CN108629708A

CN108629708A - Build the method and computer readable storage medium of patent value degree evaluation model

Info

Publication number: CN108629708A
Application number: CN201810276652.6A
Authority: CN
Inventors: 陈伟然; 郭永红; 何佳; 辛莹; 王希桢; 姜庭欣; 杨冠梅
Original assignee: Beijing Enjoy Wisdom Technology Co Ltd
Current assignee: Beijing Enjoy Wisdom Technology Co Ltd
Priority date: 2018-03-30
Filing date: 2018-03-30
Publication date: 2018-10-09

Abstract

A kind of method and computer readable storage medium of structure patent value degree evaluation model, the method for the structure patent value degree evaluation model include：Choose the factor for influencing patent value degree and the subparameter for evaluating the factor；The factor is evaluated according to the subparameter, using evaluation result as selected parameter；Patent data sample is chosen according to preset patent value degree evaluation criterion；The judgement schematics of relationship between selected parameter and patent value degree are embodied based on selected patent data Sample Establishing.Objective, accurate patent value degree evaluation model can be targetedly generated, the accuracy of patent value degree evaluation is improved.

Description

Build the method and computer readable storage medium of patent value degree evaluation model

Technical field

This disclosure relates to patent analysis technical field, and in particular to it is a kind of structure patent value degree evaluation model method and Computer readable storage medium.

Background technology

Chinese research person has carried out a large amount of research, basic ideas to the evaluation index and model of Chinese high value patent It is from the three big elements for influencing patent value：Technology, market and law set out, the index system of structure patent value evaluation.It comments Valence Model Calculating Method mainly uses expert graded, stratification, decision tree method and Field Using Fuzzy Comprehensive Assessment etc..Wan little Li and Zhu Xuezhong builds valuation of patent index system from technology, market and right angle, includes altogether 17 evaluation indexes, utilizes stratification Evaluation model is established with fuzzy synthesis theory.2012, by State Intellectual Property Office and Chinese technology business institute combined publication 《Patent value analysis indexes system operation manual》, patent is established from patent law value, technological value and economic value angle Value analysis index includes 18 evaluation indexes altogether.However, in above-mentioned patent value appraisement system influence value index and The determination of its importance mostly come from expert the logic of index significance level is speculated but it is untested it is assumed that its it is subjective because Element is affected.It is therefore desirable to establish more targeted high value valuation of patent system.

Invention content

In view of this, present disclose provides a kind of method of structure patent value degree evaluation model and computer-readable storages Medium can targetedly generate objective, accurate evaluation model, improve the accuracy of patent value degree evaluation.

According to the disclosure in a first aspect, provide it is a kind of structure patent value degree evaluation model method, including：It chooses Influence the factor of patent value degree and the subparameter for evaluating the factor；The factor is evaluated according to the subparameter, Using evaluation result as selected parameter；Patent data sample is chosen according to preset patent value degree evaluation criterion；It is based on Selected patent data Sample Establishing embodies the judgement schematics of relationship between selected parameter and patent value degree.

Preferably, the factor for influencing patent value degree includes the applicant of patent, agency, agent, invention At least one of people, industry and technology, the parameter include correspondingly at least one of the following：To the applicant's of patent Evaluation result；To the evaluation result of the agency of patent；To the procuratorial evaluation result of patent；The inventor's of patent comments Valence result；To the evaluation result of the industry of patent；And the evaluation result of the technology to patent.

Preferably, the method further includes：After choosing patent sample, based on selected patent data Sample Establishing Relation map between selected parameter and patent value degree；And the judgement schematics are built based on the relation map Vertical.

Preferably, described to include according to preset patent value degree evaluation criterion selection patent data sample：Selection meets The patent of preset first condition is as first group of patent data sample；Choose the patent work for meeting preset second preset condition For second group of patent data sample.

Preferably, second group of patent data sample is chosen from first group of patent data sample, and described first Condition includes whole inventions and the utility model being recorded in state's patent database, the second condition include it is following at least One：Action for infringement of a patent occurred and obtains the patent won a lawsuit；The patent for maintaining payment to expire to patent life；It puts into effect simultaneously Obtain the patent of return；It is requested invalid and is judged to maintain effective Chinese patent；Form the patent of industry standard；By with super Cross the patent of preset times；The patent of multinational application.

Preferably, it is described based on selected patent data Sample Establishing embody selected parameter and patent value degree it Between the judgement schematics of relationship include：Calculate distribution of the selected parameter in first group of patent data sample；Selected by calculating Distribution of the parameter in second group of patent data sample；Calculate point of the selected parameter in first group of patent data sample Comparison result between the distribution of cloth and selected parameter in second group of patent data sample；And it ties based on the comparison Relational expression between the selected parameter of fruit fitting and the comparison result, the parameter selected as embodiment and patent value degree Between relationship judgement schematics.

Preferably, it is described based on selected patent data Sample Establishing embody selected parameter and patent value degree it Between the judgement schematics of relationship include：The corresponding patent of each of extracting parameter value；Patent corresponding to each value is carried out special The evaluation of sharp action value, and based on evaluation result come the value of setup parameter to the influence power of patent value degree；Based on each of parameter A value obtains the judgement schematics for embodying relationship between parameter and patent value degree to the influence power of patent value degree.

Preferably, the patent data sample includes the patent data sample of single-piece patent or is converted by more than one piece patent Compound patent patent data sample..

Preferably, selected parameter has multiple, described based on selected by selected patent data Sample Establishing embodiment Parameter and patent value degree between the judgement schematics of relationship include：For each selected parameter, it is based respectively on selected Patent data Sample Establishing embody the sub- judgement schematics of relationship between the parameter and patent value degree；By adding for sub- judgement schematics Quan He, as the judgement schematics for selected all parameters.

Preferably, selected parameter is divided into multigroup, described based on selected by selected patent data Sample Establishing embodiment The judgement schematics of relationship include between the parameter taken and patent value degree：For each each group of selected parameter, difference base The sub- judgement schematics of relationship between this group of parameter and patent value degree are embodied in selected patent data Sample Establishing；Son is commented The weighted sum of valence formula, as the judgement schematics for selected all parameters.

Preferably, the method for the structure patent value degree evaluation model further includes：Receive other parameters；Based on it is described its His parameter is established at least one other model from evaluation patent value degree, by other described models and the evaluation established Patent value degree evaluation model of the formula combinations finally to be built.

According to another aspect of the present disclosure, a kind of computer readable storage medium is provided, for storing instruction, the finger Order makes the method that processor executes above-mentioned structure patent value degree evaluation model when being executed by processor.

Description of the drawings

In order to illustrate more clearly of the technical solution of the embodiment of the present disclosure, simple be situated between will be made to the attached drawing of embodiment below It continues, it should be apparent that, the attached drawing in description below only relates to some embodiments of the present disclosure, rather than the limitation to the disclosure.

Fig. 1 shows the schematic flow diagram of the method for the structure patent value degree evaluation model according to the embodiment of the present disclosure.

Fig. 2A shows that the patent of different claim quantity in the Chinese patent according to the embodiment of the present disclosure is special first The schematic diagram of the probability occurred in sharp data sample.

Fig. 2 B show that the patent of different claim quantity in the Chinese patent according to the embodiment of the present disclosure is special second The schematic diagram of the probability occurred in sharp data sample.

Fig. 2 C show the patent of different claim quantity in the Chinese patent according to the embodiment of the present disclosure in the first He The comparison of the probability occurred in second patent data sample.

Fig. 3 A show that the patent of different claim quantity in the United States Patent (USP) according to the embodiment of the present disclosure is special first The probability occurred in sharp data sample.

Fig. 3 B show that the patent of different claim quantity in the United States Patent (USP) according to the embodiment of the present disclosure is special second The probability occurred in sharp data sample.

Fig. 3 C show the patent of different claim quantity in the United States Patent (USP) according to the embodiment of the present disclosure in the first He The comparison of the probability occurred in second patent data sample.

Fig. 4 shows Chinese high value patent and common monopoly claim distributed number situation relationship matched curve.

Fig. 5 shows the schematic flow diagram of the patent value degree evaluation method according to the embodiment of the present disclosure.

Fig. 6 shows the schematic diagram of the comprehensive evaluation result of the patent value degree according to the embodiment of the present disclosure.

Fig. 7 A show first to the 18th China of patent value degree evaluation method pair using the embodiment of the present disclosure The action value analysis table of the prize-winning patent of patent prize.

Fig. 7 B show the patent value degree evaluation method with the embodiment of the present disclosure to ETSI Chinese Industrial Standards (CIS) necessity patents Action value analysis table.

Fig. 7 C are shown with the comprehensive value according to the patent value degree evaluation method of the embodiment of the present disclosure to single-piece patent The displaying of degree evaluation and the evaluation of each dimension patent valence degree.

Fig. 8 A show that carrying out patent value degree using patent value degree evaluation method according to an embodiment of the present disclosure comments The schematic diagram of the user interface of valence.

Fig. 8 B show that the high value filtered out using patent value degree evaluation method according to an embodiment of the present disclosure is special The schematic diagram of the patent distribution of the applicant of profit.

Specific implementation mode

To keep the purpose, technical scheme and advantage of the embodiment of the present disclosure clearer, below in conjunction with the embodiment of the present disclosure Attached drawing, clear, complete description is carried out to the technical solution of the embodiment of the present disclosure.Obvious described embodiment is the disclosure A part of the embodiment, instead of all the embodiments.Based on described embodiment of the disclosure, ordinary skill people The every other embodiment that member is obtained under the premise of without creative work belongs to the range of disclosure protection.

