field identification unit 307 ... Word weight adjustment unit 308 ... field adjustment unit 350 ... memory 351 ... Common word information buffer 352 ... comparison document similarity buffer unit 353 ... sector similarity integration value buffer

Claims (10)

A document classification device for classifying a classification key document that is a target for classifying a field to which a document belongs,
A comparison target document information storage unit storing information of a comparison target document to be compared with the classification key document, and comparison target document information associated with a field of the comparison target document;
A word weight information storage unit storing a word and a word weight serving as an index indicating characteristics of the field of the document in which the word is included;
The classification key document is compared with the comparison target document information to extract common usage words that are commonly used in the classification key document and the comparison target document, and at least these common usage words and the common usage words A batch processing control unit that generates common word information in which the number of uses and the word weight of the common use word read from the word weight information storage unit are associated;
A similarity between a plurality of comparison target documents and the classification key document is obtained from the common word information, a field is specified based on the obtained comparison target document having a high similarity, and further, based on an instruction from the input device A dialogue processing control unit that adjusts the identified field;
A document classification apparatus comprising: A document classification device for classifying a classification key document that is a target for classifying a field to which a document belongs,
A comparison target document information storage unit storing information of a comparison target document to be compared with the classification key document, and comparison target document information associated with a field of the comparison target document;
A word weight information storage unit storing a word and a word weight serving as an index indicating characteristics of the field of the document in which the word is included;
The classification key document is compared with the comparison target document information to extract common usage words that are commonly used in the classification key document and the comparison target document, and at least these common usage words and the common usage words A common word information generating unit that generates common word information in which the number of uses and the word weight of the common used word read from the word weight storage unit are associated;
A similarity calculation unit that calculates a similarity based on the number of times of use of each commonly used word included in the common word information and the word weight of the commonly used word;
A field identification unit that determines a comparison target document having a high degree of similarity obtained for the classification key document, and that determines a classification result that identifies a field to which the comparison target document belongs as a field to which the classification key document belongs;
An adjusting unit for adjusting the field specified by the field specifying unit;
A document classification apparatus comprising: The document classification device according to claim 2, wherein the adjustment unit includes:
A document classification apparatus, wherein the word weight is varied to generate common word information. The document classification device according to claim 2, wherein the adjustment unit includes:
A document classification apparatus, wherein a specific field is deleted from a classification result obtained from the classification result. 5. The document classification device according to claim 2, wherein the similarity calculation unit includes:
A document classification apparatus characterized in that a sum of products of the number of times each common word is used and a word weight is calculated, and the calculated total value is used as a similarity. A document classification method for classifying a classification key document that is a target for classifying a field to which a document belongs,
The classification key document is compared with the comparison target document information stored in the comparison target document information storage unit in association with the information of the comparison target document to be compared with the classification key document and the field of the comparison target document. And common use words that are commonly used in the comparison target document, and extract at least the common use word, the number of times of use of the common use word, and the field of the document including the word and the word. Generating common word information associated with the word weight of the commonly used word read from the word weight information storage unit storing the word weight as an index indicating the feature;
Find the similarity between multiple comparison target documents and classification key documents from common word information, identify the field based on the compared target documents with high similarity, and further identify based on instructions from the input device Document classification method characterized by adjusting the selected field. A document classification method for classifying a classification key document that is a target for classifying a field to which a document belongs,
The classification key document is compared with the comparison target document information stored in the comparison target document information storage unit in association with the information of the comparison target document to be compared with the classification key document and the field of the comparison target document. And common use words that are commonly used in the comparison target document, and extract at least the common use word, the number of times of use of the common use word, and the field of the document including the word and the word. Generating common word information associated with the word weight of the commonly used word read from the word weight information storage unit in which the word weight serving as an index indicating the feature is stored;
Calculating the similarity based on the number of times of use of each commonly used word included in the common word information and the word weight of the commonly used word;
A comparison target document having a high degree of similarity required for the classification key document is determined, the field to which the comparison target document belongs is specified,
Reconcile identified areas,
A document classification method characterized by comprising:   8. The document classification method according to claim 7, wherein the common word information is generated by varying the word weight when adjusting the specified field.   8. The document classification method according to claim 7, wherein when the specified field is adjusted, the specific field is deleted from the classification result obtained from the classification result. 10. The document classification method according to claim 7, wherein when calculating the similarity, a sum of products of the number of times of use of each common word and the word weight is calculated, and the calculated total value is calculated. Document classification method characterized by similarity.

Images