CN104112170B

CN104112170B - The constructing method of patent leading indicators and application

Info

Publication number: CN104112170B
Application number: CN201410283508.7A
Authority: CN
Inventors: 车慧中
Original assignee: Shenzhen De Gaohang Intellectual Property Data Technology Co Ltd
Current assignee: Shenzhen de Gaohang intellectual property data Technology Co., Ltd.
Priority date: 2014-06-23
Filing date: 2014-06-23
Publication date: 2018-08-31
Anticipated expiration: 2034-06-23
Also published as: CN104112170A; TW201523498A; TWI526969B; CN109002983A; CN109002983B

Abstract

The present invention is a kind of constructing method of patent leading indicators and application, this method include：(1) multiple patent entities are set and to describe the multiple patent indexes and financial index of each patent entity；(2) collection period is set；(3) patent index data of the patent entity in collection period and financial index data are collected；(4) all patent index data and financial index data are formed into first panel data；(5) first panel data are formed into second panel data by translation operation program；(6) the first time Sequence Operation Theory program of setting time leading phase and offer based on the time leading phase, the independent variable of first time Sequence Operation Theory program are the patent index data of second panel data, and dependent variable is the financial index data of second panel data；(7) first threshold is set, by first time Sequence Operation Theory program, operation second panel data screen from multiple patent indexes and obtain the patent leading indicators for meeting first threshold.

Description

The constructing method of patent leading indicators and application

【Technical field】

The present invention relates to digital content treatment technology, refer in particular to excavate patent information to enterprise by the operation of patent information The association of industry financial information, and then establish the prediction technique and relative computer system of the leading financial information of patent information.

【Background technology】

As the fast-developing and intellectual property of scientific and technological industry is by attention, patent has been considered as the weight of industry or technology Want index.For obligee, patent has been not only the defence tool for protection and creation and product, is more become and competition pair The optimal attack weapon that hand competes on industry stage.If company can possess more patents, more have on its competitive field Representative and influence property, therefore patent becomes as a particularly important competition force information.

World Intellectual Property Organization (WIPO) once reported, the technology contents that patent includes, have 80% not other opinion It is disclosed in text, magazine, encyclopedia.Patent is the specific manifestation of innovation.List of the patent numbers with quality compared with tool advantage Position, energy and the quality of innovation also have more advantage compared with other competitors.Since patent has jural exclusive right, Has the effect of oligopoly to market.Therefore for the scientific and technological enterprises based on technical research, patent numbers and quality Compared with tool dominant, product sale and achievement also have considerable degree of superiority.

In the prior art, many paper studies are pointed out, the sales information of the leading product of quantity information of patent, for solution Releasing market development situation has the effect of primacy, such as following documents：

(1)Griliches,Z.(1720),Patent statistics as economic indicators：a survey,Journal of Economic Literature,28(4),PP.1661-1707.

(2)Ernst,H.(1725),Patenting strategies in the German mechanical engineering industry and their relationship to firm Performance,Technovation, 5(4),PP.225-240.

(3)Adams,K.,D.Kim,F.L.JoLz,Trost,R.P.and Mastrogianis,G.(1727), Modeling and Forecasting U.S.Patent Application Filings,Journal of Policy Modeling,19(5),PP.491-535.

(4)Ernst H.(1727),The Use of Patent Data for Technological Forecasting：The Diffusion of CNC-Technology in Machine Tool Industry,Small Business Economics,9(4),PP.361-381.

(5)Ernst H.(2001),Patent applications and subsequent changes of Performance：evidence from time-series cross-section analyses on the firm level,Research Policy,30(1),PP.143-157.

The above-mentioned prior art points out that patent information becomes more important compared to other market informations, more merits attention, therefore with The method of patent information prediction markets information is just increasingly becoming investment and an important research topic in valuation of enterprise.

In the prior art, United States Patent (USP) US6175824 and US6832171 is put forward for the first time with patent information to listed company's wealth Business performance is assessed.United States Patent (USP) US6175824 and US6832171 leads to for the more high-tech stock of listed company's patent number Multivariable regressive analysis model is crossed, analyzes the association of listed company's patent index over the years (PI) and financial index over the years (FI), most The Estimate equation formula based on patent index (PI) is exported afterwards, its Value of Science ＆ Technology is calculated with this Estimate equation formula, then with this Innovative value compares its price value ratio (Market-to-Book Ratio), and being considered as if Value of Science ＆ Technology is more than price value ratio has investment latent Power is considered as no investment value, the tool thereby selected stocks as investment if Value of Science ＆ Technology is less than price value ratio.Its patentee CHI More by this, the Estimate equation formula based on patent information pushes masses to Research, establishes its commercial pattern of exclusive business accordingly.

