JP2010231564A - Patent application evaluation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a patent application evaluation device for rationally and quantitatively evaluating the value of a patent application. <P>SOLUTION: The patent application evaluation device is configured to identify the number of applications similar to an application to be analyzed as the number of similar applications to be analyzed; to identify an analysis increase rate that is an increase per unit period of the applications similar to the application to be analyzed; to identify the number of applications similar to an application to be evaluated as the number of similar applications to be evaluated; to identify an evaluation increase rate that is an increase per unit period of the applications similar to the application to be evaluated; to identify a similar position that is a relative position of the number of similar applications to be evaluated to the number of similar applications to be analyzed; to identify an increase rate position that is a relative position of the evaluation increase rate to a plurality of analysis increase rates; and to evaluate the value of the patent application to be evaluated from the similar position and relative position, and a patent prediction value showing the possibility that the patent application to be evaluated is patented. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a patent application evaluation apparatus that evaluates the value of a patent application.

  In order to obtain a patent right in Japan, an application examination request must be made within three years from the filing date of the patent application (Article 48-3). When an application examination request is made, financial costs and human costs are incurred. Therefore, it is necessary for the applicant to carefully judge whether or not a return corresponding to the cost can be expected and to decide whether or not to request for examination of the application.

JP 2007-265039 A JP 2007-265040 A Japanese Patent Laid-Open No. 2002-15001

  At present, many of the applicants are companies, and in many cases, the person in charge of intellectual property of the company evaluates the value of the patent application based on its own experience and determines whether or not to request an application examination. However, a technique for quantitatively evaluating the value of a patent application has not been established, and it is not easy to determine whether or not to request application examination. The value of a patent application is required not only when deciding whether to request application examination, but also when a notice of reasons for refusal is received or at the time of filing.

  An object of the present invention is to provide a patent application evaluation apparatus that rationally and quantitatively evaluates the value of a patent application.

  In order to solve the above problems and achieve the above object, a patent application evaluation apparatus of the present invention includes a predicted value acquisition unit that acquires a patent predicted value indicating a patentability of a patent application to be evaluated, and a patent application A data acquisition unit that acquires document data from a database that stores document data, and an analysis that is a patent application that is analyzed to contribute to the evaluation of the patent application to be evaluated using the document data acquired by the data acquisition unit An analysis similarity number identifying unit that identifies the number of patent applications that are similar to the analysis target patent application in terms of content as an analysis similarity number, and a document acquired by the data acquisition unit. An analysis increase rate specifying unit for specifying an analysis increase rate, which is the number of increases per unit period of patent applications similar in content to the analysis-target patent application, using the data; Using the literature data acquired by the data acquisition unit, the number of patent applications that are filed earlier than the patent application to be evaluated and similar in content to the patent application to be evaluated are specified as the evaluation similarity number An evaluation for specifying an evaluation increase rate, which is an increase per unit period of a patent application similar in content to the patent application to be evaluated, using the evaluation similarity number specifying unit and the literature data acquired by the data acquisition unit Similar position specification that specifies a relative position that is a relative position of the evaluation similarity number with respect to a plurality of analysis similarity numbers obtained for a plurality of patent applications to be analyzed by the increase rate specifying unit and the analysis similarity number specification unit And the evaluation increase rate relative to the plurality of analysis increase rates obtained for the plurality of analysis-target patent applications by the analysis increase rate specifying unit. An increase rate position specifying unit for specifying an increase rate position, a value evaluation unit for evaluating the value of the patent application to be evaluated, using the predicted patent value, the similar position, and the increase rate position. Prepare.

  In addition, a program for causing a computer to realize the functions of the constituent elements of the patent application evaluation apparatus of the present invention is also an aspect of the present invention.

  The present invention can provide a patent application evaluation apparatus that rationally and quantitatively evaluates the value of a patent application.

It is a block diagram of the patent application evaluation system of this Embodiment. It is a schematic diagram which shows the process which the patent application evaluation apparatus of this Embodiment performs. It is a flowchart for demonstrating grouping of a final application. It is a figure which shows an example of an analysis result. It is a figure which shows an example of the determination method of the evaluation value of a verification application. It is a block diagram of the patent application evaluation apparatus of this Embodiment. It is a figure which shows an example of the structure of the data which the data holding part of the patent application evaluation apparatus of this Embodiment hold | maintains. It is a figure which shows an example of the structure of the data which the data holding part of the patent application evaluation apparatus of this Embodiment hold | maintains. It is a graph which shows the relationship between the number of preceding similarities and the amount of preceding information, and the examination result when a group is not considered. It is a graph which shows the relationship between the number of preceding similarities and the amount of preceding information, and the examination result in consideration of the group. It is a graph which shows the relationship between an unpublished similar application and a published similar application. It is a flowchart which shows the process of a patent possibility analysis phase. It is a flowchart which shows the process in a patentability prediction phase. It is a flowchart which shows the process of an evaluation analysis phase. It is a flowchart which shows the process in a comprehensive evaluation phase. It is a figure which shows comprehensive evaluation of a verification application.

  EMBODIMENT OF THE INVENTION Below, the form for implementing this invention is demonstrated with reference to drawings.

  First, a patent application evaluation system including a patent application evaluation apparatus 1 according to the present embodiment will be described with reference to FIG. FIG. 1 is a configuration diagram of the patent application evaluation system of the present embodiment. As shown in FIG. 1, the patent application evaluation system is a system for evaluating the value of a patent application to be evaluated, which includes a patent application evaluation apparatus 1, a client terminal apparatus 2, a patent database 3, and a communication network. 4 and. In the following description, the patent application to be evaluated is referred to as a “verification application”. The verification application may be a pre-application manuscript created for a patent application.

  The patent application evaluation apparatus 1 predicts the possibility of being patented (hereinafter referred to as “patentability”) for a verification application, and relates to whether or not the application is in accordance with a technology trend, and a pioneering invention. It is a device that judges the value by judging whether or not it is an application. The client terminal device 2 is a terminal device having a web browser. The patent application evaluation apparatus 1 and the client terminal apparatus 2 can be realized by a general computer. The patent database 3 is a database that holds document data of patent applications that have been published, and includes patent data of patent applications for which patents or rejections have been finalized (hereinafter referred to as “final applications”), patents or rejections. The document data of the patent application (hereinafter referred to as “unconfirmed application”) that is not finalized is stored. The document data is not limited to the document itself attached to the application such as the claims, but also includes accompanying bibliographic information such as progress information. The patent database 3 may be an existing database such as “Patent Electronic Library”. The communication network 4 is a network for communicating data such as the Internet.

  The patent application evaluation device 1, the client terminal device 2, and the patent database 3 are connected to a communication network 4, and the patent application evaluation device 1 and the client terminal device 2 can communicate with each other. The user accesses the patent application evaluation apparatus 1 via the client terminal apparatus 2, and the patent application evaluation apparatus 1 provides various information to the client terminal apparatus 2 for the access. Further, the patent application evaluation apparatus 1 acquires document data from the patent database 3 via the communication network 4.

  The patent application evaluation apparatus 1 analyzes the causal relationship between the application contents and examination results for a finalized application to predict the patentability of a verification application, and analyzes the number of finalized and unconfirmed applications, the number of increases, etc. The value of the verification application is evaluated from the obtained results.

  Next, processing performed by the patent application evaluation apparatus 1 will be described with reference to FIG. FIG. 2 is a schematic diagram showing processing performed by the patent application evaluation apparatus 1. As shown in FIG. 2, the process performed by the patent application evaluation apparatus 1 is roughly divided into an “analysis phase F1” and an “evaluation phase F2”.

  The analysis phase F1 is a phase of analyzing document data in order to evaluate the value of the verification application, and is roughly divided into a “patentability analysis phase F11” and an “evaluation analysis phase F12”. In the patentability analysis phase F11, the causal relationship between the application contents and the examination results for the finalized application is analyzed, and a “patent prediction formula” for obtaining the patentability of the verification application is specified. In the evaluation analysis phase F12, the rate of increase in the number of patent applications similar to the patent application to be analyzed, which is a patent application analyzed to contribute to the evaluation of the verification application, and the patent application similar to the patent application to be analyzed Thus, the number of patent applications filed prior to the patent application to be analyzed and the amount of information of the patent application to be analyzed are analyzed.

  The evaluation phase F2 is a phase for evaluating the value of the verification application using the analysis result in the analysis phase F1, and is roughly divided into a “patentability prediction phase F21” and a “total evaluation phase F22”. In the patentability prediction phase F21, a “patent prediction value” indicating the patentability of the verification application is calculated from the patent prediction formula specified in the patentability analysis phase F11 and the application contents of the verification application. In the comprehensive evaluation phase F22, the value of the verification application is calculated from the patent predicted value and the result obtained in the evaluation analysis phase F12.

  Next, the theoretical background on which the evaluation is based will be described. Thereafter, the analysis phase F1 and the evaluation phase F2 will be described.

  First, the theoretical background will be explained.

  Since the technical scope of the patented invention is defined by the “claims” (Article 70 (1)), the breadth of the claims determines the breadth of rights. Right holders prefer a wide range of claims, but it is generally difficult to obtain a patent with a wide range of claims in the technical field where development competition is intense (hereinafter referred to as “dense technical field”). . On the other hand, in a technical field with few competitors (hereinafter referred to as “sparse technical field”), it is relatively easy to obtain a patent with a broad claim. Based on this empirical rule, the present inventor thought that a statistical correlation could be found between the density of the technical field, the breadth of claims, and the patentability.

