JP2005128978A - Apparatus, program and method for automatic preparation of information analysis report - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To analyze information of a research object document against a comparison object document to prepare a report. <P>SOLUTION: This apparatus is an information analysis report preparation apparatus which automatically analyzes information of a research object document against a document used as a comparison object document, and it is provided with; an input means for inputting a research object document and a comparison object document; an input means for inputting conditions of information analysis; a selection means of a population document consisting of document groups similar to the research object document; an extraction means of an index word; an operation means of significance of the index word; an output means to output the selection result of the population and the operation result of the index word in a display means, a recording means, or a communication means. Thereby, information on the research object documents can be expressed exactly without reading the contents of research object documents and a vast quantity of comparison object documents at all. A program is also provided. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a document analysis apparatus, and more particularly to an information analysis report automatic creation apparatus, program, and method for analyzing a document or group of documents to be investigated and expressing the characteristics thereof.

  The amount of technical documents and other documents including patent documents has been steadily increasing year by year. In recent years, since document data has been digitized and circulated, a system for automatically searching only a document similar to a document to be investigated from a large number of documents has been put into practical use. However, the amount of similar documents as a search result is still large, and in order to know the content or character of the document to be investigated, a skilled person has to read the similar document as a search result.

For example, in the “similar document search device and similar document search method” of Patent Document 1, an index word included in a document or group of documents to be searched is compared with an index word included in a group of documents to be compared, and a similar index The similarity is calculated from the word type and the number of appearances, and the documents with the highest similarity are output in order. FIG. 6 is an overall configuration diagram of the apparatus disclosed in Patent Document 1. In FIG. Conventionally, an investigation target document input from the input device 602 is subjected to similarity calculation processing by comparing with a document group in a database in the external auxiliary storage device 603 under a certain extraction condition by a similarity calculation system in the control device 601. Then, the content of the document with high similarity is read based on the result of the output document list that is output by the output device 604, and an expert evaluator reads it to evaluate the document to be investigated. The evaluator had to read several to thousands of documents in order to know the contents of documents with high similarity.
Japanese Patent Laid-Open No. 11-73415 “Similar Document Retrieval Device and Similar Document Retrieval Method”

  However, conventionally, the similar document automatic search system similar to the above-described Patent Document 1 outputs a list of documents similar to the search target document from the comparison target document group as a search result, and the evaluator is similar to the search target document. From this list of documents, extract several to thousands of documents with high similarity, read them, find documents that are similar to the documents to be investigated, read them further, evaluate them, Since the personality was positioned, the evaluator had to extract several to thousands of documents and read them, so that there was a problem that an expression that accurately represents the character of the document to be investigated could not be found.

  Therefore, the present invention automatically generates an information analysis report that enables a human to accurately report information on a document to be investigated without reading the contents of any document to be investigated and a large number of documents to be compared. The purpose is to create.

  In order to solve the above-mentioned problems, the present invention specifies and inputs a survey target document and a comparison target document, a population comprising a group of documents similar to the survey target document, input means for inputting information analysis conditions Selection means for selecting a document from the comparison target document, extraction means for extracting a characteristic index word for the population document of the search target document, and a population or index word representing the characteristics of the search target document Output means for outputting to recording or communication.

In order to solve the above problems, the present invention provides an input means for specifying and inputting the investigation target document and the comparison target document, and inputting conditions for performing the information analysis, and a document similar to the investigation target document. A selection means for selecting a group document consisting of groups from the comparison target document, an extraction means for extracting characteristic index words of the survey target document with respect to the population document, and a population or An information analysis of a document to be investigated, comprising an output means for displaying, recording, or outputting an index word for communication,
In the selection means for selecting a population document consisting of a document group similar to the survey target document from the comparison target documents, the selection means is a selection means for selecting based on a result of a calculation means for calculating a similarity ratio for the comparison target document. It is characterized by that.

In order to solve the above problems, the present invention provides an input means for specifying and inputting the investigation target document and the comparison target document, and inputting conditions for performing the information analysis, and a document similar to the investigation target document. A selection means for selecting a group document consisting of groups from the comparison target document, an extraction means for extracting characteristic index words of the survey target document with respect to the population document, and a population or An information analysis of a document to be investigated, comprising an output means for displaying, recording, or outputting an index word for communication,
The selection means is a selection means that is selected based on a result of a calculation means for calculating a similarity ratio for a comparison target document, and the calculation means for calculating the similarity ratio includes an appearance frequency and a document frequency for each index word of each document. It is a means for calculating the similarity based on the function value of.

In order to solve the above problems, the present invention provides an input means for specifying and inputting the investigation target document and the comparison target document, and inputting conditions for performing the information analysis, and a document similar to the investigation target document. A selection means for selecting a group document consisting of groups from the comparison target document, an extraction means for extracting characteristic index words of the survey target document with respect to the population document, and a population or An information analysis of a document to be investigated, comprising an output means for displaying, recording, or outputting an index word for communication,
In the output means for displaying, recording, or outputting the population or index word representing the characteristics of the survey object document, the display means for distributing and displaying the population or the index word in the form of a map; The display means for displaying a part of the data of the population or the index word, and the output means automatically or select a fixed comment according to the contents, or enter or select a free comment Output means for displaying is included.