Prior art US6175824 and US6832171 have its uniqueness, but disadvantage be its analyze patent index (PI) with The equation of financial index (FI) is typical Multivariable regressive analysis model, and independent variable is the patent index of a time series (PI), dependent variable is the financial index (FI) of same time sequence.Therefore the multiple regression equation established, when inputting some Between put patent index (PI) when, obtained output the result is that the time point corresponding financial index (FI), rather than " future " Financial index (FI).

In the prior art, United States Patent (USP) US6556992, US7657476 and US7716226, using the sustainment rate of patent as base Plinth, then other patent indexes (PI) of arranging in pairs or groups are developed another set of to listed company's innovation ability with Multivariable regressive analysis model Appraisal procedure, and 300 listed companies of the U.S. are picked according to this as constituent stocks, after being weighted calculating, issue the whole world first A patent index (OT300Patent Index).However it is polynary used in US6556992, US7657476 and US7716226 Regression analysis model, independent variable patent index (PI) and dependent variable financial index (FI) be all belong to same time sequence, though Its right independent variable is different from the prior art US6175824 and US6832171, but the multiple regression equation of its foundation, works as input When patent index (PI) at some time point, obtained output result is still the time point corresponding financial index (FI), and Be not " future " financial index (FI).

, it must be understood that investing the Practical Operation selected stocks, investment institution is not intended to just make a profit instantly in investment for we, And it is desirable to just make a profit in some following time point arbitrage.That is, investment institution is in investment, it is desirable to which grasp is worked as Lower information to " future " and profit have " prediction " and effect, investment risk could be reduced, it is ensured that returns of investment.The above-mentioned prior art US6175824, US6832171, US6556992, US7657476 and US7716226 do not have substantially " prediction " effect.

Another aspect, the above-mentioned prior art US6175824, US6832171, US6556992, US7657476 and US7716226 It is all based on information possessed by American invention granted patent information and Index Establishment model.Wherein US6175824 with The most crucial patent index (PI) that US6832171 is used is that novelty (NoveLy) examines and creativeness/unobviousness (Nonobviousness) quotation of Patent Citation and non-patent literature used in examining, including preceding quotation (Backward Citation) with rear quotation (Forward Citation)；The most core that US6556992, US7657476 and US7716226 are used Heart patent index (PI) is the sustainment rate of invention granted patent.However, there is no utility model systems for United States Patent (USP), but to China For the vast patent information in continent, more without the utility model patent of entity examination procedure, patent of invention is less；To invention For patent, the granted patent examined by entity is less, and early stage publication is more.Again for invention granted patent, specially The citation information that sharp database is issued is that the granted patent of recent years just gradually starts to disclose, and is not traced back to fully previously Granted patent；Again for citation information, used comparison when only preceding quotation, i.e. patent entity (PE) examination at present File, but rear quotation is not disclosed, i.e., other patents examined as documents after license.So existing skill The content and method of its exposure of art US6175824, US6832171, US6556992, US7657476 and US7716226, is not suitable for The patent information of China's Mainland, therefore the financial information of the listed company of China's Mainland can not be assessed.

Edge this, the financial data for the listed company of China's Mainland and its patent information in China's Mainland, if there is tool It is standby " prediction " effect association model and realize the computer system of this association model, not only help patent information analysis Develop with the technical strength that utilizes, more can investment promotion field capitalized method positive development, and to the research and development of industrial technology Positive support effect is played with innovation.

【Invention content】

Based on above-mentioned defect in the prior art is improved, the main purpose of the present invention is to provide a kind of building for patent leading indicators Structure method (100), patent leading indicators (172) are predicting the financial information of patent entity (PE), and patent leading indicators (172) there are one the information of the financial index (FI) of information leads patent entity (PE) the preset time leading phase (L), The constructing method (100) of patent leading indicators includes：

(1) set multiple patent entities (PE) and to describe multiple patent indexes (PI) of each patent entity (PE) with Financial index (FI), each patent index (PI) are obtained by the patent information operation of each patent entity (PE).