  “Similarity (x)” is introduced as an index value indicating the density of the technical field. Among patent applications filed prior to a certain patent application A, a patent application similar in content to the patent application A is defined as “similar application (of patent application A)” and the number of similar applications is defined as “(patent Number of similarities (of application A) (x) ”. It can be said that the greater the number of similarities, in other words, the denser the technical field to which patent application A belongs, the more documents that can be the basis for denying the patentability of patent application A.

  The determination of the similarity between the patent application A and the other patent application B can be performed by applying a known natural language processing technique. For example, if the morpheme group included in claim 1 of patent application A includes a morpheme of a predetermined ratio or more in the specification of other patent application B, it is determined that patent application B is similar to patent application A. To do. Like the TF (Term Frequency) method and the IDF (Inverse Document Frequency) method, the similarity between the patent application A and another patent application B may be determined in consideration of the importance of each morpheme. Similarity may be determined by detecting whether or not the degree of similarity is equal to or greater than a predetermined threshold by using a concept search technique that has already been put into practical use (see Patent Document 3 for details of concept search).

  Next, “information amount (y)” is introduced as an index value indicating the width of the claim. In the present embodiment, the number of morphemes included in a predetermined claim of patent application A, for example, the number of morphemes included in claim 1 is defined as “information amount (y) of (patent application A)”. It is estimated that the greater the amount of information, the narrower the scope of rights, and the smaller the amount of information, the wider the scope of rights.

  Next, “increase rate (f)” is introduced as an index value indicating whether or not the technical field to which the verification application belongs is a trendy field. The increase rate (f) is the increase in the number of patent applications in the technical field per predetermined unit period (for example, one year). By utilizing the increase rate (f), the value of the verification application can be evaluated.

  Next, the analysis phase F1 and the evaluation phase F2 will be described.

  First, the analysis phase F1 will be described. As described above, the analysis phase F1 is a phase for analyzing document data in order to evaluate the value of the verification application, and is roughly divided into a patentability analysis phase F11 and an evaluation analysis phase F12.

  The patentability analysis phase F11 is a phase for identifying the “patent prediction formula” for obtaining the patentability of the verification application by analyzing the causal relationship between the application content and the examination result for the finalized application. The process can be divided into three processes: a process S1, a first analysis material specifying process S2, and a patent foreseeing type specifying process S3.

  First, the first collection process S1 will be described. The first collection process S <b> 1 is a process of acquiring document data of a final application from the patent database 3.

  The patent application evaluation apparatus 1 acquires document data of a final application from the patent database 3 and stores it in a storage area that the patent application evaluation apparatus 1 has. The patent application evaluation apparatus 1 may acquire document data when executing the patentability analysis phase F11, or may acquire document data periodically by batch processing. The user inputs “search conditions” to the patent application evaluation apparatus 1 via the client terminal apparatus 2 for specifying a finalized application to be analyzed. The search conditions are, for example, application date, applicant, IPC (International Patent Classification), examiner, agent, application number, or publication number. The “application date” may be a period including a plurality of consecutive days such as “April 1, 1997 to March 31, 2008”.

  The patent application evaluation apparatus 1 specifies the finalized application to be the analysis target corresponding to the input search condition after holding the document data of the finalized application in the storage area of the patent application evaluation apparatus 1. Note that the input of search conditions by the user is not essential. If the search condition is not input, the patent application evaluation apparatus 1 sets all finalized applications acquired from the patent database 3 as analysis targets. In addition, the patent application evaluation apparatus 1 may obtain a finalized application to be analyzed that corresponds to the search condition from the patent database 3 after the search condition is input by the user, and hold it in a storage area that the patent application evaluation apparatus 1 has. . In the following description, the “final application subject to analysis” is referred to as a “final application” unless otherwise specified.

  Next, the first analysis material specifying process S2 will be described. The first analysis material identification process S2 is a process for identifying the number of similarities and the amount of information for each finalized application and identifying the examination result indicating whether it is patented or rejected.

The patent application evaluation apparatus 1 specifies the number of similarities and the amount of information for each finalized application. For example, when specifying the number of similarities and the amount of information of the finalized application D, the patent application evaluation apparatus 1 is a patent application that has been filed prior to the finalized application D and is similar to the finalized application D. Is identified as “similar application (of finalized application D)” and the number of similar applications is identified as “number of similarities (of finalized application D) (x ₁ )”. In the following description, the similar application of the final application is referred to as “preceding similar application”, and the number of similarities of the final application is referred to as “preceding similar number (x ₁ )”. Next, the patent application evaluation apparatus 1 specifies the information amount of the finalized application D. In the following description, the information amount of the final application is referred to as “preceding information amount (y ₁ )”.

  Finally, the patent application evaluation apparatus 1 indexes the examination result of the finalized application D as “result index value (z)”. In this embodiment, a finalized application whose rejection is finalized with a result index value (z) of “1” as a result of a finalized application whose patent has been finalized (hereinafter referred to as “finalized application (patent)”). The result index value (z) of (hereinafter referred to as “final application (rejection)”) is set to “0”.

In this way, the patent application evaluation apparatus 1 sets the preceding similarity number (x ₁ ), the preceding information amount (y ₁ ), and the result index value (z) as one set for the finalized application D. Hold in the storage area that has The patent application evaluation apparatus 1 also specifies the number of preceding similarities, the amount of preceding information, and the result index value for other confirmed applications. That is, the patent application evaluation apparatus 1 specifies and holds “a set of the preceding similarity number (x ₁ ), the preceding information amount (y ₁ ), and the result index value (z)” for each of the plurality of finalized applications. To do.

  Next, the patent predictive expression specifying process S3 will be described. The patent preview formula specifying process S3 is a process for specifying a patent preview formula for predicting the patentability of the verification application using the result obtained in the first analysis material specifying process S2.

The patent application evaluation apparatus 1 performs multiple regression analysis using log (preceding similarity number (x ₁ )) and the preceding information amount (y ₁ ) as explanatory variables, and the result index value (z) as an objective variable. The base of the logarithm of log (preceding similarity number (x ₁ )) is “10”. That is, the patent application evaluation apparatus 1 sets “z = a · logx + b · y + c” as a regression equation for performing multiple regression analysis, performs multiple regression analysis for a plurality of finalized applications, and performs coefficient a and coefficient b. , And the coefficient c.

In the present embodiment, the patent application evaluation apparatus 1 groups each finalized application as described below with reference to FIG. 3, and sets a regression equation “z = a · logx + b · y + c” for each group. . That is, as shown in FIG. 3, the patent application evaluation apparatus 1 first arranges a plurality of finalized applications in descending order of the number of preceding similarities (x ₁ ) (S100), and in order from the largest number of preceding similarities (x ₁ ). Candidate applications P (j) (k) belonging to the same group are created (S102).

In addition, when performing the process of step S106 mentioned later from step S102 for the first time, the patent application evaluation apparatus 1 performs the process with “j = 1, s = 1”. The patent application evaluation apparatus 1 arranges a plurality of finalized applications in descending order of the number of preceding similarities (x ₁ ) (S100), for example, D (1), D (2), D (3),. , D (n) in order of finalized applications. Specific examples of the candidate P (j) (k) are as follows.

P (j) (1) = {D (s), D (s + 1), D (s + 2)},
P (j) (2) = {D (s), D (s + 1), D (s + 2), D (s + 3)},
...
P (j) (n−s−1) = {D (s), D (s + 1), D (s + 2), D (s + 3),..., D (n)}
Next, the patent application evaluation apparatus 1 overlaps the regression equation “z = a · logx + b · y + c” for each candidate P (j) (k) (k = 1,..., N−s−1) created. Regression analysis is performed, and from the first candidate to the candidate P (j) (K) where the significance F becomes a predetermined value (for example, 0.05) or less and the multiple correlation (R) value becomes maximum. It belongs to one group (S104). For example, a candidate to fourth in order from the larger prior similar number (x _1), prior similar number (x ₁₎ Write result of multiple regression analysis intended for the candidate to the fifth in order from the large In both cases, the significance F was less than or equal to a predetermined value, but the multiple correlation (R) obtained by performing multiple regression analysis on the fifth candidate in order from the largest preceding similarity number (x ₁ ). When the value becomes the maximum, the patent application evaluation apparatus 1 causes the first to fifth candidates from the largest in the number of preceding similarities (x ₁ ) to belong to one group. For example, the candidates from the largest preceding similarity number (x ₁ ) to the fourth candidate are expressed as P (j) (2), and the candidates from the largest preceding similarity number (x ₁ ) to the fifth candidate are , P (j) (3).

Then, the patent application evaluation apparatus 1 uses the log (x ₁ (s)) and log (x ₁ (K + s + 1)) of the finalized applications (D (s), D (K + s + 1)) belonging to the same group. (S106). For example, in the case where up to the fifth candidate in order from the largest preceding similarity number (x ₁ ) belongs to one group, the patent application evaluation apparatus 1 determines the logarithmic value of the leading preceding similarity number (x ₁ ). And the logarithmic value of the fifth preceding similarity number (x ₁ ) are held in its own storage area. That retention is necessary later in determining which group the verification application should belong to. In addition, when the candidate with the largest number of preceding similarities (x ₁ ) to the fifth candidate belongs to one group, P (j) (3) is a candidate that maximizes the value of the multiple correlation R. Because there is, K = 3.