Present disclose provides a kind of method, patent value degree evaluation method and the calculating of structure patent value degree evaluation model Machine readable storage medium storing program for executing, the method for not using expert estimation, but each parameter is obtained respectively using the method for big data analysis Influence to patent value obtains patent value degree model to establish the judgement schematics of patent value degree.On the one hand can have Objective, accurate evaluation model is pointedly generated, the accuracy of patent value degree evaluation on the other hand can be improved.

In step S101, the parameter for influencing patent value degree is chosen.

Following table 1 gives the example for the parameter for influencing patent value degree.It can be chosen from table 1 as needed desired Parameter.To ensure the comprehensive of model foundation, at the beginning of establishing patent value degree evaluation model, patent will likely should be influenced as possible All kinds of factors of action value are all found out, and also include according to Chinese patent wherein both having included the international evaluation index generally used Conservation status feature sums up the index come, and has both included the direct parameter directly related with patent itself, also include with applicant, Agency and the relevant indirect parameter of patent technical field.

Table 1

In step S102, patent data sample is chosen according to preset patent value degree evaluation criterion.

In this step, the patent for meeting preset first condition can be chosen as first group of patent data sample, choosing Take the patent for meeting preset second preset condition as second group of patent data sample.Such as it can be according to people to patent valence The general standard of value evaluation chooses the sample of the sample and common monopoly of known high value patent.

Such as the patent sample for meeting the patent of following at least one standard as known high value can be chosen：1. participating in It crosses infringement suit and obtains the substantive patent won a lawsuit；2. the long-lived patent for maintaining payment to expire to patent life；3. patentee Patent is put into effect to by modes such as implementation, license or transfers and is obtained the patent of great number value；4. being suggested invalidation request And it is judged to maintain effective patent；5. being included into standard or by practical application forms de facto standard as the base of industry standard Plinth patent；6. by with the patent more than preset times；7. the patent of multinational application, such as pass through international cooperation treaty PCT, Paris The patent of pact or other modes to multiple national applications；8. other generate the special of significant impact to the technology of industry or market Profit.

On the one hand it is differed greatly compared with the U.S. due to Chinese patent lawsuit development level, no matter from the data sample that can be chosen It cannot all meet analysis demand in this quantity or quality.By taking the maintenance deadline of patent as an example, payment to patent life is maintained to expire Chinese invention patent share 2156 by December, 2017, wherein 20% belongs to Sinopec Group, it is seen that China works as The patent that preceding patent life expires mostlys come from state-owned enterprise, smaller with specific value relation.On the other hand, the missing of related data It will also result in the obstacle of high value patent data sample extraction.For example, patentee can specify patent applicable cases and operation The patent of value lacks enough valid data supports in China and few.Therefore, our state's high value patent numbers in selecting It when according to sample, should proceed from the reality, theory cannot be indiscriminately imitated.

As signal, the present embodiment is invalid and be maintained effective 2305 with the generation chosen in Chinese patent database Patent of invention analyzes sample as simplified high value patent.In contrast, Chinese patent is suggested invalidation request and is judged to tie up It is sufficient to hold effective patent numbers sample, spends manpower financial capacity to lift patent and illustrates that the patent has been applied in the industry in vain Or certain influence is produced on industry, the value of patent can be confirmed from side.

And common monopoly sample as a comparison, it can directly choose the whole inventions and practicality of Chinese patent database New patent, statistical data reflect the distribution situation of claim quantity in full library patent comprehensively.

Certainly, embodiment of the disclosure is without being limited thereto, and the selection criteria of patent data sample can be adjusted as required by, Such as whole patents of some applicant can be chosen as common value patent data sample, and from the applicant all specially The patent for meeting preset condition is chosen in profit as high or low value patent data sample.In some embodiments, high or low valence The selection of value patent data sample can not also be limited to commonly to be worth patent data range of the sample, can also be out of other It is chosen.

In addition, above-mentioned patent data sample can be the data sample about common single-piece patent, it can also be special by more than one piece Profit is processed into the data sample of the compound patent of single-piece patent form.For example, to generally include more than one piece associated with each other for patent families Patent, such as based on identical priority respectively to the patent of multiple national applications.It in some cases, can be in the database The characteristic parameter of each patent in patent families is integrated, to be converted to the form of single-piece patent, in order to retrieving and Analysis.For example, the applicant for the single-piece patent being converted to by patent families can be each applicant of patent families Intersection, claim number can be the sum of claim number of each patent families or average value etc..The present embodiment Patent data sample cover such case, such as this patent families extraction for being converted into single-piece patent form can be directed to Patent data sample evaluates the action value of patent families to establish patent value degree evaluation model.In this case, by It is had differences in the relationship etc. of unitary patent and patent family between parameter, weight and parameters, the present embodiment can It can be on the basis of the patent family data to the whole world be counted, is analyzed with the action value evaluation model for patent families Build, rather than to single patent value scoring simple cumulative in family or by patent family all patents it is corresponding Parameter evaluated according to the action value evaluation model of single patent after simply adding up.

In step S103, the relation map of selected parameter and patent value degree is established.

After the analysis sample of high value patent determines, so that it may be calculated one by one according to the previously selected evaluation parameter list in front The incidence relation of itself and patent value degree.In the present embodiment, claim quantity is chosen as the ginseng for influencing patent value degree Number, illustrates as example.

According to the regulation of Patent Law, the protection domain of invention or utility model patent right is with the content of its claim It is accurate.In fact, each the scope of the claims is by verbal description, it is difficult to specifically quantify.Although the number of claim Amount cannot directly reflect the range of patent requirements protection, but the quantity of claim can reflect patent requirements to a certain extent Protection domain and right stability.On the one hand, claim quantity is more, may illustrate that the technical solution of patent is more multiple Miscellaneous or invention content is more abundant.Wherein independent claims quantity is more, illustrates to regard to what same inventive concept was protected Angle is bigger, or more according to the technical solution that this inventive concept is derived, and comparatively the protection domain of patent is also It is bigger；Dependent claims quantity is more, also illustrates that the technical solution that corresponding independent claims are derived is more or right The improvement of technical solution is more careful, untenable in independent claims, can retain the protection of bigger as far as possible Range maintains the stability of right as far as possible.On the other hand, as previously mentioned, according to the regulation of patent law of china, claim number When amount is more than 10, additional fee is needed in application, meanwhile, agent props up to write the labour paid in the process more The patent agency expense paid may also can increase, and cause the cost that applicant pays when applying for a patent is also opposite to improve. Therefore, it theoretically analyzes, the patent more than claim quantity, the probability for forming high value patent also can bigger.

In order to verify above-mentioned theory, illustrated by taking Chinese patent and United States Patent (USP) as an example below.

In this step, the distributed number situation that can be directed to claim establishes pair of high value patent and common monopoly Compare relation table.Calculate separately high value patent sample and full library patent (note：The full library patent used in the present embodiment include by Chinese whole patent applications and whole utility model patents before 25 days October in 2017) in claim number be equal to n (n be big In the natural number equal to 1) patent numbers and patent numbers ratio shared in high value patent and common monopoly respectively Example.

Fig. 2A shows that the patent of different claim quantity in the Chinese patent according to the embodiment of the present disclosure is special first The schematic diagram of the probability occurred in sharp data sample (common monopoly).Fig. 2 B show the Chinese patent according to the embodiment of the present disclosure The schematic diagram for the probability that the patent of middle difference claim quantity occurs in the second patent data sample (high value patent sample). Fig. 2 C show the patent of different claim quantity in the Chinese patent according to the embodiment of the present disclosure in the first and second patents The comparison of the probability occurred in data sample.

It no matter can be seen that from Fig. 2A and 2B for common monopoly, or for high value patent, when right requested number When being incremented by since 1, icon, which is presented, first rises downward trend again.Only common monopoly is when claim quantity is more than 3 It begins to have dropped, and high value patent, it is just begun to decline after claim quantity is more than 10.

Next the probability that the high value patent that Patent right requirement quantity is n occurs in high value patent is calculated separately The probability multiple occurred in full library for the common monopoly of n with Patent right requirement quantity, obtains relationship between the two, such as schemes Shown in 2C.It can be seen that from Fig. 2 C when right requested number is less than or equal to 6, the claim quantity accounting of high value patent Less than the quantity of common monopoly, when right requested number is more than 6, the quantity accounting of high value patent is totally higher than common valence Value, and the trend that the multiple being higher by linearly increases substantially.Illustrate for common monopoly, the right of high value patent is wanted Ask quantity relatively more, and claim quantity is bigger, the probability that high value patent occurs is also bigger, this and our originally pre- Phase trend is consistent.

More clearly to indicate the relationship of patent value and claim quantity, we can be simply quasi- by the above scatter plot The linearity curve of synthesizing one-dimensional：Y=0.085x+1.0695, slope 0.085 then represent claim quantity this parameter pair The degree of association of patent value degree.

In fact, the trend of linear increase is not presented always with the relationship of claim quantity for the value of patent.From upper Figure is as can be seen that when right requested number is less (when such as claim quantity is less than 10), and patent value is with claim number The variation tendency steeper increased is measured, and continues to increase with claim quantity, influence meeting of the quantity to patent value It gradually reduces and tends towards stability.Therefore, when building the relationship of patent parameter and patent value, we can provide one it is more multiple Miscellaneous accurate fitting strategy.