  According to the present invention, the comparison target document and the comparison target document are designated and input, the input means for inputting the condition for information analysis, and the population document composed of a document group similar to the survey target document is set as the comparison target. Selection means for selecting from documents, extraction means for extracting characteristic index words for the population document of the survey target document, and displaying, recording, or communicating a population or index word representing the characteristics of the survey target document Therefore, in the information analysis of the survey target document, the characteristics of the survey target document with respect to the comparison target document can be accurately and automatically created.

  Further, according to the present invention, an input unit for specifying and inputting a search target document and a comparison target document and inputting a condition for information analysis, and a population document composed of a document group similar to the search target document Selection means for selecting from comparison target documents; extraction means for extracting characteristic index words for the population document of the search target document; and displaying or recording a population or index word representing the characteristics of the search target documents; Or an output means for outputting to communication, wherein the selecting means calculates a similarity ratio with respect to the comparison target document in a selection means for selecting a population document consisting of a document group similar to the survey target document from the comparison target documents. Since there is a selection means to select based on the result of the calculation means, in the information analysis of the survey target document, the characteristics of the survey target document with respect to the comparison target document are created accurately and automatically Rukoto can.

  Further, according to the present invention, in the information analysis of the survey target document, an information analysis report automatic creation device that accurately and automatically creates the characteristics of the survey target document with respect to the comparison target document, the survey target document and Specifying and inputting the comparison target document, input means for inputting conditions for performing the information analysis, selection means for selecting a population document consisting of a document group similar to the investigation target document from the comparison target document, Extraction means for extracting a characteristic index word for the population document of the survey target document, and output means for displaying, recording, or outputting the population or index word representing the characteristics of the survey target document, The selecting means is a selecting means that selects based on the result of the calculating means for calculating the similarity to the comparison target document, and the calculating means for calculating the similarity is Since it has a means for calculating the similarity based on the appearance frequency for each reference and the function value of the document frequency, it accurately and automatically creates the characteristics of the target document for comparison in the information analysis of the target document be able to.

  Further, according to the present invention, in the information analysis of the survey target document, an information analysis report automatic creation device that accurately and automatically creates the characteristics of the survey target document with respect to the comparison target document, the survey target document and Specifying and inputting the comparison target document, input means for inputting conditions for performing the information analysis, selection means for selecting a population document consisting of a document group similar to the investigation target document from the comparison target document, Extraction means for extracting a characteristic index word for the population document of the survey target document, and output means for displaying, recording, or outputting the population or index word representing the characteristics of the survey target document, In the output means for displaying, recording, or outputting to the communication a population or index word representing the characteristics of the document to be investigated, the population or the index word is distributed in a map form The display means for displaying the data, the display means for displaying a part of the data of the population or the index word, and the output means automatically or select a fixed comment according to the contents, or freely Therefore, in the information analysis of the survey target document, the characteristics of the survey target document with respect to the comparison target document can be accurately and automatically created.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

Define or explain the vocabulary used in this specification.
Survey target document d: A certain project related to the survey. For example, a document such as a patent gazette number or a collection thereof.
Comparison target document P: a document to be compared with a survey target document or a set thereof. including d.
Population document S: A group of documents similar to the survey target document d among the comparison target documents P. including d.

“D” or “d”, “P” or “P”, or “S” or “S” attached to the component in the figure means the survey target document, the comparison target document, or the population document, respectively. It is also attached to components and operations so that it can be easily identified. For example, the index word (d) means the index word of the survey target document d, the index word (P) means the index word of the comparison target document P, and the index word (S) means the population. This means an index word of the document S.
The TF calculation is a calculation of Term Frequency, which is a calculation of an appearance frequency (index word frequency) of an index word included in a document in the document.
The DF of the DF operation is a calculation of Document Frequency, which is a calculation of the number of hits when the document frequency is searched from the search target document group using an index word included in a certain document.
The IDF calculation is, for example, the reciprocal of the DF calculation result or the logarithm obtained by multiplying the reciprocal by the number of documents P or S.

To simplify the following explanation, abbreviations are determined.
N: Number of documents of comparison target document P N ′: Number of documents of population document S TF (d): Frequency of occurrence in index word TF (P): P according to index word of P Appearance frequency DF (P) in P: Document frequency in P DF (S): Document frequency in P DF (S): Document frequency in S according to S index word IDF (P): DF (P ) X logarithm of document number: ln [N / DF (P)]
IDF (S): reciprocal of DF (S) × logarithm of the number of documents: ln [N ′ / DF (S)]
TFIDF: product of TF and IDF. Calculated for each index word in the document.
Similarity: The degree of similarity between the survey target document d and a certain document belonging to the comparison target document P.