(2) setting collection period (121), collection period (121) is by time interval (T) and time issue (N) institute structure At time issue (N) is the integer not less than two；

(3) each patent entity (PE) is collected in collection period (121), and the corresponding patent of each time interval (T) institute refers to Mark data (131) and financial index data (132)；

(4) by multiple patent index data (131) of multiple patent entities (PE) and multiple financial index data (132) groups At first panel data (141)；

(5) first panel data (141) are formed into normal distribution and standard scores by translation operation program (152) The second panel data (151) of numberization；

(6) the first time Sequence Operation Theory program of setting time leading phase (L) and offer based on the time leading phase (L) (161), the time leading phase (L) includes at least one time interval (T), the independent variable of first time Sequence Operation Theory program (161) For a patent index data (131) of second panel data (151), the dependent variable of first time Sequence Operation Theory program (161) For the financial index data (132) of second panel data (151)；

(7) setting first threshold (171) passes through first time Sequence Operation Theory program (161) and time leading phase (L), fortune Second panel data (151) are calculated, is obtained from screening in multiple patent indexes (PI) and meets at least one special of first threshold (171) Sharp leading indicators (172).

The constructing method (100) of patent leading indicators proposed by the invention is objective and rigorous, is not only particularly suitable for China Continent patent information, including disclosure of the invention patent, invention granted patent, utility model patent and design patent；Also simultaneously It is suitble to be directed to other various countries' patent information, excavates the patent leading indicators (172) with leading companies financial information.

It is yet another object of the invention to provide a kind of constructing method of the leading equation of patent (500), the leading equations of patent For formula (501) to predict the financial information of patent entity (PE), the leading equation of patent (501) generates a leading score of patent (502), there are one the preset time leading phases for the financial information of the leading patent entity (PE) of the leading score of patent (502) (L).The constructing method (500) of the leading equation of patent (501) includes：

(1) obtain second panel data (151) with multiple patent leading indicators (172), second panel data (151) with it is more A patent leading indicators (172) are obtained by the constructing method (100) of patent leading indicators above-mentioned.

(2) multiple patent leading indicators (172) are based on, third panel data is formed from second panel data (151) screening (521)；

(3) the second time series operation program (531) based on the time leading phase (L), the second time series operation are provided The independent variable of program (531) is multiple patent leading indicators (172) of third panel data (521), the second time series operation The dependent variable of program (531) is the financial index (FI) of second panel data (151)；

(4) setting second threshold (541) passes through the second time series operation program (531) and time leading phase (L), fortune Third panel data (521) is calculated, is obtained from screening in multiple patent leading indicators (172) and meets the multiple of second threshold (541) Patent core index (542) simultaneously generates the leading equation of patent (501), and the leading equation of patent (501) is substantially by multiple special Sharp core index (542) and its corresponding weight coefficient (543) are formed.

It is a further object of the present invention to provide a kind of methods (600) of assessment enterprise investment potentiality, including：

(1) patent information (612) of multiple enterprises (611) is collected；

(2) a leading equation of patent (501) is provided, the leading equation of patent (501) is by the leading equation of aforementioned patent The constructing method (500) of formula is obtained, the leading equation of patent (501) substantially by multiple patent core index (542) and its Corresponding weight coefficient (543) is formed；

(3) patent information (612) for being based on each enterprise (611) calculates each corresponding patent core index (542) of enterprise Data (631)；

(4) data (631) of the patent core index (542) of each enterprise (611) are based on, are passed through leading equation (501) Calculate and generate the leading score of patent (502) of each enterprise (611)；

(5) the leading score of patent (502) of each enterprise (611) is ranked up by a collator (651), is sorted As a result (652) represent the investment potential of enterprise (611).

The above-mentioned higher person of the leading score of patent (502), represent enterprise the time leading phase (L) later corresponding finance refer to The data for marking (FI) are higher；The lower person of the leading score of patent (502), represent enterprise the time leading phase (L) later it is corresponding The numerical value of financial index (FI) is lower.Since the data height of business finance index (FI) directly expresses the good of its business performance Bad, the higher person of numerical value of financial index (FI), performance better, more has investment value.Due to the leading score of patent (502) represent enterprise the time leading phase (L) later corresponding financial index (FI) numerical value, therefore possessed by enterprise specially The height ranking of the leading score (502) of profit, just can be from wherein picking out the object with investment potential.