It can be seen that when right requested number rises to certain amount (in this example, when n is more than 23 by Fig. 2A to 2C When), discrete trend is presented in the distribution of data.It is used this is because above-mentioned high value patent sample is merely possible to case demonstration , therefore sample size is less, when right requested number is larger, the case where there are sample size missings.If high value is special The data sample of profit is sufficiently large, and above-mentioned discrete phenomena will be weakened gradually, but not disappears, because of actually high value patent Also it has the right the less situation of requirement, low value patent is also had the right the more situation of requested number, and the quantity of claim is only It can reflect the probability that high value patent occurs, cannot absolutely represent the height of patent value.

In addition, can be seen that from Fig. 2A to 2C when claim number is 10, high value distribution accounting with it is common specially Profit is compared to there is the phenomenon that " diving ", according to the trend of front and back data, when claim number is 10, the accounting of high value patent It should be 2 times or so of common monopoly, and in statistical chart be only then 0.75, than claim quantity is 6,7,8,9 right Ask numerical value small, smaller than claim quantity is more than 10 high value patent accounting.Be not claim quantity be 10 The value of patent is bound to small from probability, and problem is still from patent sample.According to high value patent recited above Screening conditions, most in obtained sample is all the patent of invention authorized.Patent of invention needed before mandate by Examination as to substances, in examination as to substances, the claim of most patents all can meet with the query of auditor and carry out local modification, lead Cause claim number and quantity when initial application of the patent of ultimate authority variant (typically reducing).And it represents common special The patent in " the full library " of sharp sample includes then Chinese whole patent application and utility model patent, these patents are not pass through Cross examination as to substances.And in view of in Chinese patent expenses standard to the regulation of claim additional cost, many applicants or agency People gets used to the quantity of claim being limited within 10 or 10, therefore results in the patent without substantive examination, and right is wanted The distribution probability for asking quantity to be 10 is higher, and statistical data is made to seem abnormal.Similarly, in above-mentioned scatter plot, power Sharp requested number is respectively that abnormal phenomenon is also all more or less presented in the statistic that 20,30,40 etc. positions are 0, also and this Or there are some relationships (some people like " making up a round number " when writing claim) more.We comment in actual implementation patent value degree It, can be by ignoring these abnormal data points, to more truly restore patent parameter and patent value when valence formula Relationship.

Certainly, the function for only a simple high value patent sample pattern being leaned on to reflect is not necessarily exactly accurate, solution Method certainly has following several：First, improving the purity of the quantity and high value patent wherein included of high value patent sample； Second is that the high value patent sample pattern that structure is different, carries out back-to-back calculating, is compared respectively later, find out original respectively Cause simultaneously adjusts accordingly；Third, the high value patent model using Primary Construction carries out the patent sample of multigroup random sampling Grouping, is analyzed by the accuracy to group result, is manually adjusted to the weight of relevant parameter.In short, high value The structure of the sample pattern of patent is an extremely complex process, needs to calculate repeatedly, compares, adjusts, verifying.

In addition, it is contemplated that patent database is a dynamic change, ever-increasing data acquisition system on the whole, with the time Passage, the result and the rule of reflection of big data statistics also can constantly change, and therefore, establish the evaluation of patent value degree When model, also need that periodically model is verified and adjusted.

During the foundation of above-mentioned patent numbers and the relation map of evaluation parameter, there is no according to Chinese patent for we Parameter characteristic be especially arranged by the method for expert estimation, but this relation map really reflects Chinese patent True feature, reason are that the basis that collection of illustrative plates is established is the patent data of China and is extracted from Chinese patent data Therefore high value patent will not be influenced by other countries and regional patent feature.

We can create the relation map of United States Patent (USP) quantity and evaluation parameter, and by comparing in the same way, It was found that difference of the two country variants on patent numbers and patent value degree relationship.Illustrate below with reference to Fig. 3 A to Fig. 3 C This point.As an example, we choose, the U.S. applies so far after 2000 and disclosed whole patents are as basis sample, and Choose wherein with ETSI (European Telecommunications Standards Institute, European Telecommunication Standard Association) the relevant standard necessity patent of standard (total 18393) is used as high value patent sample, divided using the above method After analysis, following relation map is obtained：

Fig. 3 A show that the patent of different claim quantity in the United States Patent (USP) according to the embodiment of the present disclosure is special first The probability occurred in sharp data sample.Fig. 3 B show different claim quantity in the United States Patent (USP) according to the embodiment of the present disclosure The probability that occurs in the second patent data sample of patent.Fig. 3 C are shown in the United States Patent (USP) according to the embodiment of the present disclosure not With the comparison for the probability that the patent of claim quantity occurs in the first and second patent data samples.

High value patent is can be seen that from Fig. 3 C and the probabilistic relation of common monopoly claim distributed number is：

Y=0.0297x+0.5452 (1)

It can show that this parameter of claim quantity is 0.0297 to the degree of association of patent value degree from equation (1).

Comparison diagram 2 and Fig. 3 can be seen that：

(1) it compares the probability that the patent that Sino-U.S. claim is n occurs in whole samples and can be seen that China specially Benefit number is distributed in section of the claim number less than or equal to 10, and claim quantity is respectively that 1 and 2 patent numbers account for Than having respectively reached 9.6% and 9.8%.Other than claim is equal to 10, probability of occurrence it is most be claim number etc. In 3 patent, patent numbers are in the trend that is gradually reduced in section of the claim quantity equal to 3 to 9, but work as claim number When amount is more than 10, there is cliff of displacement formula bust in patent numbers, and when claim number is more than 12, accounting substantially remains in 1% or less. And United States Patent (USP), when right requested number is less than 10, patent numbers are in more slow ascendant trend, are equal to 10 in claim It is in that when right requested number is more than 20 slow downward trend is presented, right is wanted in gentle straight line substantially in 20 section The patent accounting in 20 to 30 section is asked to be all higher than 1% substantially.Illustrate that the claim quantity of U.S.'s common monopoly is most It is maintained at a larger numberical range, overall patent drafting level is higher than Chinese patent.In addition, in claim quantity When equal to 20, the distribution of U.S.'s high value patent and common monopoly, which is all presented, significantly flies up situation, this is obviously and the U.S. The rule of collecting of Patent right requirement additional cost has direct relation.

(2) distribution probability for the high value patent that Sino-U.S. claim is n is compared, it can be seen that Chinese high value is special The quantity of sharp claim is substantially in rising trend within 10, when right requested number is more than 10, although in quantity point There is obvious bust on cloth, but can slowly decline from 3% or so below, when right requested number is equal to 21, still keeps 1% or more.The high value patent in the U.S. with claim quantity increase, it is still basic to keep first rising becoming of declining afterwards Gesture, but generally, the patent accounting more than claim quantity is higher than the high value patent of China.Illustrate that the high value of China is special Although common monopoly of the profit than China on average entitlement requested number has larger offer, with U.S.'s high value patent phase Than still there is certain gap.

(3) in general, with the increase of claim quantity, the average accounting of U.S.'s high value patent is also carrying Height, but growth rate (slope) be significantly lower than Chinese patent, especially when right requested number be less than 20 when, high value patent with The quantity of common monopoly has no clear superiority in probability distribution, illustrates that the quantity of claim is in patent for Chinese patent Performance in terms of value becomes apparent, and in other words, relative to common monopoly, Chinese high value patent is quantitative in claim Difference is more notable.

In step S104, the judgement schematics of patent parameter and patent value relationship are built

Although the relevant parameter for theoretically, including in model is more, summary and description to patent value will get over Careful, evaluation result should be more accurate.It bears in mind that, evaluation index increases the complexity that also will increase model and operation Amount, and some indexs (such as patent examination duration) are sufficiently complex with the incidence relation of patent value, need to be directed to specific feelings Condition is made a concrete analysis of, and only merely mechanically applies mechanically the result of system data statistics and can generate back of the body road to the judgement of patent value instead sometimes And the effect speeded；And some to study a question have an impact but the index that is temporarily difficult to obtain comprehensively (such as partial invalidity, which examines, to be determined In be deactivated the importance of claim) can also reduce the feasibility of this research method.These patents are needed to experiment As a result it carries out selectively being given up on the basis of serious analysis.It would therefore be desirable to consider factors above and reality Collection of illustrative plates test result evaluation parameter that model is used select.

After determining the evaluation parameter selected, followed by the actual features for analyzing each parameter, the parameter and patent valence are built The judgement schematics of value relationship.Below still by taking claim number parameter as an example, illustrate that claim quantity is evaluated with patent value The building process of formula.

(1) meet the structure of the patent value judgement schematics of the direct parameter of curved line relation

Fig. 4 shows Chinese high value patent and common monopoly claim distributed number situation relationship matched curve.It is logical It crosses and depth analysis is carried out to the scatter plot of Fig. 4, following nonlinear mathematical model can be fitted：

Wherein, Vi refers to contribution of i-th of patent parameter to patent value.M is the maximum value that Vi can be got, and p is parameter Value, the value of p when K tends towards stability for Vi.

China's high value patent and common monopoly claim distributed number situation relationship matched curve shown in Fig. 4 and Speech, p can indicate that the quantity of Current patents claim, M indicate the maximum value that claim parameter influences patent value degree, I.e. when right requested number p is more than threshold k, the increase of claim quantity no longer embodies the influence of patent value degree.This reality The value for applying K in example is 50.

Compared to above-mentioned equation (1) this linear fit as a result, equation (2) can more accurately reflect the power of Chinese patent Influence situation of the sharp requested number parameter to patent value.

Certain above-mentioned fit approach can be additionally used in the other parameters such as patent citation quantity, patent families quantity, need to illustrate , different parameters shows different on the shape of the influence curve of patent value and the reason of cause this influence, needs It targetedly to be analyzed and be designed respectively, even need to be segmented in some cases to provide corresponding functional relation.