Here, the index word is a so-called keyword, which is a word cut out from all or part of the document. Words can be extracted using known methods or commercially available software to extract meaningful nouns, excluding particles and conjunctions, and to create a database of index words (thesaurus) in advance. It is also possible to use index words held in the database and obtained from the database.
In the case of a document group having a plurality of documents to be investigated, the extraction target may be the index word as described above, but the IPC classification, the group of companies, the yearly group such as the patent application year and the patent registration It is possible even in years. Hereinafter, in this specification, the index word is often used as a representative.

FIG. 1 is a diagram showing a hardware configuration of an information analysis report automatic creating apparatus according to an embodiment of the present invention.
As shown in the figure, an information analysis report automatic creation device according to an embodiment of the present invention includes a processing device 1 including a CPU (central operator) and a memory (recording device), and a keyboard (manual input instrument). ) And the like, an input device 2 as a recording means, a recording device 3 as a recording means for storing document data, conditions, work results by the processing device 1, etc., and a feature index word extraction result, etc. are displayed on a map, data, etc. It is comprised from the output device 4 which is an output means to perform.

  FIG. 2 is a diagram for explaining in detail the configuration and functions of the information analysis report automatic creating apparatus, program, and method according to an embodiment of the present invention.

  The processing apparatus 1 includes an investigation target document d reading unit 110, an index word (d) extraction unit 120, a TF (d) calculation unit 121, a comparison target document P reading unit 130, an index word (P) extraction unit 140, and a comparison target document. TF (P) calculation unit 141 for P, IDF (P) calculation unit 142 for comparison target document P, similarity calculation unit 150, population document S selection unit 160, index word (S) extraction unit 170, IDF (S) The calculation unit 171 and the feature index word TF (d) IDF (S) calculation unit 180 are included.

  The input device 2 includes an investigation target document d condition input unit 210, a comparison target document P condition input unit 220, an extraction condition and other input unit 230, and the like.

  The recording device 3 includes a condition recording unit 310, a work result storage unit 320, a document storage unit 330, and the like. The document storage unit 330 includes an external database and an internal database. The external database means a document database such as IPDL of a patent electronic library serviced by the Patent Office or PATOLIS serviced by Patrice Co., Ltd. The internal database is a database in which data such as a patent JP-ROM that is sold is stored by itself, an FD (floppy disk), a CDROM (compact disk), an MO (magneto-optical disk), a DVD (which stores documents) A device that reads from a medium such as a digital video disk), a device such as an OCR (optical information reader) that reads a document that has been output on paper or handwritten, and a device that converts the read data into electronic data such as text. Is included.

  The output device 4 includes a map creation condition reading unit 410, a map data loading unit 412, a map (graph / table) generation unit 415, a population data output condition reading unit 420, an output data loading unit 422, a comment condition reading unit 430, A standard comment capturing unit 432, a comment adding unit 435, a map / data / comment composite shaping output unit 440, and the like are included.

  1 and 2, the communication means for exchanging signals and data among the processing device 1, the input device 2, the recording device 3, and the output device 4 is directly connected by a USB (Universal System Bus) cable or the like. Alternatively, transmission / reception may be performed via a network such as a LAN (local area network), or may be performed via a medium such as an FD, CDROM, MO, or DVD storing a document. Alternatively, a part or a combination of these may be used.

  The functions of the information analysis report automatic creating apparatus, program, and method according to an embodiment of the present invention will be described in detail with reference to FIG.

  In the input device 2 of FIG. 2, the survey target document d condition input unit 210 sets conditions for reading the survey target document d through an input screen or the like. The comparison target document P condition input unit 220 sets conditions for reading the comparison target document P on an input screen or the like. The extraction condition and other input unit 230 displays the index word extraction condition of the investigation target document d and the comparison target document P, the TF calculation condition, the IDF calculation condition, the similarity ratio calculation condition, the similar document selection condition, the map, etc. Set creation conditions, data output conditions, comment addition conditions, etc. These input conditions are sent to and stored in the condition recording unit 310 of the recording device 3.

  In the processing apparatus 1 of FIG. 2, the survey target document d reading unit 110 reads the survey target document from the document storage unit 330 based on the condition of the condition recording unit 310, and the index word (d) extraction unit 120. Sent to. The index word (d) extraction unit 120 extracts an index word from the document obtained by the investigation target document d reading unit 110 based on the condition of the condition recording unit 310 and stores it in the work result storage unit 320.

  The comparison target document P reading unit 130 reads a population document from the document storage unit 330 based on the condition of the condition recording unit 310 and sends the document to the index word (P) extraction unit 140. The index word (P) extraction unit 140 extracts an index word from the document obtained by the comparison target document P reading unit 130 based on the condition of the condition recording unit 310 and stores it in the work result storage unit 320.