Another object of the present invention is to provide a kind of computer system (700) of assessment enterprise investment potentiality, including patent The leading score computing device (730) of information collecting device (710), index computation device (720), patent and score collator (740).Wherein, patent information harvester (710) collects the patent information (612) of multiple enterprises；Index computation device (720) Patent information (612) based on each enterprise calculates, and generates the data (631) of the corresponding patent core index (542) of enterprise；Patent Leading score computing device (730) passes through patent leading side according to the data (631) of the patent core index (542) of each enterprise Formula (501) calculates and generates the leading score of patent (502) of each enterprise, and the leading equation of patent (501) is by patent above-mentioned The constructing method (500) of leading equation is obtained, and the leading equation of patent (501) is substantially by multiple patent core index (542) and its corresponding weight coefficient (543) is formed；Score collator (740) by the leading score of multiple patents (502) into Row sequence, ranking results (652) represent the investment potential of multiple enterprises.

The method (600) and computer system (700) of assessment enterprise investment potentiality proposed by the invention are based on big number According to, the achievement of objective operation, rigorous verification, the technical strength for not only helping patent information analysis and utilizing develops, can more promote Positive support effect is played into the positive development of the capitalized method of investment field, and to the research and development of industrial technology and innovation.

Other particular technique characteristics for the present invention and non-obvious prominent effect, will be in following sections specifically It is bright.

【Description of the drawings】

Fig. 1 is the first preferred embodiment proposed by the present invention, is a kind of stream of the constructing method (100) of patent leading indicators Cheng Tu；

Fig. 2 is that Shanghai listed company of exchange analyzes the quantity of parent and industry is distributed；

Fig. 3 is the industry distribution that Shanghai listed company of exchange analyzes sample；

Fig. 4 is the schematic diagram of first panel data (141)；

Fig. 5 is the patent leading indicators (172) that the leading phase is 1 year；

Fig. 6 is the patent leading indicators (172) that the leading phase is 2 years；

Fig. 7 is the patent leading indicators (172) that the leading phase is 3 years；

Fig. 8 is the patent leading indicators (172) that the leading phase is 4 years；

Fig. 9 is the second preferred embodiment proposed by the present invention, is a kind of constructing method (500) of the leading equation of patent Flow chart.

Figure 10 A, Figure 10 B, Figure 10 C are the life of patent core index (542) in the second preferred embodiment proposed by the present invention At process.

Figure 11 is third preferred embodiment proposed by the present invention, for a kind of method (600) of assessment enterprise investment potentiality Flow chart.

Figure 12 is the 4th preferred embodiment proposed by the present invention, for a kind of computer system of assessment enterprise investment potentiality (700) configuration diagram.

【Specific implementation mode】

The present invention mainly discloses constructing method (100) and the application of a kind of patent leading indicators, wherein involved patent The basic knowledge of information, patent index (PI), financial index (FI) etc. has usually intellectual institute energy for correlative technology field Understand, therefore with following description, no longer make complete description.Meanwhile the schema hereinafter to compare, only expression and the present invention The related signal of feature also need not completely be drawn according to actual size, first explanation.

It is a kind of constructing method of patent leading indicators referring to FIG. 1, for the first preferred embodiment of proposition of the invention (100), patent leading indicators (172) are predicting the financial information of patent entity (PE), and the letter of patent leading indicators (172) The information of the financial index (FI) of leading patent entity (PE) is ceased there are one the preset time leading phase (L), and patent is leading The constructing method (100) of index includes：

Step 110：Set multiple patent entities (PE) and to describe multiple patent indexes of each patent entity (PE) (PI) it is obtained by the patent information operation of each patent entity (PE) with financial index (FI), each patent index (PI).

Step 120：Collection period (121) is set, collection period (121) is by time interval (T) and time issue (N) It is constituted, time issue (N) is the integer not less than two.

Step 130：Each patent entity (PE) is collected in collection period (121), each time interval (T) institute is special accordingly Sharp achievement data (131) and financial index data (132).

Step 140：By multiple patent index data (131) of multiple patent entities (PE) and multiple financial index data (132) composition first panel data (141).

Step 150：First panel data (141) are formed into normal distribution by translation operation program (152) and are marked The second panel data (151) of quasi- fractionation.