  Next, it is determined whether or not the number of finalized applications that do not belong to any group is 2 or less (S108). If it is 3 or more, the patent application evaluation apparatus 1 determines the values of the variable s and the variable j. To change. Specifically, the patent application evaluation apparatus 1 changes the value of the variable s to “s = K + s + 2” and changes the value of the variable j to “j = j + 1”. In the above example, since K = 3, s = 1, and j = 1, the patent application evaluation apparatus 1 changes the value of the variable s to “s = 5” and sets the value of the variable j to “j = 2”. Change to "". And operation | movement of the patent application evaluation apparatus 1 returns to step S102, and if it is 2 or less, operation | movement of the patent application evaluation apparatus 1 will be complete | finished. In step S102, the patent application evaluation apparatus 1 creates candidates P (j) (k), except for the confirmed applications that already belong to any group.

In this way, the patent application evaluation apparatus 1 uses the log (preceding similarity number (x ₁ )) and the preceding information amount (y ₁ ) for each group specified by the number of preceding similarity numbers (x ₁ ) in the final application. Is an explanatory variable, and a “patent prediction formula” is determined as a regression equation using the result index value (z) as a target variable. The patent application evaluation apparatus 1 holds the determined patent preview formula in a storage area that it owns. The patent application evaluation apparatus 1 may calculate a multiple correlation coefficient and the like together with the patent prediction formula. The greater the multiple correlation coefficient, the higher the reliability of the patent prediction formula. The patent prediction formula for a <0, b> 0, c> 0 is more likely to be patented as the preceding similarity number (x ₁ ) is smaller and the preceding information amount (y ₁ ) is larger (result index value (z) is smaller). This is consistent with the rule of thumb

The above is the description of the patentability analysis phase F11. In addition, the patent application evaluation apparatus 1 uses the log (preceding similarity number (x ₁ )) as the x axis and the prior information amount (y ₁ ) as the y axis after the completion of the execution of the patentability analysis phase F11. May be plotted on a graph by distinguishing whether it is patented or rejected, and the graph may be displayed on the client terminal device 2. The graph will be described later with reference to FIGS. 9 and 10.

  Next, the patentability prediction phase F21 will be described. Before describing the evaluation analysis phase F12, the patentability prediction phase F21 will be described. In the patentability prediction phase F21, as described above, the “patent prediction value” indicating the patentability of the verification application is calculated from the patent prediction formula specified in the patentability analysis phase F11 and the application contents of the verification application. Therefore, it can be divided into two processes of a first verification material specifying process S4 and a prediction process S5.

  First, the first verification material specifying process S4 will be described. The first verification material specifying process S4 is a process of acquiring the reference data of the verification application and specifying the number of similarities and the information amount of the verification application in order to predict the patentability of the verification application.

  The user inputs the document data of the verification application to the patent application evaluation apparatus 1 via the client terminal device 2. Thereby, the patent application evaluation apparatus 1 acquires document data of the verification application. The user does not need to input the entire document data of the verification application. The user may input only a part of the document data of the verification application such as claim 1 or abstract. If the verification application is published, the user may input the verification application publication number or the like, and cause the patent application evaluation apparatus 1 to acquire the verification application document data from the patent database 3.

Next, the patent application evaluation device 1 refers to the number of similarities of the verification application (hereinafter referred to as “validation similarity number (x ₂ )”) and the information amount of the verification application (hereinafter referred to as “verification information amount (y ₂ )”). ). For example, when specifying the verification similarity number (x ₂ ) and the verification information amount (y ₂ ) of a certain verification application E, the patent application evaluation apparatus 1 is a patent application filed before the verification application E. Thus, the patent application similar to the verification application E is identified as “similar application (of verification application E)”, and the number of similar applications is “number of similarities (of verification application E)”, that is, “number of verification similarities (x ₂ )” And the amount of information of the verification application E is specified as the amount of verification information (y ₂ ).

  Next, the prediction process S5 will be described. The prediction process S5 is a process for predicting the patentability of the verification application.

The patent application evaluation apparatus 1 uses the verification similarity number (x ₂ ) and the verification information amount (y) to the variable x and the variable y of the patent prediction formula “z = a · logx + b · y + c” obtained in the patentability analysis phase F11. ₂ ) are respectively substituted to obtain a result index value (z). The calculated result index value (z) is a “patent prediction value” that quantitatively indicates the patentability of the verification application. The patent predicted value is a value that quantitatively indicates the patentability of the verification application based on the result of analyzing the causal relationship between the application content and the examination result for the final application group. It is predicted that the greater the patent predicted value, that is, the closer the patent predicted value is to 1.0, the higher the possibility that the verification application will be patented. The patent application evaluation apparatus 1 holds the calculated predicted patent value in a storage area that it owns. The patent application evaluation apparatus 1 displays the analysis result 180 shown in FIG. 4 on the client terminal apparatus 2, and the patentability of the verification application (patent prediction value), the number of prior applications (number of verification similarities), multiple correlation coefficient, verification The number of similarities, the P value of the verification information amount, etc. may be shown to the user. As a result, the user can change the search condition with reference to the multiple correlation coefficient value, and cause the patent application evaluation apparatus 1 to determine a patent prediction formula having a high multiple correlation coefficient, in other words, a highly reliable patent prediction formula. it can.

  By such processing, the patent application evaluation apparatus 1 calculates the patentability of an unexamined verification application rationally and quantitatively based on the density of the technical field, the breadth of claims and the past examination results. Can do.

  The above is the description of the patentability prediction phase F21.

  Next, the evaluation analysis phase F12 will be described. As described above, the evaluation analysis phase F12 includes a rate of increase in the number of patent applications similar to the patent application to be analyzed, and a patent application similar to the patent application to be analyzed. This is a phase for analyzing the number of patent applications that have been applied and the amount of information of the patent application to be analyzed, and includes three phases: a second collection process S6, a second analysis material identification process S7, and a variance calculation process S8. Can be divided into processing.

  First, the second collection process S6 will be described. The second collection process S6 is a process for acquiring document data from the patent database 3.

  The patent application evaluation apparatus 1 acquires document data from the patent database 3 and stores it in its own storage area, as in the first collection process S1 of the patentability analysis phase F11. The patent application evaluation apparatus 1 may acquire document data when executing the evaluation analysis phase F12, or may periodically acquire document data by batch processing. The user inputs “search conditions” for specifying a patent application to be analyzed to the patent application evaluation apparatus 1 via the client terminal apparatus 2. The search condition is a condition described when the first collection process S1 is described, and is an application date or the like.

  The patent application evaluation apparatus 1 specifies a patent application to be analyzed corresponding to the input search condition after holding the document data in the storage area of the patent application evaluation apparatus 1. Note that the input of search conditions by the user is not essential. If no search condition is input, the patent application evaluation apparatus 1 analyzes all patent applications acquired from the patent database 3. The patent application evaluation apparatus 1 may acquire a patent application to be analyzed that corresponds to the search condition from the patent database 3 after the search condition is input by the user and hold it in a storage area that the patent application evaluation apparatus 1 has. .

  Next, the second analysis material specifying process S7 will be described. The second analysis material specifying process S7 includes, for each patent application to be analyzed, the rate of increase in the number of similar patent applications, and the number of similar patent applications that have been filed before the patent application to be analyzed. This is processing for specifying the number and the amount of information. In the following description, an increase rate that is an increase per unit period of the number of patent applications similar to the patent application to be analyzed is referred to as an “analysis increase rate”. The number of similar patent applications that have been filed before the patent application to be analyzed is referred to as “analysis similarity number”.

  The patent application evaluation apparatus 1 specifies the number of similarities of analysis, the analysis increase rate, and the amount of information for each patent application to be analyzed. For example, when specifying the number of analysis similarities of a patent application P to be analyzed, the patent application evaluation apparatus 1 is a patent application that has been filed before the patent application P and is similar to the patent application P. Is identified as “analysis similar application of (patent application P)”, and the number of analysis similar applications is identified as “analysis similarity number (e) of (patent application P)”.

  When the analysis increase rate of the patent application P is specified, the patent application evaluation device 1 is a patent application that has been filed after the patent application P and is similar to the patent application P. The number of increases per unit period (for example, “1 year”) of similar applications of the patent application P is specified as the analysis increase rate (f). In the present embodiment, the patent application evaluation apparatus 1 divides the analysis increase rate (f) by the analysis similarity number. According to the inventor's investigation, the analysis increase rate (f) is high if the number of patent applications similar to the patent application P is large when the patent application P is filed, and the number of patent applications similar to the patent application P when the patent application P is filed is high. It was found that the analysis increase rate (f) was low when the number of cases was small. That means that if the analysis increase rate (f) is simply used as a comparative material, it cannot be accurately determined whether or not the technical field to which the patent application P belongs is a current field. Therefore, in order to know with accuracy whether the technical field to which the patent application P belongs is a trendy field, the patent application evaluation apparatus 1 divides the analysis increase rate (f) by the analysis similarity number (e). The analysis increase rate for evaluation (f ′) is obtained.

  The patent application evaluation apparatus 1 divides the analysis increase rate (f) by the sum of the analysis similarity number (e) and all applications similar to the patent application P filed after the filing of the patent application P. The analysis increase rate for evaluation (f ′) may be obtained. Further, the patent application evaluation apparatus 1 analyzes all applications similar to the patent application P filed after the filing of the patent application P into the analysis similarity number (e) and the patent application P filed after the filing of the patent application P. By dividing by the sum of all similar applications, the analysis increase rate for evaluation (f ′) may be obtained.