  The comparison target document P reading unit 130 and the index word (P) extraction unit 140 are usually all publications such as patent publications, which are one of the comparison target documents. If you prepare and save it, you do not need to cut it out again, so you can omit it.

  The TF (d) calculation unit 121 performs TF calculation on the work result of the index word (d) extraction unit 120 for the investigation target document d stored in the work result storage unit 320 based on the condition of the condition recording unit 310. Then, it is stored in the work result storage unit 320 or directly sent to the similarity calculation unit 150 or the feature index word TF (d) IDF (S) calculation unit 180.

  Based on the condition of the condition recording unit 310, the TF (P) calculation unit 141 performs TF calculation on the work result of the index word (P) extraction unit 140 for the comparison target document P stored in the work result storage unit 320. And stored in the work result storage unit 320 or directly sent to the similarity ratio calculation unit 150.

  Based on the condition of the condition recording unit 310, the IDF (P) calculation unit 142 performs IDF calculation on the work result of the index word (P) extraction unit 140 for the comparison target document P stored in the work result storage unit 320. And stored in the work result storage unit 320 or directly sent to the similarity calculation unit 150.

  Based on the conditions of the condition recording unit 310, the similarity ratio calculation unit 150 can directly calculate the calculation results of the TF (d) calculation unit 121, the TF (P) calculation unit 141, and the IDF (P) calculation unit 142 from each of them. Obtained from the work result storage unit 320, the similarity ratio of each of the comparison target documents P with respect to the survey target document d is calculated, and is added to each of the comparison target documents P as similarity ratio data. It is sent to the population document S selection unit 160.

  The similarity ratio calculation unit 150 calculates the similarity ratio for each index word, for example, a calculation represented by a TFIDF calculation, and the like, and the similarity ratio of each document of the comparison target document P to the survey target document d is calculated. Calculated. The TFIDF calculation is a product of the TF calculation result and the IDF calculation result. A method of calculating the similarity rate will be described in detail later.

  Since there are various methods for calculating the similarity ratio, the similarity ratio calculation section 150 based on the TF (d) calculation section 121, the TF (P) calculation section 141, and the IDF (P) calculation section 142 is not changed. However, when the method of calculating similarity is not required, the TF (d) calculation unit 121, TF (P) calculation unit 141, and IDF (P) calculation unit 142 are all omitted. Needless to say, only the similarity calculation unit 150 may be used.

  Based on the condition of the condition recording unit 310, the population document S selection unit 160 records the conditions in the similarity calculation result of the population document S from the results of the work result storage unit 320 or the direct similarity calculation unit 150. Selection is performed by the selected selection method, for example, the documents are sorted in descending order of similarity, and the necessary number indicated in the condition is selected and sent to the work result storage unit 320 or the index word (S) extraction unit 170 directly.

  In addition, since the output from the population document S selection unit 160 may proceed directly to the map data capturing unit 412 or the output data capturing unit 422, it goes without saying that the subsequent configuration is unnecessary.

  The index word (S) extraction unit 170 extracts the index word (S) from the population document S as a result of the work result storage unit 320 or the population document S selection unit 160 based on the condition of the condition recording unit 310. The result is sent to the work result storage unit 320 or directly to the IDF (S) calculation unit 171.

  The IDF (S) calculation unit 171 performs IDF calculation on the work result of the work result storage unit 320 or the direct index word (S) extraction unit 170 based on the condition of the condition recording unit 310, and stores the result in the work result storage unit 320. It is stored or directly sent to the feature index word TF (d) IDF (S) calculation unit 180.

  The feature index word TF (d) IDF (S) calculation unit 180 is based on the condition of the condition recording unit 310, or from the work result storage unit 320 or the result of the TF (d) calculation unit 121 and the direct IDF (S). From the result of the calculation unit 171, select a population document, for example, in ascending order of similarity, by the required number written in the selection condition or the number selected by the calculation result based on the condition, The result is sent to the work result storage unit 320.

  In the recording device 3 of FIG. 2, the condition recording unit 310 records information such as conditions obtained from the input device 2, and sends necessary data to each based on a request from the processing device 1 or the output device 4. The work result storage unit 320 stores the work result of each component in the processing device 1 and sends necessary data to each based on a request from the processing device 1 or the output device 4.

  The document storage unit 330 stores and provides necessary document data obtained from an external database or an internal database based on a request from the input device 2 or the processing device 1.

  In the output device 4 of FIG. 2, the map creation condition reading unit 410 reads map creation conditions based on the conditions of the condition recording unit 310 and sends them to the map data fetching unit 412. The population data output condition reading unit 420 reads out the output condition of the population data based on the condition of the condition recording unit 310 and sends it to the output data capturing unit 422. The comment condition reading unit 430 reads out the comment output condition and the additional writing condition based on the conditions of the condition recording unit 310 and sends them to the comment additional writing unit 432.