(2) it is unsatisfactory for the structure of the patent value judgement schematics of the direct parameter of curved line relation

And the not all parameter that may influence patent value degree can be transferred through the fitting structure of above-mentioned curve.For example, application Mankind's type (enterprise, universities and colleges, other mechanisms, individual), the type of patent (invention, utility model), validity are (effective, careful In, failure) or statutory status (lawsuit, license, transfer, invalid etc.) action value of patent can all be had an impact, but these are joined Number is difficult to be analyzed by the statistics to variable.For the parameter of this type, the side of big data analysis again may be by Formula counts distribution probability of the parameter in high value patent and general patent, and estimates each case to action value It influences and provides corresponding weighing factor.For example, by taking applicant's type as an example, it can be with special corresponding to each of extracting parameter value Profit, to the patent corresponding to each value carry out patent value degree evaluation (established model, which may be used herein, to be evaluated, such as The rudimentary model of the other parameters structure for meeting curved line relation according to above-mentioned evaluates, or otherwise evaluates), and base Carry out the weight of influence of the value of setup parameter to patent value degree in evaluation result, each value based on parameter is to patent value The weight of the influence of degree embodies the judgement schematics of relationship between parameter and patent value degree to obtain.

Can be that enterprise, universities and colleges, other mechanisms, individual will be special according to applicant by taking applicant's type parameter as an example Profit is divided into group, utilizes the rudimentary model built with parameters such as claim number, patent citation quantity, patent families quantity The action value of each patent in every group is obtained, and then obtains the comprehensive value degree of this group of patent, it is specific to obtain comprehensive value degree Mode can refer to the content described in step S105 below.Relevant parameter can be designed according to every group of comprehensive value degree The weighing factor of value, and the influence factor of the parameter is included in patent value degree evaluation model.

Still by taking applicant's type parameter as an example, it is assumed that set through the above way enterprise, universities and colleges, other mechanisms, These four personal applicant's types are respectively tetra- values of a, b, c, d to the influence power of patent value, then can be in patent value degree It is evaluated according to following rule when evaluation：

If applicant's type=enterprise：V1=a

If applicant's type=universities and colleges：V2=b

If applicant's type=other mechanisms：V3=c

If applicant's type=individual：V4=d

Wherein V1, V2, V3 and V4 indicate enterprise, universities and colleges, other mechanisms, these four personal applicant's types pair respectively The contribution of patent value.

(3) structure of the patent value judgement schematics of indirect parameter

In above-mentioned table 1 in addition to direct parameter, " patent agency's general patent Bid-winning rate index " etc. 14 is also listed A indirect parameter.Each indirect parameter be by carrying out COMPREHENSIVE CALCULATING acquisition to several subparameters for influencing its numerical value, In some embodiments, each indirect parameter is also to be obtained by different value assessment models.For example, " patent agency integrates This indirect parameter of patent application performance figure " is by the patent comprehensive value degree scoring of agency, proxy machine The multiple parameters such as the patent of invention quantity of structure agency and its average entitlement requested number of accounting, invention patent mandate rate, patent Overall merit is carried out to obtain.When building judgement schematics for indirect parameter, the factor for influencing patent value degree can be chosen, together When can also choose subparameter for evaluating the factor, which is evaluated according to subparameter, evaluation result is used as selected by The indirect parameter taken.Specifically, indirect parameter can be obtained by following equation：

Wherein, p_iIndicate that i-th of indirect parameter, s indicate to influence the quantity of the subparameter of the indirect parameter number, r_tRefer to wherein The contribution that t-th of subparameter generates the indirect parameter numerical value, w_tIndicate the weight of t-th of subparameter.

Relationship between indirect parameter and patent value degree can be expressed as：

V_i=f (p_i) (4)

Wherein V_iIndicate contribution of i-th of indirect parameter to patent value degree, f () representative function relationship.

It should be pointed out that further including that " the preliminary value of single patent is commented in the indirect parameter listed in this chapter subordinate lists The preparation method of valence index ", the parameter is as follows：1, the preliminary assessment of patent value is built using other indexs in addition to this index Model, patented preliminary Indexes of Value Assessment；2, be directed to the preliminary Indexes of Value Assessment to part direct parameter or indirectly Parameter is iterated, optimizes, the Optimized model of patented action value evaluation；3, to obtain patent using the Optimized model final Patent value degree.

In step S105, the overall merit formula of patent value degree is established.

After the judgement schematics of above-mentioned direct parameter and indirect parameter are established, it can set about building the comprehensive of patent value degree Close judgement schematics.

The existing parameter system for evaluating patent value degree is roughly divided into a meta structure, linear structure and tower structure Etc. several structure types.Wherein a meta structure is evaluated using single parameter, it is clear that is not suitable for carrying out the action value of patent Evaluation.Linear structure uses multiple parameters, but these parameters are coordination between each other.Judge due to influencing patent value degree Factor it is more, system can be made to become excessively complicated using linear structure, it is difficult to hold the relationship between each parameter, and can The case where capable of causing to take a part for the whole, the reliability of analysis result is made to reduce.According to the parameter for evaluating patent value degree Feature can select tower evaluation structure, i.e., all evaluation parameters is different according to the angle for embodying patent value, will join Number is divided into many levels.For example, parameter can be divided into multiple groups, one evaluative dimension of every group of correspondence.In the present embodiment may be used Parameter is divided into group according to three dimensions such as technology (technical advance), law (technical stability), market (protection domain) Three groups, wherein technical stability mainly considers whether patent passes through substantive examination and obtain mandate or remain effective, if warp It the movable inspection of the market operations such as crosses action for infringement of a patent, pledge, and whether having the patent families etc. abroad authorized；Technology is first It is mainly in view of patent and its patent families into property and is cited situation, research staff's quantity, the patent of patent input in the whole world Technical field that technology is related to (IPC classify quantity), patent transfer perhaps can situation etc.；Protection domain Primary Reference patent Claim number, the remaining protection period, patent distribution country quantity etc..Since the parameter of selection has hierarchy, Every weight is controllable in each level, and can be expanded with increasing for parameter and accordingly, thus more flexible, office It is sex-limited also smaller.

The comprehensive value degree of patent is obtained by the sum of comprehensive value of above three dimension：

Wherein, when i value ranges respectively in [1, n1], [n1+1, n2], [n2+1, n] when, V_iPresentation technology (technology respectively It is advanced), the influence of law (technical stability), the patent parameter under market (protection domain) three dimensions to patent value degree Power is contributed.Can by the method that extreme value limits the Numerical Control of each dimension it is desirable that in the range of.

In addition, can also be established for a patent set (such as on the basis of above-mentioned Chinese patent value degree evaluation model All Chinese patents of a certain applicant application) action value evaluation model, thereby come special to the synthesis in multiple and different set Profit value is evaluated, is compared.

In addition, other than parameter listed by table 1, User Defined parameter can also be set, in this case, user can be with Desired other parameters type is selected as needed, can also provide corresponding data list if necessary as patent data sample This part.When building patent value degree evaluation model, selected part or whole parameters can be given tacit consent to from table 1 to build Preliminary patent action value model, and other patent value degree evaluation model is built based on user-defined parameter, by it It is combined with preliminary patent action value model, the patent value degree evaluation model to finally be built.

In addition, the action value of patent is presented as strategic value, technological value, legal value, market value and the warp of patent Many levels, the patent value degree evaluation models of above-described embodiment such as Ji value can provide action value at least partly level and comment Valence as a result, for the not covered level of above-described embodiment, can provide by other means patent value degree evaluation model or Person even directly obtains evaluation result, is combined with the patent value degree evaluation model of above-described embodiment and obtains final patent valence Value degree is evaluated.For example, can be in evaluation of the evaluation model to technological value, legal value and market value of above-mentioned patent value As a result on the basis of, increase the evaluation of the relevant evaluation dimension of strategic value or economic value, to by application range from macroscopic view Analysis is extended in the value analysis for case.The strategic value and economic value of patent are largely dependent upon patent sheet Some specific properties of body, for example, patent the relevant technologies or product practice situation, the evading property of the technology of patent, can The complexity that confidentiality or substitutability, patent infringement evidence obtain, and to applicant's and patent assets resetting The excess earnings etc. that cost, Recent Activity price, the patent assets of same in the market or similar assets are brought, and these belong to Property is much personalized, it tends to be difficult to by being obtained in traditional patent information or legal information, be needed in light of the circumstances Independent assignment.

It should be noted that above example is merely to illustrate that the principle of patent value degree evaluation model structure, it is real The influence factor for the more complexity being contemplated that in the building process of border.For example, there may be stronger correlations between partial parameters Property (such as authorization conditions etc. of the authorization conditions of patent and patent families), may cause to repeat the case where weighting.In addition, this reality It is the comprehensive evaluation model for invention and utility model patent to apply the model in example, wherein patent application without substantive examination and Parameter involved by utility model and patent of invention after substantive examination is different, the patent value feelings of same parameters reflection Condition is also not quite similar, and for these intricate situations, multi-angle judgement can be carried out in logical relation, and according to difference The case where carry out selective merging or shared weight suitably adjusted.It therefore, in actual operation, can be in conjunction with association Degree test as a result, being accepted or rejected, being integrated to index of correlation.

In step S201, patent value degree evaluation model is built.It in this step, can be by above with reference to Fig. 1 to Fig. 3 The method of description builds patent value degree evaluation model.The quantity of constructed model can be one or more.For example, can To build the first model based on one group of patent data sample, and the second model is built based on another group of patent data sample, So that subsequent step uses.