  Next, the patent application evaluation apparatus 1 specifies the information amount of the patent application P. In the following description, the information amount of the patent application to be analyzed in the patentability prediction phase F21 is referred to as “analysis information amount (g)”. When the disposal of the patent application to be analyzed is unconfirmed, the patent application evaluation apparatus 1 multiplies the analysis information amount (g) obtained simply by the correction coefficient and uses the analysis information amount ( g) may be specified. For example, the correction coefficient is an average value of values obtained by paying attention to a plurality of patent applications and dividing “the number of words in claim 1 when patented” by “the number of words in claim 1 at the time of filing”. is there.

  Next, the variance calculation process S8 will be described. The variance calculation process S8 calculates an average and variance for the logarithmic value of the analysis similarity number (e) of each of the plurality of analysis-target patent applications, and also calculates an analysis increase rate for evaluation of each of the plurality of analysis-target patent applications ( f ′) and the reciprocal of the analysis information amount (g) are processes for calculating an average and a variance.

  The patent application evaluation apparatus 1 calculates an average and a variance for the logarithmic value of the analysis similarity number (e) of each of the plurality of analysis target patent applications. The calculation of the average and variance of the logarithmic value of the analysis similarity number (e), not the analysis similarity number (e) itself, indicates the order of the number, not just the size of the number, as a result of the investigation by the present inventors. This is because it has been found that the logarithm better explains the patentability.

  Next, the patent application evaluation apparatus 1 calculates an average and a variance for the evaluation analysis increase rate (f ′) of each of the plurality of analysis target patent applications. Further, the patent application evaluation apparatus 1 calculates an average and a variance for the reciprocal of the analysis information amount (g) of each of the plurality of analysis target patent applications.

  Then, the patent application evaluation apparatus 1 includes the average and variance of logarithmic values of the number of analysis similarities (e) and the average and variance of the analysis increase rate for evaluation (f ′) for each of the plurality of patent applications to be analyzed. The average and variance of the reciprocal of the analysis information amount (g) are held in the storage area that the device itself has.

  The above is the description of the evaluation analysis phase F12. Next, the comprehensive evaluation phase F22 will be described.

  The comprehensive evaluation phase F22 is a phase in which the value of the verification application is calculated from the patent prediction value obtained in the patentability prediction phase F21 and the result obtained in the evaluation analysis phase F12, and the second verification material specifying process S9. And the comprehensive evaluation process S10.

  First, the second verification material specifying process S9 will be described. The second verification material specifying process S9 includes the number of patent applications that are filed earlier than the verification application and have similar contents (hereinafter referred to as “evaluation similarity number”), and the unit period of the patent application similar to the verification application ( For example, a value obtained by dividing the number of increases per “one year” by the number of evaluation similarities (hereinafter referred to as “evaluation increase rate”) and the reciprocal of the information amount are specified, and the average obtained in the evaluation analysis phase F12 And the standard deviation of the verification application using the variance.

  The patent application evaluation apparatus 1 specifies the logarithmic value of the evaluation similarity number, the evaluation increase rate, and the reciprocal of the information amount for the verification application. Then, the patent application evaluation apparatus 1 uses the mean and variance of the logarithm of the analysis similarity number (e) of each of the plurality of analysis target patent applications based on the following formula (1). The standard deviation of the logarithmic value of the evaluation similarity number (hereinafter referred to as “similar number standard deviation (E)”) is calculated.

E = {log (evaluation similarity number of verification application) −log (average of analysis similarity number (e))} / (standard deviation of log (analysis similarity number (e)) (1)
Similarly, the patent application evaluation apparatus 1 verifies the case where the average and variance of the analysis analysis increase rate (f ′) for each of the plurality of analysis-target patent applications are used based on the following equation (2). The standard deviation of the evaluation increase rate of the application (hereinafter referred to as “increase rate standard deviation (F)”) is calculated.

F = {(Evaluation increase rate of verification application) −Average of analysis increase rate for evaluation (f ′)} / Standard deviation of analysis increase rate for evaluation (f ′) (2)
Furthermore, the patent application evaluation apparatus 1 uses the following equation (3) to verify the verification application information when using the mean and variance for the reciprocal of the analysis information amount (g) of each of the plurality of analysis target patent applications. The standard deviation of the reciprocal quantity (hereinafter referred to as “information standard deviation (G)”) is calculated using the reciprocal of the specified information quantity.

G = {(reciprocal of information amount of verification application) −average reciprocal of analysis information amount (g)} / standard deviation of reciprocal of analysis information amount (g) (3)
Then, the patent application evaluation apparatus 1 holds the calculated similarity number standard deviation (E), increase rate standard deviation (F), and information amount standard deviation (G) in its own storage area.

  Next, the comprehensive evaluation process S10 will be described. The comprehensive evaluation process S10 is a process for comprehensively evaluating the value of the verification application.

  The patent application evaluation apparatus 1 stores the patent predicted value (z), the similarity number standard deviation (E), the increase rate standard deviation (F), and the information amount standard deviation (G) held in its own storage area. Substituting it into the evaluation formula “v = α · z + β · E + γ · F + δ · G” calculates the value v that comprehensively indicates the value of the verification application.

  By such processing, the value of verification applications that comprehensively consider whether it is a patentability, whether it is an application according to a technology trend, whether it is an application for a pioneering invention, etc., can be rationally and quantified. Can be calculated automatically. In particular, when deciding whether to request examination of an application for the patent application by three years after the patent application, the applicant or the person in charge of intellectual property when the applicant is a company evaluates the patent application. It is effective for the device 1 to calculate the value v of the patent application. This is because it becomes easy to determine whether or not to request examination based on the value v.

  In the embodiment described above, the patent application evaluation apparatus 1 calculates the increase rate of the number of patent applications filed after the patent application to be analyzed when calculating the analysis increase rate (f). If it is desired to obtain the value v after filing the verification application, such as when determining whether to request application examination, the patent application evaluation apparatus 1 may evaluate the value of the verification application at the time of filing. The rate of increase in the number of patent applications filed prior to the verification application is calculated. Then, the patent application evaluation apparatus 1 obtains an analysis increase rate (f ′) for evaluation by dividing the increase rate of the number of patent applications prior to the verification application by the analysis similarity number (e). Thereby, at the time of filing of the verification application, the value of the verification application including whether or not the verification application is on the trend at the time of filing can be obtained.

  Note that the coefficient α, the coefficient β, the coefficient γ, and the coefficient δ in the evaluation formula “v = α · z + β · E + γ · F + δ · G”, for example, calculate the value v for many verification applications, It is determined experimentally based on the result of comparing and verifying the actual value of each verification application specified by a standard that takes into account the economic effect and the like.

  The evaluation formula may be “v ′ = α · z + β · E + γ · F”.

  Further, the patent application evaluation apparatus 1 may determine the evaluation value v of the verification application without using the evaluation formula. For example, the patent application evaluation apparatus 1 makes the patent predicted value (z) belong to “A rank when it is 0.67 or more and 1 or less”, and when it is “0.33 or more and less than 0.66, B rank And belong to “C rank when it is 0 or more and less than 0.33”. In addition, the patent application evaluation apparatus 1 causes each of the similarity number standard deviation (E), the increase rate standard deviation (F), and the information amount standard deviation (G) to belong to “A rank when it is +1 or more”, If it is “−1 or more and less than +1”, it belongs to “B rank”, and if it is less than “−1”, it belongs to “C rank”. In addition, the patent application evaluation apparatus 1 determines which rank each of the patent predicted value (z), the similarity number standard deviation (E), the increase rate standard deviation (F), and the information amount standard deviation (G) belongs to. May be evaluated comprehensively to determine the evaluation value v.

  Specifically, the patent application evaluation apparatus 1 may determine the evaluation value v of the verification application based on the table shown in FIG. For example, when the patent predicted value (z), the similarity standard deviation (E), the increase rate standard deviation (F), and the information standard deviation (G) all belong to the A rank, the patent application evaluation apparatus 1 May be determined as the highest value “1”. Similarly, when the patent predicted value (z), the similarity number standard deviation (E), the increase rate standard deviation (F), and the information amount standard deviation (G) all belong to the C rank, the patent application evaluation apparatus 1 verifies The value v of the application may be determined as the minimum value “0.2”. Further, any one of patent predicted value (z), similarity number standard deviation (E), increase rate standard deviation (F), and information amount standard deviation (G) belongs to rank B, and the others belong to rank A. In this case, the patent application evaluation apparatus 1 may determine the evaluation value v of the verification application as “0.9”.

  In this way, the patent application evaluation apparatus 1 always uses the patent predicted value (z), the similarity number standard deviation (E), and the increase rate standard deviation (F), and arbitrarily sets the information amount standard deviation (G). May be used to determine the value v. Alternatively, the patent application evaluation apparatus 1 always uses the patent predicted value (z), the analysis increase rate (f), and the analysis similarity number (e), and arbitrarily uses the analysis information amount (g) to determine the value v. May be determined. That is, the patent application evaluation apparatus 1 may or may not use a logarithm or a reciprocal. The patent application evaluation value (z), the analysis increase rate (f), and the analysis similarity number (e) are always used. The value v may be determined by arbitrarily using the analysis information amount (g).

  Next, the configuration of the patent application evaluation apparatus 1 will be described. FIG. 6 is a configuration diagram of the patent application evaluation apparatus 1.

  As shown in FIG. 6, the patent application evaluation apparatus 1 includes an IF (interface) unit 110, a data processing unit 130, and a data holding unit 150.

  The IF unit 110 is in charge of an interface with the client terminal device 2 and the patent database 3. The IF unit 110 displays various information on the client terminal device 2. The data processing unit 130 executes various data processing based on data acquired from the IF unit 110 and the data holding unit 150. The data processing unit 130 also serves as an interface between the IF unit 110 and the data holding unit 150. The data holding unit 150 is a storage area for holding various data.