  The map data capturing unit 412 calculates the result of the population document S selection unit 160 and the feature index word TF (d) IDF (S) stored in the work result storage unit 320 according to the conditions of the map creation condition reading unit 410. The result of the unit 180 is fetched together with the data of the document storage unit 330 and sent to the work result storage unit 320 or the direct map (graph / table) generation unit 415.

  The map (graph / table) generation unit 415 generates a graph, a table, a title, a legend, and the like using the data from the map data acquisition unit 412, and sends the graph, table, title, legend, etc.

  The output data capturing unit 422 is a result of the population document S selection unit 160 and a feature index word TF (d) IDF (S) calculation unit stored in the work result storage unit 320 according to the conditions of the data output condition reading unit 420. The result of 180 or the like is taken together with the data in the document storage unit 330 and sent to the work result storage unit 320 or directly to the map / data / comment composite shaping output unit 440.

  The fixed comment capturing unit 432 captures the data of the work result storage unit 320 and the document storage unit 330 in accordance with the conditions of the comment condition reading unit 430, and inputs the data to the comment addition unit 435 or the direct map data / comment composite output unit 440. Sent.

  According to the conditions of the comment condition reading unit 430, the comment adding unit 435 directly prepares an external input device such as a keyboard or an OCR or an internal database of the document storage unit 330 in advance by the evaluator of the survey target document d. Data to be added as a comment is prepared and sent to the work result storage unit 320 or directly to the map / data / comment composite shaping output unit 440.

  The map / data / comment composite shaping output unit 440 outputs the conditions and data output from the map (graph / table) generation unit 415, the conditions and data output from the output data acquisition unit 422, and the standard comment acquisition unit 432. Conditions, data and conditions and data output from the comment adding section 435 are obtained directly or from the work result storage section 320, and map data, comments are formed into an optimal form as paper output, and then output in combination. Prepare, display on map, output to data list, and output comments or part of them so that they can be displayed, printed, or stored as data.

  3, 4, and 5 are diagrams for explaining typical examples of operations in the information analysis report automatic creating apparatus, program, and method according to an embodiment of the present invention.

  FIG. 3 is a condition setting flowchart showing the operation of each component by the input device 2. First, after initialization (step S201), input conditions are distinguished (step S202). At this time, when the condition is the condition input of the survey target document d, the condition of the survey target document d is input in the survey target document d condition input unit 210 (step S210). Next, the input condition is confirmed on the display screen. If it is good, the setting is selected and stored in the condition recording unit 310 (step S310). If it is bad, return is selected and the process returns to step S210 (step S211).

  On the other hand, when the condition is the condition input of the comparison target document P in step S202, the comparison target document P condition input unit 220 inputs the condition of the comparison target document P (step S220). Next, the input condition is confirmed on the display screen. If it is good, the setting is selected and stored in the condition recording unit 310 (step S310). If it is bad, return is selected and the process returns to step S220 (step S221).

  When the condition is the extraction condition or the like in step S202, the extraction condition or the like is input in the extraction condition or other input unit 230 (step S230). Next, the input condition is confirmed on the display screen. If it is good, the setting is selected and stored in the condition recording unit 310 (step S310). If it is bad, return is selected and the process returns to step S230 (step S231). In step S230, both the extraction condition for the survey target document d and the extraction condition for the population document S from the comparison target document P are set.

  FIG. 4 is a flowchart showing the operation of each component by the processing apparatus 1. First, after initialization (step S101), based on the condition of the condition recording unit 310, a document to be read from the document storage unit 330 is classified into a survey target document d and a comparison target document P (step S102). When the document to be read is the survey target document d, the survey target document d reading unit 110 reads the survey target document from the document storage unit 330 (step S110). Next, the index word (d) extraction unit 120 extracts the index word of the investigation target document d (step S120). Subsequently, the TF (d) calculation unit 121 performs TF calculation for each of the extracted index words (step S121).

  On the other hand, when the document to be read is the comparison target document P in step S102, the comparison target document P is read by the comparison target document P reading unit 130 (step S130). Next, the index word (P) extraction unit 140 performs index word extraction of the comparison target document P (step S140). Subsequently, for each of the extracted index words, the TF (P) calculation unit 141 performs TF calculation (step S141) and the IDF (P) calculation unit 142 performs IDF calculation (step S142).

  Next, the TF (d) calculation result output from the TF (d) calculation unit 121, the TF (P) calculation result output from the TF (P) calculation unit 141, and the IDF output from the IDF (P) calculation unit 142 (P) Based on the calculation result, the similarity calculation unit 150 outputs the calculation result for each index word of the document, for example, the average value of the index words is calculated and used as the document similarity, and the similarity ratio is calculated. Calculation is performed (step S150).

  When the method of calculating the similarity rate is not TFIDF or the like, the similarity rate is calculated by another method from the index word (d) extraction unit 120 of the investigation target document d and the index word (P) extraction unit 140 of the comparison target document P. Sometimes required.