Step 160：The first time Sequence Operation Theory journey of setting time leading phase (L) and offer based on the time leading phase (L) Sequence (161), this time leading phase (L) include at least one time interval (T), first time Sequence Operation Theory program (161) from Variable is the patent index data (131) of a patent index (PI) of second panel data (151), first time Sequence Operation Theory The dependent variable of program (161) is the financial index data (132) of the financial index (FI) of second panel data (151).

Step 170：First threshold (171) is set, gradually uses first time Sequence Operation Theory program (161) and time leading Phase (L), operation second panel data (151) obtain from screening in multiple patent indexes (PI) and meet first threshold (171) extremely A few patent leading indicators (172).

In above-mentioned steps 110, patent entity (PE) is to possess patent right and can be by right master that patent right operation is made a profit Body, the preferably listed company of public publication, but be not limited with listed company；The present embodiment is also suitable for private company, As long as external sources of finance entrance can be received, shares the subject of right of equity and return on equity, all belong to the scope of application of the present embodiment. Also, the patent that the first preferred embodiment is targeted, it is granted patent not limit, as long as that is announced in patent database is special Profit all can, including disclosure of the invention patent, invention granted patent, utility model patent, design patent etc..Meanwhile first compared with The method that good embodiment is proposed is different from asking for United States Patent (USP) in addition to can effectively solve the information content of China's Mainland patent Topic, more can be adapted for all parts of the world area patent in fact.

In terms of patent index (PI), by taking the patent of China's Mainland as an example, patent index (PI) includes but not limited to following Items can be by the quantitative index of the automatic operation of computer, such as：

P1：Patent sum

P2：Disclosure of the invention patent sum

P3：Utility model patent sum

P4：Design patent sum

P5：Invention granted patent sum

P6：Disclosure of the invention patent is averaged the patent service life

P7：Utility model patent is averaged the patent service life

P8：Design patent is averaged the patent service life

P9：Invention granted patent is averaged the patent service life

P10：Invention granted patent is averaged the period under review

P11：Current disclosure of the invention patent number

P12：Current utility model patent number

P13：Current design patent number

P14：Current invention granted patent number

P15：Current invention granted patent is averaged the period under review, stops to authorized announcement date from the applying date

P16：Current disclosure of the invention patent IPC code sum

P17：Current utility model patent IPC code sum

P18：Current invention granted patent IPC code sum

P19：Current disclosure of the invention patent IPC code average

P20：Current utility model patent IPC code average

P21：Current invention granted patent IPC code average

P22：Current disclosure of the invention patent specification total page number

P23：Current utility model patent specification total page number

P24：Current invention granted patent specification total page number

P25：Current disclosure of the invention patent specification average number of pages

P26：Current utility model patent specification average number of pages

P27：Current invention granted patent specification average number of pages

P28：The claim sum of current disclosure of the invention patent

P29：The claim sum of current utility model patent

P30：The claim sum of current invention granted patent

P31：The claim average of current disclosure of the invention patent

P32：The claim average of current utility model patent

P33：The claim average of current invention granted patent

P34：The exclusive rights sum of current disclosure of the invention patent

P35：The exclusive rights sum of current utility model patent

P36：The exclusive rights sum of current invention granted patent

P37：The exclusive rights average of current disclosure of the invention patent

P38：The exclusive rights average of current utility model patent

P39：The exclusive rights average of current invention granted patent

P40：The attached drawing number of current disclosure of the invention patent specification

P41：The attached drawing number of current utility model patent specification

P42：The attached drawing number of current invention granted patent specification

P43：The attached drawing average of current disclosure of the invention patent specification

P44：The attached drawing average of current utility model patent specification

P45：The attached drawing average of current invention granted patent specification

In above-mentioned patent index (PI), " current " to refer to that the data operation of a patent index (PI) is limited in some specific In time interval (T), citing, if time interval (T) is year, P11 (current disclosure of the invention patent number) represents some year All invention disclosed patent numbers within interior 1~December；(current invention granted patent is averaged the period under review P15, from the applying date Stop to authorized announcement date) then represent all bulletins within 1~December in some year invention granted patent average the period under review. If time interval (T) is season, P11 (current disclosure of the invention patent number) represents in some season all disclosures within 3 months Patent of invention number；P15 (current invention granted patent is averaged the period under review, stops to authorized announcement date from the applying date) is then represented In some season within 3 months the invention granted patent of all bulletins average the period under review.Yu this analogize.