  The IF unit 110 includes a reception unit 112 and a transmission unit 120. The receiving unit 112 receives search conditions input by the user to the client terminal device 2. The receiving unit 112 acquires document data from the patent database 3. The transmission unit 120 transmits various types of information to the client terminal device 2.

  The receiving unit 112 includes an analysis application acquisition unit 114, a search condition acquisition unit 118, and a verification application acquisition unit 116. The analysis application acquisition unit 114 acquires document data of the patent application to be analyzed from the patent database 3. The search condition acquisition unit 118 acquires a search condition. The verification application acquisition unit 116 acquires document data of the verification application.

  The transmission unit 120 includes a graph display unit 122 and an evaluation result notification unit 124. The graph display unit 122 displays the correlation between the number of similarities (x) and the amount of information (y) and patent / rejection on a graph. Details of the graph will be described later with reference to FIGS. 9 and 10. The evaluation result notifying unit 124 notifies the client terminal device 2 of the value of the verification application including the patent predicted value and the like. The evaluation result notification unit 124 may display the analysis result 180 illustrated in FIG. 4 on the screen of the client terminal device 2.

  The data processing unit 130 includes an analysis unit 132, a prediction / evaluation unit 140, and a classification unit 148. The classification unit 148 groups the patent application or verification application to be analyzed according to the number of similarities. The grouping by the classification unit 148 will be described later with reference to FIG.

  The analysis unit 132 includes a preceding information amount specifying unit 134, a preceding similarity number specifying unit 136, a patent prediction formula determining unit 138, an analysis information amount specifying unit 200, an analysis similarity number specifying unit 202, and an analysis increase rate specifying unit. 204, an analysis information amount distribution specifying unit 206, an analysis similarity number distribution specifying unit 208, and an analysis increase rate distribution specifying unit 210.

The prior information amount specifying unit 134 specifies the prior information amount (y ₁ ) of the final application. The preceding similarity number specifying unit 136 specifies the preceding similarity number (x ₁ ) of the final application. The patent preview formula determination unit 138 determines a patent preview formula based on the number of prior similarities (x ₁ ), the amount of prior information (y ₁ ), and the result index value (z) for a plurality of finalized applications. The analysis information amount specifying unit 200 specifies the analysis information amount (g). The analysis similarity number specifying unit 202 specifies the analysis similarity number (e). The analysis increase rate specifying unit 204 specifies the analysis increase rate for evaluation (f ′). The analysis information amount variance specifying unit 206 specifies an average and a variance for the reciprocal of the analysis information amount (g) of each of the plurality of analysis target patent applications. The analysis similarity number variance specifying unit 208 specifies an average and a variance for the logarithmic value of the analysis similarity number (e) of each of the plurality of analysis target patent applications. The analysis increase rate variance specifying unit 210 specifies an average and a variance for the evaluation analysis increase rate (f ′) of each of the plurality of analysis target patent applications.

  The prediction / evaluation unit 140 includes a verification information amount specifying unit 142, a verification similarity number specifying unit 144, a patentability prediction unit 146, an evaluation information amount specifying unit 212, an evaluation similarity number specifying unit 214, and an evaluation increase rate specifying Unit 216, information amount standard deviation calculation unit 218, similarity number standard deviation calculation unit 220, increase rate standard deviation calculation unit 222, and comprehensive evaluation unit 224.

The verification information amount specifying unit 142 specifies the verification information amount (y ₂ ) of the verification application. The verification similarity number specifying unit 144 specifies the verification similarity number (x ₂ ) of the verification application. The patentability prediction unit 146 calculates a patent predicted value of the verification application from the verification similarity number (x ₂ ), the verification information amount (y ₂ ), and the patent prediction formula. The evaluation information amount specifying unit 212 specifies the information amount of the verification application and its reciprocal. The evaluation similarity number specifying unit 214 specifies the evaluation similarity number and the logarithmic value of the verification application. The evaluation increase rate specifying unit 216 specifies an increase number (increase rate) per unit period of a patent application similar to the verification application and an evaluation increase rate that is a value obtained by dividing the number by the evaluation similarity number. The information amount standard deviation calculation unit 218 calculates an information amount standard deviation (G). The similarity number standard deviation calculation unit 220 calculates a similarity number standard deviation (E). The increase rate standard deviation calculation unit 222 calculates the increase rate standard deviation (F). The comprehensive evaluation unit 224 comprehensively evaluates the value v of the verification application.

  Note that the functions of the respective components constituting the IF unit 110 and the data processing unit 130 of the patent application evaluation apparatus 1 are realized, for example, by the cooperation of a computer CPU as a hardware and a computer program. However, the function of each component described above may be realized in any form such as realized by a dedicated circuit.

  The data holding unit 150 is a device (hardware) having a storage area or storage medium such as a semiconductor memory or a hard disk, and characterizes various types of information such as the IPC, publication number, and amount of prior information of the patent application to be analyzed. Hold as information. The structure of data held by the data holding unit 150 will be described with reference to FIGS.

  7 and 8 are diagrams illustrating an example of the structure of data held by the data holding unit 150. FIG.

  After obtaining the document data of the patent application to be analyzed, the analysis application acquisition unit 114 registers the characteristic information of the patent application to be analyzed in the data holding unit 150. The IPC column 160 is the IPC of the patent application to be analyzed, the publication number column 162 is the publication number, the application date column 164 is the filing date or priority date, the publication date column 166 is the application publication date, and the applicant column 168 is the application Each person's name is shown. The preceding information amount column 170 indicates the amount of preceding information to be analyzed, the preceding similarity number column 172 indicates the preceding similarity number, and the earliest similar application date column 174 indicates the earliest application date (or earliest priority date) of the preceding similar applications. ) Respectively. The examination result column 176 shows the examination result of the final application. A circle indicates that it has been patented, and a cross indicates that it has been rejected. In addition to these, the data holding unit 150 may hold the application number, the inventor's name, the name of the invention, claim 1, and the summary as characteristic information.

The data shown in FIG. 7 is an example of data held in the storage area of the patent application evaluation apparatus 1, that is, the data holding unit 150 in the patent possibility analysis phase F11. In the example of the data structure shown in FIG. 7, Japanese Patent Application Laid-Open No. 2008-054324X is a publication of a final application filed on August 24, 2007 with August 24, 2006 as a priority date. The final application is patented with a preceding information amount (y ₁ ) of “51” and a preceding similarity number (x ₁ ) of “104”.

  The data shown in FIG. 8 is an example of data held in the storage area of the patent application evaluation apparatus 1, that is, the data holding unit 150 in the evaluation analysis phase F12. In the example of the data structure shown in FIG. 8, Japanese Patent Application Laid-Open No. 2001-22222X is a patent application filed on February 1, 2000. In the patent application, the analysis information amount is “30”, the analysis similarity number is “100”, and the disposal is not finalized. The number of similar applications of the patent application is “200” until 1.5 years have passed since the filing of the patent application, “300” until 3 years have passed, and 5 years have passed. Up to “500”, and up to 10 years, “800”.

FIG. 9 is a graph showing the relationship between the number of preceding similarities (x ₁ ) and the amount of preceding information (y ₁ ) when the group is not considered, and the examination results. The group will be described later with reference to FIG.

The horizontal axis indicates the logarithmic value of the preceding similarity number (x ₁ ), and the vertical axis indicates the preceding information amount (y ₁ ). A circle indicates a finalized application (patent), and a cross indicates a finalized application (rejection). FIG. 9 is a graph created by the present inventor for schematically showing the relationship between the number of preceding similarities (x ₁ ) and the amount of preceding information (y ₁ ) and the examination results based on the survey results. As a result of the inventor's investigation, it was possible to read a statistical tendency that an application with a larger amount of prior information (y ₁ ), that is, an application with a narrower scope of rights, is more likely to be patented. It was also possible to read the statistical tendency that the smaller the number of similarities in advance, in other words, the more sparse technical field the more patented.

When the patent prediction formula “z = a · log (x) + b · y + c” is transformed with respect to y, “y = − (a / b) · log (x) + (z−c) / b” is obtained. Since the coefficient a, the coefficient b, and the coefficient c have already been calculated by the patentability analysis phase F11, y where the patent expected value (z) is 0.5 or more is
y ≧ − (a / b) · log (x) + (0.5−c) / b
Exists in the area. This region is a region where patenting can be expected with a probability of 50% or more, and is hereinafter referred to as an “A (Allowance) region”. On the other hand, y that the patent expected value (z) is less than 0.5 is
y <-(a / b) .log (x) + (0.5-c) / b
Exists in the area. This area is an area that can be patented only with a probability of less than 50%, and is hereinafter referred to as an “R (Rejection) area”. A straight line “y = − (a / b) · log (x) + (0.5−c) / b” indicating a boundary line between the A region and the R region is referred to as a “B (Boundary) straight line”.

  After the patentability analysis phase F11 ends, the transmission unit 120 may transmit a graph obtained by processing the A region, the R region, and the B straight line to the client terminal device 2 so that the graph is displayed on the screen. For example, the graph display unit 122 of the transmission unit 120 colors the A region in blue, the R region is colored in red, and the evaluation result notification unit 124 of the transmission unit 120 transmits the colored graph to the client terminal device 2. It may be displayed on the screen.

  By such an aspect, the user can visually grasp the correlation between the number of similarities and the amount of information and the patentability. By setting various search conditions, patent trends by technical field, applicant, agent, examiner and period can be visualized. Note that the threshold of the expected patent value (z) that determines the B straight line is not limited to 0.5, and can be arbitrarily set by the user.