  Next, the population document S selection unit 160 rearranges the documents calculated in step S150 in the order of similarity, and selects the number of population documents S according to the extraction conditions and other conditions set in the input unit 230 ( Step S160).

  These data may be directly used by the map (graph / table) generation unit 415 or the map / data / comment combined shaping output unit 440 of the output device 4.

  Next, the index word (S) extraction unit 170 of the population document S extracts the index word (S) of the population document S selected in step S160 (step S170).

  Next, IDF calculation is performed by the IDF (S) calculation unit 171 for each index word (d) (step S171).

  Next, the result of the IDF (S) calculation for each index word (d) in the population document S in step S171, and the TF (d) for each index word (d) in the survey target document d in step S121. Based on the result of the calculation, a feature index word calculation TF (d) IDF (S) is performed (step S180).

  FIG. 5 is a flowchart of outputting maps, data, and / or comments by the output device 4. First, after initialization (step S401), conditions to be read from the condition recording unit 310 are classified into map creation conditions, data output conditions, and comment addition conditions (step S402).

  When the condition read from the condition recording unit 310 is the map creation condition (step S410), if the map requires the condition (step S411), the map data capturing unit 412 captures the map data from the work result storage unit 320. (Step 412). A map such as a graph or a table is generated in accordance with the map creation condition of the map creation condition reading unit 410 (step S415), and then the map is prepared to be displayed (step 419). The data is sent to the output unit 440.

  On the other hand, when the condition read from the condition recording unit 310 is a population data output condition (step S420), if the condition requires data (step S421), the data is output from the work result storage unit 320 by the output data capturing unit 422. The data is taken in (step S422). Next, the data is output in accordance with the data output condition of the data output condition reading unit 420 (step S423), and then the data is prepared to be output (step 429). To 440.

  On the other hand, when the condition read from the condition recording unit 310 is a comment condition (step S430), if the condition requires a comment (step S431), the map data / comment combined output unit 440 Prepare a frame that can be appended, and manually enter it from the keyboard or OCR (step S435), or use a comment prepared in advance in the internal database of the document storage unit 330, A comment is taken in (step S432), and subsequently, a preparation for outputting a comment is made (step 439), and the comment is sent to the map / data / comment combined output unit 440.

  If it is not a condition for displaying a map in step S411, or if it is not a condition for outputting data in step S421, or if it is not a condition for adding a comment in step S431, the process ends at that point, and the map data comment Data is not sent to the composite shaping output unit 440.

  FIG. 7 is an input condition setting screen of the input device 2 of one embodiment of the information analysis report automatic creating device according to the present invention.

  FIG. 7 is a display example of the input condition setting (1) screen of the input device 2 of the information analysis report automatic creation device. In FIG. 7, “Survey target document” is selected from “Survey target document” and “Comparison target document” in the “Target document” window, and then “Public patent” and “Registered patent” in the “Document content” window. Select “Public Patent” from “Utility Model”, “Academic Literature”, etc., then “Read Data” window “Company DB1” “Company DB2” “Patent Office IPDL” “PATOLIS” “Other Commercial DB1” “FD” is selected from “Other Commercial DB 2”, “FD”, “CD”, “MO”, “DVD”, “Other”, etc., and “Document 1” “Document 2” “Document 3” “Document 4” of “FD”. In this example, “Document 3” is selected from “Document 5”, “Document 6”, and the like.

  FIG. 8 is a display example of the input condition setting (2) screen of the input device 2 of the information analysis report automatic creation device. In FIG. 8, a “comparison target document” is selected from “investigation target document” and “comparison target document” in the “target document” window, and then “published patent” and “registered patent” in the “document content” window. "Utility model", "Academic literature", etc., select "Public patent" and "Registered patent", then "extracted contents" window "claim" "prior art" "invention problem" Select both “Claim” and “Summary” from “Means / Effects”, “Examples”, “Explanation of Figures”, “Drawings”, “Summary”, “Bibliographic Items”, “Progress Information”, “Registration Information”, “Others”, etc. Next, in the “data read” window, “in-house DB1” is selected from the same items as described above. The setting conditions on the input condition setting screen in this example set the investigation target document d condition input unit 210 and the comparison target document P condition input unit 220.

  FIG. 9 is a display example of an input condition setting (3) screen of the input device 2 of the information analysis report automatic creation device. Select “In-house keyword extraction 1” from “In-house keyword extraction 1”, “In-house keyword extraction 2”, “Commercial keyword extraction 1”, “Commercial keyword extraction 2”, etc. Next, “similarity 1” from among “similarity 1” “similarity 2” “similarity 3” “similarity 4” “similarity 5” “similarity 6” of the window of “similarity calculation method” Next, select “Number of Population Documents” from the “Number of Population Documents”, “Number of Non-Population Documents”, etc. in the “Select Population Document” window, and then select “Top 100” and “Top 1000”. This is an example of a state in which “higher 3000 items” is selected from “items”, “higher 3000 items”, “higher 5000 items”, “numerical value input” and the like. The setting condition on the extraction condition setting screen in this example sets the extraction condition other input unit 230.