Above-mentioned 45 patent indexes (PI) can repartition as disclosure of the invention patent class, invention granted patent class, practicality newly Type patent class and design patent class, wherein having to the index for describing disclosure of the invention patent：

P2：Disclosure of the invention patent sum

P6：Disclosure of the invention patent is averaged the patent service life

P11：Current disclosure of the invention patent number

P16：Current disclosure of the invention patent IPC code sum

P19：Current disclosure of the invention patent IPC code average

P28：The claim sum of current disclosure of the invention patent

P31：The claim average of current disclosure of the invention patent

P34：The exclusive rights sum of current disclosure of the invention patent

Index to describe invention granted patent has：

P5：Invention granted patent sum

P9：Invention granted patent is averaged the patent service life

P10：Invention granted patent is averaged the period under review

P14：Current invention granted patent number

P18：Current invention granted patent IPC code sum

P21：Current invention granted patent IPC code average

P24：Current invention granted patent specification total page number

P27：Current invention granted patent specification average number of pages

P30：The claim sum of current invention granted patent

P33：The claim average of current invention granted patent

P36：The exclusive rights sum of current invention granted patent

P39：The exclusive rights average of current invention granted patent

Index wherein to describe utility model patent has

P3：Utility model patent sum

P7：Utility model patent is averaged the patent service life

P12：Current utility model patent number

P17：Current utility model patent IPC code sum

P20：Current utility model patent IPC code average

P23：Current utility model patent specification total page number

P26：Current utility model patent specification average number of pages

P29：The claim sum of current utility model patent

P32：The claim average of current utility model patent

P35：The exclusive rights sum of current utility model patent

P38：The exclusive rights average of current utility model patent

P41：The attached drawing number of current utility model patent specification

Index to describe design patent has

P4：Design patent sum

P8：Design patent is averaged the patent service life

P13：Current design patent number

The patent index (PI) of above-mentioned P1 to P45 is the description most sufficient quantitative index of China's Mainland patent items speciality, Which just have the effect of predicting financial information as wherein patent index (PI)It needs through rigorous analysis and verification, Where core of the invention, it will continue to illustrate in follow-up length.

In terms of financial index (FI), used in the present invention refers to the index for expressing performance, can repay Debt capacity index, operation ability index, profitability guideline, developing ability index and stock price indices etc., the present invention does not add To limit.Wherein profitability guideline can be net assets income ratio ROE (Rate of Return on Common Shareholder ' s Equity), assets return ROA (Rate of Return on Assets), earnings per share EPS (Earnings Per Share), price value ratio MTB (Market-to-Book Ratio) etc..

In above-mentioned steps 120, the unit of time interval (T) can be week, the moon, season, half a year or year etc..The setting time phase Number (N) and time interval (T) are to be used with verification to collect enough sample datas to establish model.If phase time Number (N) is set as 5, and time interval (T) is set as year, and collection period (121) is 5 years, subsequently just needs to collect 5 years samples Notebook data；If time issue (N) is set as 6, time interval (T) is set as the moon, and collection period (121) is 6 months, after The continuous sample data for just needing to collect 6 months.Due to it is proposed by the present invention be prediction model, that is, it is pre- with the data of previous phase The data of current period are surveyed, or with the data of the data prediction of a current period lower phase, therefore, time issue (N) at least needs as 2 phases, It is able to verify that the conspicuousness of prediction result.

In above-mentioned steps 130：If collection period (121) is this 5 years 2008 to 2012, time interval (T) is one Year, time issue (N) is 5, then must collect each patent entity (PE) in this 5 years 2008 to 2012, each time All patent indexes (PI) patent index data (131) and financial index (FI) financial index data (132).

In step 140, so-called panel data, also known as panel data or comprehensive column data are time series data and cross The mixing of cross-section data refers to M cross section object being observed in the data set of time issue (N), a shared M × N number of data set. By taking the present embodiment as an example, if having 375 patent entities (PE), collection period (121) be 5 years, i.e., 2008 to 2012, Time interval (T) be year, at this time just formed 375 cross section objects being observed, 5 times time (2008,2009,2010, 2011,2012) data set of 45 patent index (PI) and financial index (FI).