The present inventor actually verified the validity of the patent prediction formula for 11 finalized applications extracted under predetermined conditions. Eight out of eleven are patented and three are rejected. First, the number of prior similarities (x ₁ ) and the amount of prior information (y ₁ ) of each final application were substituted into the patent prediction formula to calculate a patent predicted value (z). As a result, the number of finalized applications with a predicted patent value ≧ 0.5 is 7 (patents: 7 and rejected: 0), and the finalized applications with a predicted patent value <0.5 are 4 (patent: 1). Case, rejection: 3 cases). That is, one out of 11 finalized applications (patents) were determined to be “low patentability (patent predicted value (z) = 0.425)” by the patent prediction formula, but actually patented and predicted And the results do not agree with each other, but the predictions and results agree with each other. As a result, it was confirmed that there is a consistency between the prediction result based on the patent prediction formula and the actual examination result.

FIG. 10 is a graph showing the relationship between the number of preceding similarities (x ₁ ) and the amount of preceding information (y ₁ ) and the examination results in consideration of the group classified as described with reference to FIG. .

It is assumed that the classification unit 148 performs classification as described with reference to FIG. 3 based on the number of published similar applications, and classifies the finalized applications into four groups G1 to G4. The patent preview formula determination unit 138 determines a separate patent preview formula for each group. The horizontal axis indicates the logarithmic value of the preceding similarity number (x ₁ ), and the vertical axis indicates the preceding information amount (y ₁ ). A circle indicates a finalized application (patent), and a cross indicates a finalized application (rejection). FIG. 10 is also a graph created by the present inventor for schematically showing the relationship between the number of preceding similarities (x ₁ ) and the amount of preceding information (y ₁ ) and the examination results based on the investigation results.

  It is assumed that a curve as shown in FIG. 10 (hereinafter referred to as “B curve”) appears when the different patent prediction formulas calculated in G1, G2, G3, and G4 are connected. The A region where the patent expected value (z) is 0.5 or more is a region above the B curve in FIG. 10, and the R region is a region below the B curve.

Regardless of the group, when a single patent prediction formula is calculated from all of the finalized applications to be analyzed, a straight line B shown in FIG. 10 is obtained. The range of the A region and the R region is different between the B curve and the B straight line. Focusing on the A region and the R region based on the B curve, G1 is patented even if the preceding information amount (y ₁ ) is considerably small. In other words, a patent application belonging to a particularly sparse technical field is highly likely to be patented even with a simple claim. On the other hand, in G2, the rate of increase of the preceding information amount (y ₁ ) with respect to the preceding similarity number (x ₁ ) is high. Therefore, in G2, as the number of preceding similarities increases, patenting becomes difficult more rapidly. In G3 and G4, if the amount of prior information (y ₁ ) is at least 60 or more, patenting is almost impossible.

  A triangle mark shown in FIG. 10 indicates the verification application F. The verification application F exists in the R region based on the B line, but exists in the R region based on the B curve. In this way, the patent predicted value of the verification application F changes depending on whether the B straight line or the B curve is adopted. Employing the B curve makes it easier to predict patentability with a higher accuracy than adopting the B straight line.

Note that the final application H (patent) shown in FIG. 10 is patented even though the number of preceding similarities (x ₁ ) is very large and the amount of preceding information (y ₁ ) is small. Therefore, the finalized application H may have a certain right including a reason for invalidity, or may be a strong patent that secures a wide range of rights in a dense technical field. As described above, by referring to the graph shown in FIG. 10, it becomes easy to visually find “a unique patent application in view of statistical tendency”.

The patent application evaluation apparatus 1 calculates the coefficient a, the coefficient b, and the coefficient c of “z = a · log (x) + b · y + c” in the patent predictive expression specifying process S3 in the patentability analysis phase F11. When determining the patent preview formula, the patent preview formula may be determined as follows. That is, the patent application evaluation apparatus 1 includes a plurality of sets of “preceding similarity number (x ₁ ), preceding information amount (y ₁ ), and result index value (z)” for determining a patent prediction formula. Then, n (n ≧ 3) finalized applications in descending order of the number of preceding similarities are set as candidates for determining the patent prediction formula. Then, the patent application evaluation apparatus 1 changes the value of n and performs multiple regression analysis using n candidates for each n to determine a patent prediction formula and calculate a significant F. According to the statistical processing theory, when the significance F is equal to or smaller than a predetermined reference (for example, “0.05”), the reliability of the calculated patent prediction formula becomes high. In addition, the reliability of the calculated patent prediction formula increases as n increases. Therefore, the patent application evaluation apparatus 1 uses the patent prediction formula calculated when the significance F is equal to or less than a predetermined reference and n is the maximum, and the patent prediction formula used when calculating the patent prediction value indicating the patentability. May be determined.

  FIG. 11 is a graph showing the relationship between an unpublished similar application and a published similar application.

  The horizontal axis indicates time, and the vertical axis indicates the number of applications. Let T2 be the filing date of a verification application J. A part of the verification similar application for the verification application J has been published in T2, but a part thereof is not yet published. Let S be the number of verification similar applications published in T2. Hereinafter, a method for obtaining the total number S ′ of verification similar applications including unpublished verification similar applications will be described.

The filing date of the earliest verification similar application K is T1. The verification similar application K is published on T3 1.5 years after T1. Therefore, T3 = T1 + 1.5. Other verification similar applications filed after the filing date of the verification similar application K will be published after T3. Graph C ₁ shown in FIG. 11 indicates the number of verification similar application. On the other hand, Graph C ₂ indicates the number of verification similar application of which was filed published. The number of verified similar applications has increased with the passage of time, and the number of published applications has increased with a time lag of 1.5 years. Graph C ₂ is a graph by translating the graph C ₁ to 1. 5 years to the right. Therefore,
(T2-T3): S = (T2-T1): S ′
The relationship
S ′ = {(T2-T1) / (T2-T3)} · S
It becomes. In this way, the total number S ′ of verification similar applications can be obtained.

  By such processing, the preceding similarity number specifying unit 136 may specify the net similarity number including unpublished similar applications. Then, the classification unit 148 may perform grouping of finalized applications and verification applications including unpublished similar applications. The patent preview formula determination unit 138 may determine the patent preview formula based on the number of previous similarities including unpublished previous similar applications. The patentability prediction unit 146 may calculate a patent prediction value based on the number of verification similarities including unpublished verification similar applications.

  FIG. 12 is a flowchart showing the process of the patentability analysis phase F11.

The analysis application acquisition unit 114 acquires document data from the patent database 3. The document data acquired by the analysis application acquisition unit 114 is stored in the data holding unit 150. The search condition acquisition unit 118 receives an input of search conditions for specifying a confirmed application from the user (S10). The analysis unit 132 refers to the characteristic information or the like, and acquires document data of the finalized application that matches the search condition from the data holding unit 150 (S12). The preceding information amount specifying unit 134 specifies the preceding information amount (y ₁ ) (S14). The preceding similarity number specifying unit 136 specifies the preceding similarity number (x ₁ ) (S16). The preceding similarity number specifying unit 136 may specify the preceding similarity number including the number of unpublished preceding similar applications by the method described with reference to FIG. The classification unit 148 classifies the finalized application into any group as described with reference to FIG. 3 (S18).

  If there is a finalized application that matches the search conditions (Yes in S20), the process returns to S12. If it does not remain (No in S20), the patent preview formula determining unit 138 determines a separate patent preview formula for each group classified by the classifying unit 148 (S22).

  In relation to FIG. 2, S10 and S12 correspond to the first collection process S1, S14, S16, and S18 correspond to the first analysis material identification process S2, and S22 corresponds to the patent predictive expression identification process S3. To do.

  In addition, the graph display part 122 creates the graph which plotted each finalized application in the aspect which can visually recognize the A area | region, R area | region, and B curve shown in FIG. The created graph may be transmitted to the client terminal device 2 for display.

In S12, the analysis unit 132 establishes a final application including specific reasons for refusal, such as Patent Law Article 36 (requirement requirement) and Article 29 paragraph (1) column (invention requirement), among the literature data that matches the search conditions. (Rejection) document data may be excluded. In other words, the analysis unit 132 may analyze only document data of a final application (rejection) and a final application (patent) that do not include a specific reason for refusal such as a description requirement. According to such processing, it is possible to take measures so that the characteristics of the finalized application (rejection) including the reason for refusal that is not related to the preceding similarity number (x ₁ ) like the description requirement are not reflected in the patent prediction formula. . As a result, it is possible to determine a patent prediction formula that more accurately represents the relationship between the number of preceding similarities (x ₁ ) and the amount of preceding information (y ₁ ) and the examination results.

  Further, the analysis unit 132 may exclude the document data of the finalized application that has been rejected without being supported by the applicant after receiving the reason for rejection from the document data that matches the search condition. This is because a patent prediction formula with higher accuracy can be calculated by preventing the characteristics of a finalized application (rejection) whose rejection has been confirmed due to the circumstances of the applicant from being reflected in the patent prediction formula.

  FIG. 13 is a flowchart showing the process in the patentability prediction phase F21.

The verification application acquisition unit 116 acquires document data of the verification application (S30). Here, it is assumed that the verification application acquisition unit 116 has acquired the text data of claim 1 of the verification application. The verification information amount specifying unit 142 specifies the verification information amount (y ₂ ) (S32). The verification similarity number specifying unit 144 specifies the verification similarity number (x ₂ ) (S34). The verification similarity number specifying unit 144 compares the degree of similarity between the text data of claim 1 regarding the verification application and the full text data of the patent application filed prior to the verification application. If the similarity is greater than or equal to a predetermined threshold, the verification similarity number specifying unit 144 specifies the patent application as a verification similar application. The verification similarity number specifying unit 144 may specify the verification similarity number (x ₂ ) including the number of unpublished verification similar applications. The classification unit 148 causes the verification application to belong to any group of G1 to G4 (S36).