In step S202, patent value degree evaluation model is optimized.The mode of optimization include but not limited to iteration and Compare.

For example, it is assumed that constructing a model, i.e. the first model in step S201, can will be by the first model evaluation then Action value is used as in the target patent of expected range and refers to patent, is used to build the first model to update using the referenced patent Patent data sample.Here so-called expected range can select as needed, for example, higher than preset value, be less than preset value And/or in a default range.Patent for that for screening the model of high value patent, can will be above preset value is made Patent data sample is updated for referenced patent；For for screening the model of low value patent, preset value can be will be less than Patent is used as updates patent data sample with reference to patent.In some embodiments, equally the case where constructing a model Under, the model can also be optimized in the following manner：Previously selected verify data is evaluated to obtain using the model The first evaluation result is obtained, evaluation builds patent data sample used in the model and commented first to obtain the second evaluation result Valence result adjusts the model compared with the second evaluation result, and based on comparative result.Here, evaluation patent data sample with Obtain the second evaluation result operation can rule of thumb either preset standard come execute or by other means or profit It is executed with other tools.According to the deviation between two evaluation results after comparing, to optimize and revise the model, to improve Its accuracy.

It, then can will be by likewise, it is assumed that construct two models, i.e. the first model and the second model in step S201 One model and/or the second model evaluation are that action value is used as in the target patent of expected range with reference to patent, utilize the reference Patent updates the patent data sample for building the first model and/or the second model.Similarly, so-called expected range is same It can select as needed, details are not described herein.As another example, one group of patent for verifying model can also be prepared Data evaluate the patent data using the first model and the second model respectively and obtain evaluation result, by commenting for the first model The reason of valence result compares with the evaluation result of the second model, finds out difference adjusts the first model and/or the second mould accordingly Type.

It in some embodiments, can also be when using model come evaluation goal patent, if the target patent meets Preset condition, the then weight for evaluating corresponding sub- judgement schematics in model used in the target patent increase or reduce.Example Such as high value patent, it is clear that the remaining term of validity is longer, and patent value is higher.But for low value patent, The value of its patent and the length of the remaining term of validity have no apparent incidence relation, if according to similarly square with high value patent Method is weighted, then will certainly influence the objectivity of evaluation result.It therefore, can be by the method for above-mentioned iteration, in target patent After being be evaluated as high value patent, it is believed that it is necessary to reinforce the contribution of remaining life this parameter to patent value degree, its is right The weight for the subformula answered increases, otherwise reduces, and then reuses the model after optimization to reappraise target patent, thus To more accurate action value evaluation result.

In step S203, using constructed patent value degree evaluation model come the action value of evaluation goal patent.For example, If in the patent value degree evaluation model that step S201 is obtained being built based on three groups of parameters, this three groups of parameters correspond to respectively In technical stability, technical advance, the evaluation criterion of protection domain these three levels.Can so utilize in this step should Model obtains target patent respectively in the evaluation result of these three levels, by these evaluation result weighted sums, you can obtains comprehensive Evaluation result is closed, can be sorted accordingly according to patent value degree to retrieval result.

It, can be as needed using suitable model come the action value of evaluation goal patent in this step.Here target is special Profit includes but not limited to single-piece patent or patent group.

For example, if target patent is common single-piece patent, it can be using the data sample with single-piece patent come component Patent value degree evaluation model evaluate its action value.If target patent is the single piece being processed into patent families Compound patent then can evaluate its valence using the patent value degree evaluation model with corresponding compound patent data sample structure Value degree.If target patent is patent group (such as using applicant, IPC code, applicant region, patent families as condition The patent group being divided into), then every patent in group can individually be evaluated, and evaluation result is performed mathematical calculations to be directed to The evaluation result of patent group, mathematical operation described here include but not limited to averaging, geometric average, square mean (root mean square Average value, rms), harmonic average, weighted average etc..It can also be located using statistical tools such as box figure statistical tools The evaluation result of each patent in reason group, to obtain the more accurate rational evaluation result for patent group.

Patent classification can be carried out before or after the mathematical operation (that is, as described above will be at result of calculation normalization Manage into the score value in preset range), if it is the former, then the action value of every patent is classified first, then to each patent Action value score value perform mathematical calculations to obtain the evaluation result for patent group, then will be for the evaluation result of patent group Rounding；If it is the latter, then the evaluation result of every patent is performed mathematical calculations first to obtain the evaluation knot for patent group Fruit, then by the evaluation result normalized of patent group at the score value in preset range, that is, by the action value of patent group point Grade.

Fig. 6 shows the schematic diagram of the comprehensive evaluation result of the patent value degree according to the embodiment of the present disclosure.Such as Fig. 6 institutes Show, user is more intuitive, evaluation result of quickly value -capture degree in order to make, and realizes the screening to patent value and system Meter, normalized has been carried out to above-mentioned overall merit score value.For example, can be according to overall merit score value of the patent in full library Distribution situation, segment processing is carried out according to the marking mechanism of ten point system to the overall merit score value of patent, ensures each score value Patent data ratio shared in full library according to certain regular distribution, the probability distribution of patent value is from 10 points to 1 point Successively decrease.Equally, ten point system processing can also be carried out after the same method for the evaluation score value of each classification.

In addition, the model in the present embodiment is the evaluation for the design of the comprehensive value of Chinese invention and utility model patent Model, although the patent value degree of acquisition reflects the probability distribution of patent value on the whole, if quickly to judge patent valence Overall merit score value of the similar above-mentioned mode to three evaluative dimensions of patent value may be used in the core influence factor of value Also the grading design of ten point system is carried out.User is on the basis of obtaining patent comprehensive value degree, in conjunction with the row of relevant evaluation foundation The comprehensive scores of table displaying and each dimension value, you can fully understand the value concrete condition of patent.

Chinese design patent is compared with inventing with utility model, and the object of protection is different, the selection in parameter and valence It is distinct in terms of the judgement of value, it is contemplated that these differences in model construction and evaluation, but its basic mentality of designing and Method is identical, and details are not described herein.

Although giving some sample implementations optimized to patent value degree evaluation model, this public affairs above The embodiment opened is without being limited thereto, and the optimization of model will be further described below.Above example is directed to Chinese high value Patent value degree evaluation model is described for the realistic model of the evaluation of comprehensive value of patent establishes process.In above-mentioned implementation In example, the demand of high patent value is quickly identified in retrieving for user, makes full use of China to implement from 1985 special Disclosed patent document information and relevant patent law and operation information since sharp system, in the base of many experiments and analysis On plinth, based on the patent of invention that China authorizes, the basic model that China authorizes patent of invention is established, and basic herein On, design adjustment has been carried out respectively for Chinese patent application and Chinese utility model, Chinese design patent, has been realized pair The objective evaluation of the comprehensive value of Chinese patent and classification.From the point of view of the actual verification effect of the model, this model is to China High value patent can reach desired identification and resolving effect.

In addition, user in retrieving, except the high value patent for wanting to quickly identify China, also often is intended to energy The value of the patent of other countries and area is classified simultaneously, and is shown simultaneously in retrieval result.For this purpose, can be with needle To the patent feature of country variant and area, the setting of parameter and its weight in model is adjusted, keeps it special with China Sharp action value evaluation model combines together, forms an integral framework of patent comprehensive value assessment.

On this basis, finer personalized model design design can be also carried out, such as China Chemical Industry material The thematic data base in field can choose high value patent sample therein after basic database is built up, and pass through analysis, shape At the patent value degree assessment models of professional domain.

Specifically, patent information can be applied to technical information analysis, technology or novelty of patent analysis, freely implement (FTO) several scenes such as analysis, Competitive Intelligence Analysis, patent operation analysis, anti-infringement analysis, technological cooperation, talent introduction.No The purpose of same application scenarios retrieval is different, is also different to the understanding or demand of " high value patent ".Therefore, we also need Will on the basis of above-mentioned comprehensive evaluation model, can be directed to different application scenarios in model relevant parameter and weight into The corresponding adjustment of row, forms the action value evaluation model of fining.

When stressing contribution of some parameter to patent value degree, the weight of the parameter can be increased in judgement schematics Greatly, otherwise reduce.It is briefly described below for the emphasis of different application scenarios and its patent value degree evaluation model.

(1) patented technology

Patented technology value analysis is chiefly used in using acquiring technology solution and its Technology Evolution process, development trend as mesh Research activities, such as specific technique basic research and analysis, approve and initiate a project before technical investigation, process of approving and initiate a project In technical solution choose and Project-developing in the relevant technologies solution inquiry etc. scenes.In such application scenarios, use The emphasis of family concern is the novelty and its origin, evolution, applicable cases of technical solution itself, when especially those technologies originate from Between not long, novelty degree is higher and development compared with rapidly, the larger patent of influence power.For this purpose, for the purpose of designing technique intelligence use Patent value degree evaluation model when, direct parameter should pay close attention to date of application (Patent Age) of patent, novelty degree, specially The type and authorization conditions of profit, cited literature 2 average age, IPC classification quantity, patent or patent family are cited quantity (especially It is that patent numbers are drawn by him) and its IPC classification quantity, the Patent Citation that are related to the IPC classification quantity, the independent claims that are related to Corresponding dependent claims quantity, specification quantity and inventor's quantity etc.；Indirect parameter should pay close attention to applicant's Complex art advanced index, the technology correlation degree of affiliated industry, the emerging degree of technology, technology temperature and technology life cycle etc..

(2) risk of infringement is analyzed

Risk of infringement analysis includes for the patent infringement retrieval analysis of specific technical solution and for particular technology area Patent infringement Risk-warning analysis.