  FIG. 10 is a display example of an output condition setting screen of the output device 2 of the information analysis report automatic creation device. Select “x axis: number of index words” and “y axis: index word ranking” for “x axis” in the “map calculation method” window, and then “map 1” in the “map position” window. "Sheets" "two maps" "one map with data" "two maps with data" "one map with comments" "two maps with comments" "map one with data and comments" "maps" 2. Select “1 map” from “Data, with comments”, etc., and then “Original words” from the “Output words” window, “Original words”, “Technical terms”, “Population characteristic words”, etc. And “Top 20” is selected from “None”, “Top 5”, “Top 10”, “Top 15”, “Top 20”, “Numeric Input”, and the like. Next, “(free entry)” in the window of “Comment” was left blank. Thus, the extraction conditions and other output conditions of the input unit 230 are set.

  FIG. 11 shows the selection result of the population document S selection unit 160 and the characteristic index word TF (d) IDF when the information analysis report automatic creating apparatus of the present invention is inputted in the examples of FIGS. (S) This is a specific example of the output result output by the map / data / comment composite shaping output unit 440 through the map (graph / table) generation unit 415 and the extraction result of the calculation unit 180.

  From FIG. 11, in the information analysis report automatic creating apparatus according to the present invention, the published patent gazette related to the “system for promoting access to the WEB site” of the surveyed document d is about 10 years as a comparative document. As a result of examining characteristic index words compared to the minute document, it is understood that “member”, “image”, “WEB”, “display”, and the like are characteristic index words.

  FIGS. 12 to 31 show another example of the information analysis report automatic creating apparatus of the present invention. The selection results and characteristics of the population document S selection unit 160 when information is analyzed under the same conditions as in FIG. A map of an output result output from the index word TF (d) IDF (S) operation unit 180 by the map / data / comment composite shaping output unit 440 via the map (graph / table) generation unit 415 It is a specific example of a display.

Is a block diagram according to the present invention FIG. 1 is a detailed configuration diagram of an embodiment according to the present invention. These are flowcharts which show operation | movement of the input device 2 of one Embodiment which concerns on this invention. These are flowcharts which show operation | movement of the processing apparatus 1 of one Embodiment which concerns on this invention. FIG. 5 is a flowchart showing the operation of the output device 4 according to the embodiment of the present invention. Is a conventional example Is an example of input condition setting of an embodiment according to the present invention Is an example of input condition setting of an embodiment according to the present invention Is an example of input condition setting of an embodiment according to the present invention Is an example of output condition setting of an embodiment according to the present invention Is an embodiment of the invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention Is a display example of another embodiment of the present invention

Explanation of symbols

1: processing device, 2: input device, 3: recording device, 4: output device,
110: Survey target document d reading unit, 120: Index word (d) extraction unit, 130: Comparison target document P reading unit, 140: Index word (P) extraction unit, 121: TF (d) calculation unit, 141: TF (P) calculation unit, 142: IDF (P) calculation unit, 150: similarity ratio calculation unit, 160: population document S selection unit, 170: index word (S) extraction unit, 171: IDF (S) calculation unit, 180: feature index word TF (d) IDF (S) calculation unit,
210: Survey target document d condition input unit, 220: Comparison target document P condition input unit, 230: Extraction condition other input unit,
310: Condition recording unit, 320: Work result storage unit, 330: Document storage unit,
410: Map creation condition reading unit, 412: Map data loading unit, 420: Population data output condition reading unit, 422: Output data loading unit, 430: Comment condition reading unit, 432: Standard comment loading unit, 435: Comment Appending part, 440: Map, data, and comment combined shaping output part.

Claims (12)