  The patentability prediction unit 146 calculates a patent expected value based on the patent prediction formula associated with the group to be classified (S38).

  In relation to FIG. 2, S30, S32, S34, and S36 correspond to the first verification material specifying process S4, and S38 and S40 correspond to the prediction process S5.

  Note that after S38, the graph display unit 122 plots the position of the verification application on the graph shown in FIG. 10, and the evaluation result notification unit 124 transmits the generated graph to the client terminal device 2 to display it. Good.

  The verification application may be a manuscript before being completed as a patent specification. For example, the user can obtain the patent predicted value of the verification application by inputting the data of claim 1 in S30 when only claim 1 is created. For this reason, even before the specification in the patent application is completed, it is easy to reasonably and quantitatively determine whether or not it is a claim that can be expected to be patented by provisional claim 1. Become.

  FIG. 14 is a flowchart showing the process of the evaluation analysis phase F12.

  The analysis application acquisition unit 114 acquires document data from the patent database 3. The document data acquired by the analysis application acquisition unit 114 is stored in the data holding unit 150. The search condition acquisition unit 118 receives an input of search conditions for specifying a patent application to be analyzed from the user (S60). The analysis unit 132 refers to the characteristic information or the like, and acquires document data of the patent application to be analyzed that matches the search condition from the data holding unit 150 (S62). The analysis information amount specifying unit 200 specifies the analysis information amount (g) (S64). The analysis similarity number specifying unit 202 specifies the analysis similarity number (e) (S66). The analysis similarity number specifying unit 202 may specify the analysis similarity number including the number of unpublished similar applications by the method described with reference to FIG. The analysis increase rate specifying unit 204 specifies the analysis increase rate for evaluation (f ′) (S68).

  If there remains a patent application to be analyzed that matches the search conditions (Yes in S70), the process returns to S62. If it does not remain (No in S70), the analysis information amount variance specifying unit 206 calculates the average and variance for the reciprocal of the analysis information amount (g) of each of the plurality of analysis target patent applications (S72). The analysis similarity number variance specifying unit 208 calculates an average and a variance for the logarithmic value of the analysis similarity number (e) of each of the plurality of analysis target patent applications (S74). The analysis increase rate variance specifying unit 210 calculates an average and a variance for the analysis increase rate for evaluation (f ′) of each of the plurality of analysis target patent applications (S76).

  In relation to FIG. 2, S60 and S62 correspond to the second collection process S6, S64, S66, and S68 correspond to the second analysis material specifying process S7, and S72, S74, and S76 are distributed. This corresponds to the calculation process S8.

FIG. 15 is a flowchart showing the process in the comprehensive evaluation phase F22. The verification application acquisition unit 116 acquires document data of the verification application (S80). Here, it is assumed that the verification application acquisition unit 116 has acquired the text data of claim 1 of the verification application. The evaluation information amount specifying unit 212 specifies the information amount of the verification application and specifies the reciprocal of the information amount of the verification application (S82). The evaluation similarity number specifying unit 214 specifies the evaluation similarity number of the verification application, and specifies the logarithmic value of the evaluation similarity number (S84). The evaluation similarity number specifying unit 214 compares the degree of similarity between the text data of claim 1 for the verification application and the full text data of the patent application filed prior to the verification application, and then the evaluation similarity number of the verification application. Specify the logarithm of. The evaluation increase rate specifying unit 216 specifies an increase number (increase rate) per unit period of a patent application similar to the verification application, and specifies an evaluation increase rate that is a value obtained by dividing the increase rate by the evaluation similarity number ( S86). The information amount standard deviation calculation unit 218 calculates the information amount standard deviation (G) (S88). The similarity number standard deviation calculation unit 220 calculates the similarity number standard deviation (E) (S90). The increase rate standard deviation calculation unit 222 calculates the increase rate standard deviation (F) (S92). The comprehensive evaluation unit 224 comprehensively evaluates the value v of the verification application (S94). The evaluation result notifying unit 124 notifies the client terminal device 2 of the comprehensively evaluated value of the verification application (S96). For example, the evaluation result notification unit 124 notifies the client terminal apparatus 2 of the comprehensive evaluation of the verification application shown in FIG. FIG. 16 is a diagram showing a comprehensive evaluation of a verification application.

  In relation to FIG. 2, S80, S82, S84, S86, S88, S90, and S92 correspond to the second verification material specifying process S9, and S94 and S96 correspond to the comprehensive evaluation process S10.

  The verification application may be a manuscript before being completed as a patent specification. For example, the user can acquire the total value of the verification application by inputting the data of claim 1 in step S80 when only the claim 1 is created. For this reason, even before the specification in a patent application is completed, provisional creation of claim 1 makes it easy to judge the value of the claim reasonably and quantitatively.

  The patent application evaluation device 1 has been described above based on the embodiment.

  After a patent application is filed, the value of the patent application is not known, and it may be difficult to determine whether or not the application examination should be requested for the patent application. However, as described above, the patent application evaluation apparatus 1 reasonably and quantitatively calculates the value of the patent application. Therefore, the IP person in charge when the applicant himself or the applicant is a company can easily determine whether the patent application should be requested for examination by referring to the value calculated by the patent application evaluation device 1. Judgment can be made. In addition to determining whether a request for examination of an application should be made, there may be cases where it is desired to know the value of a patent application when receiving a notice of reasons for refusal or before filing an application. This is because knowing the value makes it easier to determine a countermeasure for the patent application. Since the patent application evaluation apparatus 1 reasonably and quantitatively calculates the value of the verification application, it is also effective in such a case. That is, the patent application evaluation device 1 can support the judgment of a person who wants to know the value of the patent application.

  In addition, it is often necessary to predict patentability before knowing the value after patent application. For example, when a notice of reasons for refusal is received, it is necessary to determine how much correction is necessary. When a decision to reject is received, it is necessary to decide whether to request a trial against decision of rejection or to give up patenting. Such judgments are often influenced by personal factors such as the experience and personality of the IP personnel. The patent application evaluation apparatus 1 rationally indicates the patentability of a verification application based on objective data such as a causal relationship between the application contents of a past finalized application and examination results. Thereby, the patent application evaluation apparatus 1 can support the determination by the person in charge of intellectual property.

  The user can make the patent application evaluation apparatus 1 create various patent preview expressions by changing the finalized application according to the search condition. For example, if the patent application evaluation apparatus 1 is made to create a patent preview formula for each applicant, the user can objectively evaluate the skill of the application strategy for each applicant. If the patent application evaluation apparatus 1 is made to create a patent preview formula for each application time, the user can easily grasp the change in examination tendency over time. If the patent application evaluation apparatus 1 creates a patent prediction formula for each country, the user can easily grasp the examination tendency for each country. In addition, the amount of information in claims may differ greatly between software-related patent applications and electrical-related patent applications. Even in the same software system, the amount of information in claims may differ greatly between a patent application relating to a business method and a patent application relating to device control. By causing the patent application evaluation apparatus 1 to create a patent prediction formula for each technical field, it is possible to derive an appropriate patent predicted value for each technical field. Furthermore, by operating the search condition while confirming the multiple correlation coefficient based on the analysis result 180, it is possible to cause the patent application evaluation apparatus 1 to obtain an appropriate patent prediction formula.

  Further, according to the inventor's investigation, it has been found that the patentability changes nonlinearly according to the number of published similar applications. Therefore, a B curve is obtained by grouping patent applications according to the number of similarities and determining a patent prediction formula for each group. If the B curve is used, the accuracy of the patent predicted value can be further improved.

  The patent application evaluation apparatus 1 supports the judgment of the person in charge of intellectual property in various situations such as patent application, request for examination of application, and reason for refusal, etc., so that a high quality invention is more easily protected and utilized. Is expected to be. Further, in view of the number of similarities, the patent application evaluation apparatus 1 can support the judgment of the person in charge of intellectual property even when paying attention to and confirming whether the claim is excessively wide or narrow. .

  In the present embodiment, the examination result value z is described as being a binary value of 0 (rejection) or 1 (patent). As a modification, the examination result value z is assigned three or more values such as 0.1 (decision of rejection), 0.2 (decision of rejection), 0.7 (decision of patent), 1.0 (patent assessment). May be. The final application that has been decided by the Patent Trial is a patent application that has been decided by refusal unless an appeal against appeal of refusal is requested. Based on such a concept, an examination result value lower than the patent appraisal may be assigned to the patent appeal decision.

  On the other hand, a finalized application that has been decided by refusal may have been patented if a trial for decision of refusal was requested. Based on this way of thinking, a final examination application with a rejection decision may be given a higher examination result value than a final application with a rejection decision. The examination result value may be given an optimum value by analyzing many finalized applications.

  In the present embodiment, the analysis-target patent application used when determining the patent prediction formula is described as a patent application (final application) for which a patent or rejection has been finalized. The disposal need not be finalized. For example, a patent application that is the object of a reason for refusal based on novelty may be reflected in the patent prediction formula on the assumption that refusal will be confirmed in the future. In addition, a patent application that is in the process of being appealed against a decision to reject may be reflected in the patent prediction formula, assuming that the rejection has been finalized.