Patent infringement retrieval is also referred to as patent infringement investigation, that is, refers to for specific technical solution, and possibility is found out by retrieval There is the patent of risk of infringement, and analyzes and determines whether existing technical solution falls into the process of its protection domain.Patent infringement is examined Rope also belongs to the microcosmic retrieval for specific technology contents, and of interest to the effect that whether there is and analyzed skill Patent similar in art solution technique content, whether the current statutory status of the patent is effective, if effectively, relevant technical solution Whether its protection domain is fallen into.

Although index infringement and the microcosmic retrieval of non-directional carried out for certain aspects, the content for analysis of encroaching right It is related to the protection of patent right, patent writes quality, the protection domain of right, the property of subject of right and its to patent Protection, right-safeguarding strategy etc., influence whether the probability that patent infringement risk may occur, and therefore, when carrying out index infringement, have Necessity distinguishes the risk probability that patent infringement risk may occur, that is to say, that, it is necessary to it is directed to the mesh of index infringement The a set of patent venture worth assessment models of design.The model is more concerned with related to Legal factors in building of digitized in the design of direct parameter Current patents validity, patent type, rights to independence requested number, dependent claims quantity, independent claims number of words Or technical characteristic number, specification number of pages, patent action and patent families lawsuit situation, the license of patent and transfer situation, patent And patent families are deactivated situation, authorization conditions of patent families etc., indirect parameter should more pay close attention to applicant Technology Correlation, General patent Bid-winning rate index, applicant's integrated computer sex index etc..

Patent Risk-warning refer to user in order to find or evade in advance may generation risk of infringement, for the skill of concern Art field carries out patent retrieval analysis, finds out the patent of possible occurrence risk and is classified, monitored and taken counter-measure Process.Unlike patent infringement retrieval, the not specific technical solution that patent Risk-warning is directed to, with concern Technology, product or the relevant all technologies of business.But it as index infringement, the emphasis of concern is all the right of related patents The patent value degree analysis model of the probability for protecting situation and risk to occur, patent Risk-warning can refer to patent infringement retrieval Model.

(3) patent operation is analyzed

The purpose of patent operation analysis is to be quickly found out the high value patent of operation value, and therefore, we first should The traffic-operating period of the patent of concern the activities such as permits, transfers the possession of, pledging, wherein patent grant and transfer are divided into as organization internal Transfer and license and tissue outside license and transfer, the former more lays particular emphasis on the tactical management of organization internal, and the latter is then The market operation value that more objective can embody patent.Since the activity such as the lawsuit of patent, invalid can also reflect specially from another side Influence of the profit to market, and then reflect its operation value, it should also the emphasis paid close attention to as us.In addition, patent is cited Situation reflects the technological influence of patent, and it is also often that operation value is bigger technically to have the patent of significant impact Patent；In addition the validity of patent, protection domain, the operation of patent families and lawsuit invalid situation and the synthesis of applicant Operation ability index etc. also has important reference role to the operation of patent value, can be used as patent operation value analysis mould The important index of type.

(4) patent application quality control

Patent value is that the comprehensive of technological value, legal value and market value embodies.Wherein technological value is high value The basis that patent is formed, and patent application is the key that then to convert technological value to legal value, technological value and law valence The combination of value becomes the support of market value.Therefore, the weight of the management to patent application quality and monitoring and public attention One of point.Patent application Environmental Evaluation Model can be used for inspecting applicant itself or rival, affiliate's patent Shen Quality please, especially can be investigating the service quality of patent agency.

It, can emphasis consideration independent claims quantity, rights to independence when building patent application quality value degree evaluation model Profit requires corresponding dependent claims quantity, independent claims number of words (or technical characteristic number), specification number of pages, specification Attached drawing quantity etc. and the relevant direct parameter of patent drafting also can refer to above each index in relevant art field if possible Average index.Applicant's general patent Bid-winning rate index, patent agency's general patent Bid-winning rate index, indirectly Although parameter is all related to the Bid-winning rate of patent, belong to the comprehensive evaluation index of patent application quality, these indexs into Row patent comprehensive value is analyzed or when competitor analysis, OA operation analysis, can be used as the reference of related side's patent quality ability according to According to, but when the Bid-winning rate to specific patent is evaluated, it is not recommended that assign excessively high weight.

(5) invalid patent

Patent is the legal exclusive authority that law assigns patentee, but this right is had ready conditions limitation.First, Patent protection is that have regional, and a patent is after foreign countries file an application, if not in China within the time of legal provisions It files an application to protect, will lose one's prior rights in China, if this application has been disclosed before this, then the corresponding technical solution of the patent will Lose the chance authorized in China；Secondly, it after patent application, needs to authorize by examining, if patent quilt after disclosure It rejects or recalls, then applicant can also lose the right of monopolization；Again, after license, applicant has to periodic payment year Normal right could be maintained by taking, and the protection period of patent is limited, once expiration of the term of protection, applicant will lose Respective right；In addition, the patent authorized may be by other people invalid challenge, if invalid by all or part, patent right People can also lose all or part of right.These patents for losing right then can home enter publicly-owned technical field, for the public Freely use.Above-mentioned several situations are referred to as invalid patent (note by us：It should be noted that a certain applicant or obligee The forfeiture of right is not meant to that the patent centainly becomes publicly-owned technology, and the public whether there is quilt using preceding it should also be noted that investigating Applicant or other people legal respective rights effectively possessed；It is to have regional in view of patent protection, a patent is in a state The failure in family or area, not equivalent to its protection in other countries or area also lose effect).

Under certain application scenarios (such as patent infringement risk analysis and patent operation analysis), invalid patent almost without Utility value.On the other hand, the purpose that patent system is set up is the achievement for allowing the public to obtain technological innovation, and by applying and changing Into promotion social progress.A large amount of invalid patent, the exactly public can convenient acquisition free repository.Invalid patent exists It determines that in the case of other lawful rights, directly utilization, does not greatly save research and development time and Innovation Input with can having no risk, Its value dimension is also unblamable.Therefore, being specially classified, being evaluated to the utility value of invalid patent is also extremely intentionally Justice.

The value assessment of invalid patent is analyzed with patent operation larger difference in terms of application purpose and scene, but in addition to Currently there are other originals whether outside the Rule of judgment that target country fails is just the opposite, judged for high value patent to patent Then substantially coincident, action value evaluation model can afterwards obtain with minor modifications on the basis of patent operation action value evaluation model It arrives.

It needs to be emphasized that practical application scene of the optimization of above-mentioned action value evaluation model only for user Increase the weight of some or certain parameters in the contribution of patent value degree, the priority level of patent be respectively set, It is retrieved and is browsed with fast and easy, should not be construed as a kind of patent screening or the pattern of exclusion.In view of patent document information with And the complexity and uncertainty of patent value itself, the patent that any mathematical model established using patent inherent parameters is formed Value appraisal system is all the evaluation and test to dependent event probability of happening, rather than to absolute judgement conclusion that patent itself is worth.This Outside, above-mentioned model can be used alone, and can also be combined with each other comprehensive use.For example, can first be carried out to retrieval result Patented technology value is analyzed, then operation value analysis is carried out to it, it is both comprehensive as a result, can rapidly find out operation value and The relatively high patent of patented technology value.

In addition, the adaptability of above-mentioned patent value degree evaluation model is total and minute design is directed to different objects respectively Patent type, application region, application scenarios etc. have carried out respective design so that and evaluation result is more targeted, thus also more Science, rationally.

In above-mentioned patent value degree evaluation model, the selection of each parameter and its setting of weight are to sample On the basis of notebook data is analyzed, obtained using corresponding analysis assignment method.For on the whole, evaluates and be classified As a result meet the rule of the probability distribution of high value patent, it is thus possible to meet most common retrievals, analysis scene.Usual feelings Under condition, an action value evaluation model once it is determined that, except occasional carry out some part optimize and revise in addition to, these parameters and Its weight is usually fixed, and user only has the right used, without the right of unrestricted choice adjustment.Such action value is commented Although valence model disclosure satisfy that general retrieval analysis demand, but for more high-end user, it is also necessary to be directed to no Application scenarios and different demands carry out personalized adjustment, the flexibility design of implementation model to model.

(1) it is directed to the flexible design of the patent classification of different characteristics

The development of technology is unbalanced, for example, the different times of same technical field, technology temperature, technology Life Cycle Phase, market competition situation etc. might have prodigious difference, if analyzing the value of patent with a same model, rather can Occur biased.At this point it is possible to which user is allowed voluntarily to input search condition, basic data and high value patent sample therein are specified This, counts corresponding patent parameter respectively, after these parameters and original model are compared, assigns new power again Weight, to form more science, accurate action value evaluation model for the technical field.

Equally, also can to different applicant's types (such as enterprise, universities and colleges, individual), different times application it is special The characteristics of profit or even the patent of different applicant's applications, is analyzed, and mould is evaluated to form corresponding personalized action value Type.

(2) it is directed to the flexible design of the individual demand of different user

The demand of user is diversified, and perfect curing model can not all adapt to the different individual character of user in any case Change demand.For example, user carry out patented technology value analysis when, if more concerned with emerging patented technology, may Wish the weight up-regulation shared by the date of application of patent and novelty degree；If more concerned with the influence power of technology, it would be desirable that increase The weight for being cited situation of patent or patent family.Therefore, on the basis of existing action value evaluation model, allow user Corresponding index is adjusted, realizes the flexibility design of patent value degree evaluation model, it will make model more rich in human nature Change, to be flexibly adapted to a variety of different application scenarios.