An information analysis report automatic creation device that accurately and automatically creates the characteristics of a survey target document with respect to a comparison target document in information analysis of the survey target document,
An input means for specifying and inputting the survey target document and the comparison target document, and inputting conditions for performing the information analysis,
A selection means for selecting, from the comparison target documents, a population document consisting of a document group similar to the survey target document;
Extraction means for extracting characteristic index terms for the population document of the survey target document;
An information analysis report automatic creation apparatus comprising: output means for displaying, recording, or outputting to a communication a population or index word representing the characteristics of the document to be investigated. The information analysis report automatic creation device according to claim 1,
In a selection means for selecting a population document consisting of a document group similar to the survey target document from the comparison target documents,
An information analysis report automatic creation device, wherein the selection means is a selection means that is selected based on a result of a calculation means that calculates a similarity to a comparison target document An information analysis report automatic creation device according to claim 2,
In the selection means, the selection means is selected based on the result of the calculation means for calculating the similarity to the comparison target document.
An information analysis report automatic creating apparatus, wherein the calculation means for calculating the similarity is a similarity calculation means based on a function value of the appearance frequency and document frequency for each index word of each document. An information analysis report automatic creation device according to claim 2,
In the output means for displaying, recording, or outputting the population or index word representing the characteristics of the document to be investigated,
The display means for distributing and displaying the population or the index terms in a map;
Display means for partially displaying data of the population or the index word;
The output means includes an output means for automatically or selecting a standard comment corresponding to the content, or inputting or selecting a free comment, and displaying the information analysis report automatically. apparatus. An input means for inputting a condition for performing information analysis by designating and inputting a survey target document and a comparison target document,
A selection means for selecting, from the comparison target documents, a population document consisting of a document group similar to the survey target document;
Extraction means for extracting characteristic index terms for the population document of the survey target document;
An information processing means for automatically generating an information analysis report, comprising: an output means for displaying, recording, or outputting the population or the index word representing characteristics of the investigation target document to communication Works
In the information analysis of the survey target document, an information analysis report automatic creation program for accurately and automatically creating the characteristics of the survey target document with respect to the comparison target document,
A function of designating and inputting the survey target document and the comparison target document, and inputting conditions for performing the information analysis;
A function of selecting a population document consisting of a document group similar to the survey target document from comparison target documents;
A function of extracting characteristic index terms for the population document of the survey target document;
An information analysis report automatic creation program characterized by realizing a function of displaying, recording, or outputting to a communication a population or index word representing the characteristics of the survey target document. It operates in the information processing means of the information analysis report automatic creation device according to claim 5,
In the information analysis of the survey target document, an information analysis report automatic creation program for accurately and automatically creating the characteristics of the survey target document with respect to the comparison target document,
In the function of selecting a population document consisting of a document group similar to the survey target document from the comparison target documents,
An information analysis report automatic creation program characterized by realizing that the selection function is a function that is selected based on a result of a calculation function for calculating a similarity to a comparison target document. It operates in the information processing means of the information analysis report automatic creation device according to claim 6,
In the information analysis of the survey target document, an information analysis report automatic creation program for accurately and automatically creating the characteristics of the survey target document with respect to the comparison target document,
In the function of selecting by the result of the calculation function that calculates the similarity to the comparison target document,
An information analysis report automatic creation program that realizes that the calculation function for calculating the similarity is a function for calculating a similarity based on a function value of the appearance frequency and document frequency for each index word of each document. It operates in the information processing means of the information analysis report automatic creation device according to claim 6,
In the information analysis of the survey target document, an information analysis report automatic creation program for accurately and automatically creating the characteristics of the survey target document with respect to the comparison target document,
In the function of displaying, recording, or outputting the population or index word representing the characteristics of the survey target document to the communication,
The display function for distributing and displaying the population or the index terms in a map form;
A display function for partially displaying the data of the population or the index word;
An information analysis report characterized by realizing that the output function includes an output function for automatically or selecting a fixed comment according to the content, or inputting or selecting a free comment for display. Automatic document creation program. An input means for inputting a condition for performing information analysis by designating and inputting a survey target document and a comparison target document,
A selection means for selecting, from the comparison target documents, a population document consisting of a document group similar to the survey target document;
Extraction means for extracting characteristic index terms for the population document of the survey target document;
Using the information analysis report automatic creation device, characterized in that it comprises an output means for displaying, recording, or outputting the population or the index word representing the characteristics of the document to be investigated,
In the information analysis of the survey target document, an information analysis report automatic creation method for accurately and automatically creating the characteristics of the survey target document with respect to the comparison target document,
Designating and inputting the survey target document and the comparison target document, and inputting conditions for performing the information analysis;
Selecting a population document consisting of a group of documents similar to the survey target document from the comparison target documents;
Extracting characteristic index terms for the population document of the survey target document;
A method of automatically generating an information analysis report, comprising the step of displaying, recording, or outputting to a communication a population or index word representing the characteristics of the document to be investigated. An information analysis report automatic creation method according to claim 9,
In the process of selecting a population document consisting of a document group similar to the survey target document from the comparison target documents,
An information analysis report automatic creation method characterized in that the selection step includes a step of selecting based on a result of a calculation function for calculating a similarity to a comparison target document. An information analysis report automatic creation method according to claim 10,
In the step of selecting by the result of the calculation step of calculating the similarity to the comparison target document,
An information analysis report automatic creation method characterized in that the calculation step of calculating the similarity rate includes a step of calculating a similarity rate based on a function value of the appearance frequency and document frequency for each index word of each document. An information analysis report automatic creation method according to claim 10,
In the step of displaying, recording, or outputting to the communication, the population or the index word representing the characteristics of the survey target document,
A display step of distributing and displaying the population or the index terms in a map;
A display step for displaying a part of the data of the population or the index word;
The output process includes an output process for automatically or selecting a standard comment according to the content, or inputting or selecting a free comment, and displaying the information analysis report automatically. Method.

Images