  In the present embodiment, the number of morphemes included in claim 1 is described as “information amount”. As a modification, the number of characters and the number of lines included in claim 1 may be regarded as “amount of information”. In that case, the amount of information may be specified excluding general morphemes such as “above”, “apparatus”, and “step”. This is because such general morphemes are considered to have low information value in specifying the scope of rights of the invention. Moreover, when the same morpheme appears in Claim 1 several times, you may count as 1 time. For example, “... determining the similarity between the first patent application and the second patent application according to the number of morphemes contained in common in the first patent application and the second patent application. In the claim including the word “Yes”, the morpheme “patent” appears four times. In that case, the morpheme “patent” is counted as one time. Since the same morpheme often appears repeatedly in the claim, such a processing method makes it easy to specify the substantial amount of information in the claim.

  Further, in the present embodiment, it has been described that the preceding similarity number, the verification similarity number, the analysis similarity number, and the evaluation similarity number are specified based on claim 1. As a modification, the number of verification similarities may be specified for each claim. With respect to claim 2 dependent on claim 1, after specifying the morpheme number of claim 1 + the morpheme number of claim 2 as the information amount of claim 2, the number of preceding similarities, the number of similarities to be verified, the number of analysis similarities, and The evaluation similarity number may be specified. According to such an aspect, patentability can be determined for each claim of the verification application, and evaluation can also be obtained for each claim.

  Further, the analysis application acquisition unit 114 and the verification application acquisition unit 116 are an example of a data acquisition unit of the patent application evaluation apparatus of the present invention, and the similarity number standard deviation calculation unit 220 is a similar position of the patent application evaluation apparatus of the present invention. It is an example of a specific part. The increase rate standard deviation calculation unit 222 is an example of an increase rate position specifying unit of the patent application evaluation device of the present invention, and the comprehensive evaluation unit 224 includes the predicted value acquisition unit and the value evaluation unit of the patent application evaluation device of the present invention. It is an example. The information amount standard deviation calculating unit 218 is an example of an information amount position specifying unit of the patent application evaluation device of the present invention, and the preceding similarity number specifying unit 136 is an example of a disposal result specifying unit of the patent application evaluation device of the present invention. is there. The patentability prediction unit 146 is an example of a patent predicted value calculation unit of the patent application evaluation apparatus of the present invention.

  DESCRIPTION OF SYMBOLS 1 Patent application evaluation apparatus, 2 Client terminal device, 3 Patent database, 4 Communication network, 110 IF part, 130 Data processing part, 150 Data holding part, 112 Receiving part, 120 Transmission part, 114 Analysis application acquisition part, 118 Search conditions Acquisition unit 116 Verification application acquisition unit 122 Graph display unit 124 Evaluation result notification unit 132 Analysis unit 140 Prediction / evaluation unit 148 Classification unit 134 Preceding information amount specifying unit 136 Preceding similarity number specifying unit 138 Patent Predictive formula determination unit, 200 analysis information amount specification unit, 202 analysis similarity number specification unit, 204 analysis increase rate specification unit, 206 analysis information amount variance specification unit, 208 analysis similarity number variance specification unit, 210 analysis increase rate variance specification unit, 142 verification information amount specifying unit, 144 verification similarity number specifying 146 patentability prediction unit, 212 evaluation information amount specifying unit, 214 evaluation similarity number specifying unit, 216 evaluation increase rate specifying unit, 218, information amount standard deviation calculating unit, 220 similarity number standard deviation calculating unit, 222 increasing rate standard deviation Calculation part, 224 Comprehensive evaluation part.

Claims (6)

A predicted value acquisition unit that acquires a patent predicted value indicating the possibility of patent application of the patent application to be evaluated;
A data acquisition unit for acquiring document data from a database holding document data of the patent application;
Using the literature data acquired by the data acquisition unit, the patent application to be analyzed is a patent application that is analyzed to contribute to the evaluation of the patent application to be evaluated, and the patent to be analyzed An analysis similarity number identifying unit for identifying the number of patent applications similar in content to the application as an analysis similarity number;
Using the literature data acquired by the data acquisition unit, an analysis increase rate specifying unit that specifies an analysis increase rate that is the number of increases per unit period of patent applications similar in content to the patent application to be analyzed,
Using the literature data acquired by the data acquisition unit, the number of patent applications that are filed earlier than the patent application to be evaluated and similar in content to the patent application to be evaluated are identified as the evaluation similarity number An evaluation similarity number identifying unit to be
Using the literature data acquired by the data acquisition unit, an evaluation increase rate specifying unit that specifies an increase rate per unit period of a patent application similar in content to the patent application to be evaluated,
A similar position specifying unit for specifying a similar position which is a relative position of the evaluation similar number to a plurality of analysis similar numbers obtained for the plurality of analysis similar patent applications by the analysis similar number specifying unit;
An increase rate position specifying unit that specifies an increase rate position that is a relative position of the evaluation increase rate with respect to a plurality of analysis increase rates obtained for a plurality of analysis target patent applications by the analysis increase rate specifying unit;
A patent application evaluation device comprising: a value evaluation unit that evaluates the value of the patent application to be evaluated using the predicted patent value, the similar position, and the increase rate position. Furthermore,
Using the literature data acquired by the data acquisition unit, an analysis information amount specifying unit for specifying the information amount of a predetermined claim of the patent application to be analyzed as an analysis information amount,
Using the literature data acquired by the data acquisition unit, an evaluation information amount specifying unit that specifies an information amount of a predetermined claim of the patent application to be evaluated as an evaluation information amount,
An information amount position specifying unit for specifying an information amount position that is a relative position of the evaluation information amount with respect to a plurality of analysis information amounts obtained for the plurality of analysis application patent applications by the analysis information amount specifying unit; ,
The patent application evaluation apparatus according to claim 1, wherein the value evaluation unit evaluates the value of the patent application to be evaluated using the information amount position. Furthermore,
A prior information amount identifying unit that identifies the amount of information of a predetermined claim of the finalized application as a prior information amount, using literature data of a finalized application that is a patent application that has been finalized by the data acquisition unit,
A prior similarity number specifying unit that specifies the number of patent applications that are filed earlier than the finalized application and similar in content to the finalized application using the literature data acquired by the data acquisition unit as a previous similar number; ,
Using the literature data acquired by the data acquisition unit, a disposal result identifying unit that identifies the disposal result of the final application,
With the preceding information amount and the preceding similarity number as explanatory variables, the disposal result as a target variable, a regression analysis is performed on a plurality of finalized applications, and the correlation between the preceding information amount and the preceding similarity number and the disposal result is calculated. A patent preview formula determination unit for determining a patent preview formula to be displayed;
The patent application evaluation apparatus according to claim 2, further comprising: a patent prediction value calculation unit that calculates the patent prediction value from the patent prediction formula, the evaluation information amount, and the evaluation similarity number. Furthermore,
A prior information amount identifying unit that identifies the amount of information of a predetermined claim of the finalized application as a prior information amount, using literature data of a finalized application that is a patent application that has been finalized by the data acquisition unit,
A prior similarity number specifying unit that specifies the number of patent applications that are filed earlier than the finalized application and similar in content to the finalized application using the literature data acquired by the data acquisition unit as a previous similar number; ,
Using the literature data acquired by the data acquisition unit, a disposal result identifying unit that identifies the disposal result of the final application,
With the preceding information amount and the preceding similarity number as explanatory variables, the disposal result as a target variable, a regression analysis is performed on a plurality of finalized applications, and the correlation between the preceding information amount and the preceding similarity number and the disposal result is calculated. A patent preview formula determination unit for determining a patent preview formula to be displayed;
Using the literature data acquired by the data acquisition unit, an evaluation information amount specifying unit that specifies an information amount of a predetermined claim of the patent application to be evaluated as an evaluation information amount,
The patent application evaluation apparatus according to claim 1, further comprising: a patent prediction value calculation unit that calculates the patent prediction value from the patent prediction formula, the evaluation information amount, and the evaluation similarity number. The analysis similarity number specifying unit and the evaluation similarity number specifying unit may be configured so that the first patent application and the second evaluation similarity number specifying unit correspond to the number of morphemes included in common in the first patent application and the second patent application. The patent application evaluation apparatus according to any one of claims 1 to 4, wherein the similarity with the patent application is determined. A function for obtaining a predicted patent value indicating the possibility of patent application of the patent application to be evaluated;
A function of acquiring document data from a database holding document data of a patent application;
Using the acquired literature data, the patent application to be analyzed is a patent application that is analyzed to contribute to the evaluation of the patent application to be evaluated, and similar in content to the patent application to be analyzed The ability to identify the number of patent applications to be analyzed as the number of similarities,
Using the acquired literature data, a function for specifying an analysis increase rate that is an increase per unit period of patent applications similar in content to the patent application to be analyzed,
Using the acquired literature data, a function that specifies the number of patent applications that are filed earlier than the patent application to be evaluated and similar in content to the patent application to be evaluated as an evaluation similarity number;
Using the acquired literature data, a function for specifying an evaluation increase rate, which is an increase per unit period, of a patent application similar in content to the patent application to be evaluated;
A function of specifying a similarity position that is a relative position of the evaluation similarity number to a plurality of analysis similarity numbers obtained for a plurality of the analysis-target patent applications;
A function of specifying an increase rate position that is a relative position of the evaluation increase rate with respect to a plurality of analysis increase rates obtained for a plurality of the patent applications to be analyzed;
The program for making a computer implement | achieve the function which evaluates the value of the patent application of the said evaluation object using the said patent prediction value, the said similar position, and the said increase rate position.

Images