The characteristics of patent value degree evaluation model, limitation and applicable application scenarios

(1) objectivity

The all objective parameters extracted from patent information of index used in evaluation model, each parameter is to value The weighing factor of degree obtains according to the result operation of the big data analysis of China patent database, the reflection of patent value degree Patent value is the real embodiment of each objective parameter of patent, removes and carries out subtle adjustment according to actual conditions in later stage iteration phase Outside, it is not affected by human factors substantially.

(2) comprehensive

In the building process of evaluation model, comprehensive survey has been carried out for the parameters that may influence patent value degree Examination and verification, have carried out comprehensive analysis for each dimension of patent value, patent are fully taken into account in the building process of model The various influence factors of parameter, the evaluation result finally obtained is the integrated application of multiple parameters, with obvious effects to be better than single ginseng Several evaluations.

(3) real-time

Evaluation index in evaluation model is the dynamic indicator acquired in real time in database, with the bulletin of patent, law shape Events real-time update, the patent value models itself such as state variation, the appearance of new citation information also can be periodically according to the changes of database Change situation to be adaptively adjusted, ensures the value situation that truly reflection patent is current.

(4) accuracy

Evaluation model is on the basis of the advanced credit card action value evaluation model in the world, according to the spy of Chinese patent information Point and rule are well-designed, can really reflect that the value distribution situation of Chinese patent, accuracy are answered in long-term practice It is verified in.

(5) convenience

The weight of parameters is calculated automatically by computer in evaluation model, is chosen in high value patent sample suitable In the case of, the participation for not needing expert substantially (does not need expert especially to be commented respectively for the index of each parameter Valence) structure of model can be completed, it is convenient and efficient.

(6) flexibility

Evaluation model is compared according to the high value patent sample parameter of selection and the mean parameter of database patent It obtains, therefore, user can choose different analysis targets and high value patent sample, construct rapidly according to actual needs Corresponding model.For example, user can by after 2000, applicant's type be colleges and universities Chinese invention and utility model patent Shen Basic data please be used as, and therefrom select part high value patent sample, after constructing Chinese Universities ' patent 2000 The action value evaluation model of patent, and the value of these patents is objectively compared, is analyzed.

Each parameter depends on the weighing factor of action value the distribution feelings of the relevant parameter of high value patent sample in model The high value patent sample of condition, selection is different, and the weight of the parameter in a model also can be different therewith, to lead to patent valence The evaluation result of value degree generates difference.Therefore, the strategy that high value degree patent sample is chosen is in action value model construction process Key factor.The weight adjustment mentioned in above example, can select suitable realization method as needed, such as can be with It manually performs, can also execute automatically.

The patent value degree evaluation model of the embodiment of the present disclosure is the recognised standard evaluated patent value degree according to industry, Using the big data statistic analysis result of Chinese patent database, the action value of amount body design the characteristics of for Chinese patent information Evaluation model.Can truly patent value degree be screened and be sorted according to the distribution probability of high value patent, can be made User quickly focuses high value patent in the patent of magnanimity, significantly promotes efficiency and effect that patent information utilizes.

Embodiment of the disclosure with by parse patent search result in high value patent quantity and the regularity of distribution, quickly Understanding lies in some of which important information, is into one to more accurately and comprehensively hold the overall condition of analysis object Step formulates relevant Decision or counter-measure provides strategy reference.For example, can be screened, be divided to company's high value patent Analysis obtains it and applies for quantity and trend, technical field distribution situation, technique influence situation, application country origin, inventor's input, tells Dispute, in vain, license and transfer the possession of feature, by analyzing the competitive strength, development strategy, core that understand it in technology and market etc. The information of the heart talent and attacking and defending feature etc., and then objective evaluation its competition treat to itself market, or hair in time Moment of cooperation between now, and formulate corresponding counter-measure or take corresponding action.

Embodiment of the disclosure allows user substantially understanding its synthesis in terms of the value analysis for specific patent On the basis of action value, by checking the detail information of patent value degree, the patent of each relevant dimension of the patent is quickly understood It is worth rank score value, and the details of each important parameter being had an impact to it, to rapidly to the reality of the patent Value dimension is tentatively judged.

The verification result of the embodiment of the present disclosure is introduced below with reference to Fig. 7 A to 7C.

In structure patent value degree evaluation model, its evaluation effect can be verified by practical application scene.This High value patent is selected to verify patent value degree evaluation model in embodiment.In order to ensure the objectivity of verification result, selecting When selecting the data sample of compliance test result, the foundation rule with model is avoided to be coupled, i.e., not to establish model when the high price chosen It is worth patent sample as verification sample, also retrieval is not carried out with the relevant parameter in model and obtain verification sample.It is additionally ensured that choosing The high patent value taken should have certain public trust, such as patent value has obtained market validation or recognized extensively by industry It can.In the present embodiment, sample selection standard when considering from different from model foundation chooses verification sample.

Fig. 7 A show first to the 18th China of patent value degree evaluation method pair using the embodiment of the present disclosure The action value analysis table of the prize-winning patent of patent prize.China's Patent Award is China State Intellectual Property Office (SIPO) and the United Nations's generation The selecting activity and China that boundary's Intellectual Property Organization (WIPO) is carried out jointly uniquely specially create patentable invention The government department's prize awarded is made, has successfully held 18 so far.The patented technology of Criteria not only noted item Horizontal and innovation height also focuses on its application in the conversion process of market, while also to its conservation status and management feelings Condition claims.Reward by various regions Department of Intellectual Property, State Council's unit, national employer's organization or academician of the Chinese Academy of Sciences/ Chinese engineering college academician recommends, and is selected after carrying out multistage evaluation via the Senior Expert of Department of Intellectual Property and every profession and trade respectively, It, can be as the verification sample of action value model with higher public credibility.

Obtain first to the 18th China's Patent Award (including China's Patent Award and Chinese patent gold medal) invention and Utility model patent is 3175 total.After carrying out patent value degree classification to it, export analysis table is as shown in Figure 6A.From Fig. 6 A As can be seen that it is 871 that wherein action value, which is the patent of full marks, account for total sample number 27.4% (score value is in common monopoly 4.5%) 10 patent only accounts for, patent of the action value more than or equal to 8 is 2664 total, accounts for the 83.9% of total sample number, and special Sharp action value is less than or equal to total only 127 of 5, only account for total sample number 4% (common monopoly Patent action value score value is 1 ~5 patent accounting be 55.5%), and by check the current statutory status of patent it is found that this 127 patents currently entirely Portion's failure (according to the competition condition of China's Patent Award, these patents should be entirely to authorize effectively when selected).The above case Example absolutely proves that the selecting result of patent value degree model and China's Patent Award that we currently build has extraordinary matching Degree can be used for screening high value patent.

Certainly, wherein there are also patent value degree be 6,7,8 so-called middle value patent, among these the reason of a part It is since the failure of which part patent leads to the reduction of patent value degree, simultaneously, it should also consider the competition item of China's Patent Award Include that economic benefit, market conversion etc. rely solely on the unpredictable factor of patent value degree model in part.On the other hand, specially Sharp action value evaluation model is not one and differentiates whether specific patent is absolutely " emperorship " of high value patent, although it cannot Guarantee is shot with unfailing accuracy, but its probability judged high value patent is still quite reliable, can work completely as us The competent assistant of middle identification high value patent.

About standard necessity patent (Standards-Essential Patents, SEP), International Telecommunication Union (ITU) will It is defined as " any possible patent of coverage criteria draft or patent application completely or partially ".Standard necessity patent is due to possible All or part of coverage criteria draft, thus its application value is very notable.We authorize invention and utility model in China It is chosen and ETSI (European Telecommunications Standards Institute, European Telecommunication Standard in patent Change association) the relevant standard necessity patent of standard (6851 total), action value the distribution situation analysis such as Fig. 7 B current to its It is shown.As can be seen from Figure 7B, patent value degree is more than or equal to 9 patent totally 6129, accounts for the 89.5% of total sample number, patent Only 151 of action value less than or equal to 7, account for the 2.2% of total sample number, and it is currently to have failed to have 64.9% in these patents Patent.

Fig. 7 C are shown with the comprehensive value according to the patent value degree evaluation method of the embodiment of the present disclosure to single-piece patent The displaying of degree evaluation and the evaluation of each dimension patent valence degree.In the example in figure 7, it is extracted in the patent from patent value degree equal to 10 Patent publication No. is CN102868498B, patent name be " method for generating codebooks, data transmission method and device " mandate hair Bright patent, technical stability, technical advance, the related keyword parameter of protection domain and scoring situation are as seen in figure 7 c.From It is only that the common parameter in the part that is used in patent value degree evaluation model refers to that Fig. 7 C, which can be seen that the index shown in list, Mark, but under the premise of " standard necessity patent " factor is not accounted for patent value degree evaluation model affecting parameters, can be by It is automatically recognized as high value patent, illustrates this model in building process, can reasonably use the big number of each patent parameter Information according to statistics, and carried out it is scientific and reasonable assemble, to form believable patent value degree appraisement system.

From the grading effect confirmatory experiment of the high value patent of front, it can be seen that the patent value of the embodiment of the present disclosure Model is spent objectively for quickly identifying that high value patent has the effect of highly significant.We recycle this model below The data such as operation, lawsuit to Chinese patent count, and therefrom understand some of which regularity of distribution.

To the Patent Application Publication in Chinese patent database, (including patent application, utility model and appearance design are not wrapped Including and authorize invention text) quantity counted, the quantity accounting that each legal matter occurs is obtained, as shown in the following Table 2.

Table 2

From table 2 it can be seen that total accounting of the high value patent in all patents that patent value degree is 9 or 10 is insufficient 10%, but contain generation more than half the generation of lawsuit, the patent and 40% permitting or pledge or more review Or customs files patent, 30% or more occurred transfer the possession of or invalid patent.Further to the relatively low (patent of patent value degree 4) above-mentioned patent that action value is less than or equal to is analyzed, it is found that wherein majority has rejected or void in whole, patent stability Difference, the problems such as protection domain is narrow.Therefore, user's Effective selection high value patent can be helped using patent value degree model.

The probability of various legal matters occurs for the patent that each action value is further analyzed below with reference to table 3：

Table 3

By table 3 it can be clearly seen that the high patent of patent value degree occur lawsuit, pledge, transfer the possession of, permitting, customs it is standby The probability of the legal matters such as case, review, invalid is far longer than the relatively low patent of action value.

Combine specific example briefly to patent value degree in patent retrieval and analysis practice below with reference to Fig. 8 A and 8B In application schematically illustrated.

Fig. 8 A show that carrying out patent value degree using patent value degree evaluation method according to an embodiment of the present disclosure comments The schematic diagram of the user interface of valence.Assuming that important special in terms of soy bean milk making machine in enterprise's plan understanding industry of production soy bean milk making machine Profit can search for the patent for including keyword " soy bean milk making machine " in title or abstract, and be selected according to patent in retrieval result Action value sorts, as shown in Figure 8 A.What is ranked the first in Fig. 8 A is a utility model patent " a kind of double-layer lower cover soy milk grinder " (application number：CN201414387Y).In the label beside patent it can be seen that, this patent once lived through lawsuit, transfer, A variety of legal matters such as license, customs files and invalid examination.Further investigation is carried out with regard to this patent to find, which is nine sun The patent of the Bubukao of limited liability company's purchase, in action for infringement of a patent thereafter, nine positive companies, which obtain, to be won a lawsuit and or pays for Repay 5,400,000 yuan.Slight utility model does not allow invention, and big effect has been played in patent right-safeguarding.

Next we analyze the action value distribution situation of applicant's patent.

Fig. 8 B show that the high value filtered out using patent value degree evaluation method according to an embodiment of the present disclosure is special The schematic diagram of the patent distribution of the applicant of profit.As can be seen from Figure 8B, the Chinese patent of JiuYang Co., Ltd's application is worked as In, lawsuit occurred, transfer, license, customs files, review, invalid case there are 4,19,21,6,6 and 8 respectively；Screening nine The Chinese patent Patent action value of positive company's application is equal to 9 or 10 patent, it is found that the case of above-mentioned legal matter occurred in it Part has 4,8,20,3,4 and 8 respectively, includes the important case that legal matter occurred for above-mentioned major part.It can be seen that the applicant Possess more high value patent.

The simple signal that the above is only the application of patent value degree model, in practice, should as needed carry out patent After careful retrieval, retrieval result is screened using patent value degree model, and combine patent actual content and technology, The applicable cases in market deeply, specifically analyze.

Embodiment of the disclosure additionally provides a kind of computer readable storage medium, and for storing instruction, described instruction exists Processor is set to execute the above method when being executed by processor.

As an example, embodiment of the disclosure can also be described in the context of machine-executable instruction, machine can Execute instruction such as includes in the program module that is executed in device on the true or virtual processor of target.It is general and Speech, program module include routine, program, library, object, class, component, data structure etc., execute specific task or realization Specific abstract data structure.In embodiments, the function of program module can merge between described program module Or segmentation.Machine-executable instruction for program module can execute in local or distributed apparatus.In distribution In equipment, program module can be located locally in remote medium storage the two.

The computer program code of method used to implement the present disclosure can be write with one or more programming languages.These Computer program code can be supplied to the processing of all-purpose computer, special purpose computer or other programmable data processing units Device so that program code by computer or other programmable data processing units execution when, cause flow chart and/ Or function/operation specified in block diagram is carried out.Program code can completely on computers, part on computers, conduct Independent software package, part are on computers and part is held on a remote computer or server on the remote computer or completely Row.

In the context of the disclosure, machine readable media can be include or storage be used for or about instruction execution system Any tangible medium of the program of system, device or equipment.Machine readable media can be that machine-readable signal medium or machine can Read storage medium.Machine readable media can include but is not limited to electronics, magnetic, it is optical, electromagnetism, infrared or partly lead Body system, device or equipment equipment or its any appropriate combination.The more detailed example of machine readable storage medium includes carrying one Or the electrical connection of multiple conducting wires, portable computer diskette, hard disk, random access memories (RAM), read-only memory (ROM), Erasable Programmable Read Only Memory EPROM (EPROM or flash memory), light storage device, magnetic storage apparatus or its is any appropriate Combination.

The foregoing is merely preferred embodiment of the present disclosure, are not limited to the disclosure, for those skilled in the art For, the disclosure can have various modifications and changes.It is all within the spirit and principle of the disclosure made by any modification, equivalent Replace, improve etc., it should be included within the protection domain of the disclosure.

Claims

1. a kind of method of structure patent value degree evaluation model, which is characterized in that including：

Choose the factor for influencing patent value degree and the subparameter for evaluating the factor；

The factor is evaluated according to the subparameter, using evaluation result as selected parameter；

Patent data sample is chosen according to preset patent value degree evaluation criterion；

The evaluation that relationship between selected parameter and patent value degree is embodied based on selected patent data Sample Establishing is public Formula.

2. the method for structure patent value degree evaluation model according to claim 1, which is characterized in that the influence patent The factor of action value includes at least one of applicant, agency, agent, inventor, industry and technology of patent, institute It includes at least one of the following to state parameter correspondingly：

To the evaluation result of the applicant of patent；

To the evaluation result of the agency of patent；

To the procuratorial evaluation result of patent；

The evaluation result of the inventor of patent；

To the evaluation result of the industry of patent；And

To the evaluation result of the technology of patent.

3. the method for structure patent value degree evaluation model according to claim 1, which is characterized in that the method is also wrapped It includes：After choosing patent sample, based on parameter and the patent value degree selected by selected patent data Sample Establishing it Between relation map；And

The judgement schematics are established based on the relation map.

4. the method for structure patent value degree evaluation model according to claim 1, which is characterized in that the basis is default Patent value degree evaluation criterion choose patent data sample include：

The patent for meeting preset first condition is chosen as first group of patent data sample；

The patent for meeting preset second preset condition is chosen as second group of patent data sample.

5. the method for structure patent value degree evaluation model according to claim 4, which is characterized in that described second group special Sharp data sample is chosen from first group of patent data sample, and the first condition includes being recorded in state's patent database Whole inventions and utility model, the second condition include at least one of the following：

Action for infringement of a patent occurred and obtains the patent won a lawsuit；

The patent for maintaining payment to expire to patent life；

Put and obtain the patent of return into effect；

It is requested invalid and is judged to maintain effective Chinese patent；

Form the patent of industry standard；

By with the patent more than preset times；

The patent of multinational application.

6. the method for structure patent value degree evaluation model according to claim 4, which is characterized in that described based on selected The judgement schematics of relationship include between the parameter selected by the embodiment of patent data Sample Establishing taken and patent value degree：

Calculate distribution of the selected parameter in first group of patent data sample；

Calculate distribution of the selected parameter in second group of patent data sample；

Distribution of the selected parameter in first group of patent data sample is calculated with selected parameter in second group of patent number According to the comparison result between the distribution in sample；And

Relational expression between the selected parameter of result fitting and the comparison result based on the comparison, selected by embodiment Parameter and patent value degree between relationship judgement schematics.

7. the method for structure patent value degree evaluation model according to claim 6, which is characterized in that described based on selected The judgement schematics of relationship include between the parameter selected by the embodiment of patent data Sample Establishing taken and patent value degree：

The corresponding patent of each of extracting parameter value；

To corresponding to each value patent carry out patent value degree evaluation, and based on evaluation result come the value of setup parameter to special The influence power of sharp action value；

Each value based on parameter obtains the relationship between parameter and patent value degree that embodies to the influence power of patent value degree Judgement schematics.

8. the method for structure patent value degree evaluation model according to claim 1, which is characterized in that the patent data Sample includes the patent data sample of single-piece patent or the patent data sample for the compound patent being converted by more than one piece patent..

9. the method for structure patent value degree evaluation model according to claim 1, which is characterized in that selected parameter There is multiple, relationship between the parameter and patent value degree selected based on selected patent data Sample Establishing embodiment Judgement schematics include：

For each selected parameter, it is based respectively on selected patent data Sample Establishing and embodies the parameter and patent value The sub- judgement schematics of relationship between degree；

By the weighted sum of sub- judgement schematics, as the judgement schematics for selected all parameters.

10. the method for structure patent value degree evaluation model according to claim 1, which is characterized in that selected ginseng Number is divided into multigroup, and described embodied based on selected patent data Sample Establishing is closed between selected parameter and patent value degree The judgement schematics of system include：

For each each group of selected parameter, be based respectively on selected patent data Sample Establishing embody this group of parameter with The sub- judgement schematics of relationship between patent value degree；

11. the method for structure patent value degree evaluation model according to claim 1, which is characterized in that further include：

Receive other parameters；

It is established at least one other model from evaluation patent value degree based on the other parameters, it will other described models The patent value degree evaluation model finally to be built is combined with the judgement schematics established.

12. a kind of computer readable storage medium, for storing instruction, described instruction makes processor hold when being executed by processor The method of row structure patent value degree evaluation model according to any one of claim 1 to 11.