Traditional time series data (is become certainly for analyzing observation of the single object being observed in multiple times Amount and dependent variable) connection.Traditional cross-sectional data, be for analyze multiple objects being observed single time point observation The connection of (independent variable and dependent variable).Both data are not all suitable for method proposed by the present invention because the present invention have it is multiple Object being observed and there are multiple time points, each time point has multiple independents variable and dependent variable again.And panel data, be for The association for analyzing observation of multiple specific cross section objects being observed at multiple time points, due to increasing for observation, The sampling precision of estimator can be increased, obtain more consistent estimators and efficient estimator and obtain more dynamic letters Breath, therefore the present invention is analyzed using panel data.

In the present embodiment, between step 140 and step 150, a normal distribution (Normal can be further included Distribution) check problem (145), to each patent index data (131) and each financial index data being collected into (132), the state of its normal distribution is examined.Because if normal distribution is not presented for data, during establishing analysis model, Often model is caused to collapse because of excessively high do not restrain of error.Therefore, for the data of normal distribution are not presented, it is necessary to impose appropriate Translation operation is first converted into the state of normal distribution, then carries out the analysis of independent variable and dependent variable.

There are commonly following several for normal distribution-test program (145)：Anderson-Darling check problems, Ryan- Joiner check problems, Kolmogorov-Smirnov check problems etc., or can more easily observe the coefficient of skewness and peak Spend coefficient can inference data normal distribution situation, which kind of check problem the present invention be not intended to limit using.For not meeting just The data of state distribution need to impose processing operation to generate the data of normal distribution, wherein Box-Cox conversion programs are common Mode, the present invention is also not limited.It must be stressed that if normal state point is substantially substantially presented in initial data Cloth, there is no need to impose processing operation again.

Above-mentioned steps 150 are further described below, and the distribution curve of normal distribution is substantially that a center line it is expected It is worth (average value) and curve symmetrical, being unfolded as unit of standard deviation and to both sides extension.When being averaged for two kinds of data sets Value and standard deviation all differ widely, even if both data sets are all in normal distribution, but scale gap can be very big, are unfavorable for point Analysis is compared.Therefore during establishing model, preferably, the data of normal distribution are passed through a translation operation program (152) try again processing, and data are done regular (Normalization) conversion.The translation operation journey being wherein relatively often used Sequence (152) is " Z score " operation.Pass through " Z score " data set after operation, desired value (average value) is all 0, and standard deviation is all It is 1.After if the data set of all independents variable and dependent variable is all converted into the data set of normal distribution and criterion score, more It is easy from wherein excavating association.

The first panel data (141) of the present embodiment, in step 150, after translation operation program (152), conversion For normal distribution and the second panel data (151) of criterion score.

In a step 160, first time Sequence Operation Theory program (161) is preferably the Granger CaFpngerusality test model of unitary (Granger Causality Test Model).Granger CaFpngerusality test model is Nobel prize in economics winner gram in 2003 Lay husband Granger (Clive W.J.Granger) is started, for leading and backward between the economic variable of analysis time sequence Relationship.Its basic concepts is if that, there are two variable X and Y, variable X occurs formerly, after variable Y is happened at, and to pass through Granger After Causality Test model, the variable X influence significant to the probability of happening of variable Y is set up in verification, then variable X is claimed to lead over change Y or variable X are measured as the leading indicators of variable Y.In the economic variable of Granger CaFpngerusality test model treatment, independent variable with Dependent variable is all the variable of time series, and basic operations model is regression analysis model, but before regression analysis, it is first right The offset of independent variable or dependent variable setting a period of time set the leading phase or fall behind the phase, then inspects with the leading phase or fall The regression analysis model in later stage is with appropriateness, to verify the leading effect of independent variable or fall behind effect.

The model of Granger CaFpngerusality test model examines obtained P_value, general P_value with the common F of appropriateness< 0.1, it is acceptable for model；If P_value<0.05, it is that model is good；If P_value<0.005, it is that model is splendid.Briefly, If independent variable is X, dependent variable Y, and via the P_value obtained after Granger CaFpngerusality test model<0.05, then it represents that hundred In/nine ten five confidence interval, independent variable X is Y for dependent variable, has leading effect.Granger CaFpngerusality test model is another One common model is R square values with appropriateness, and R square values are between 0 to 1, better closer to 1, poorer closer to 0；With 1 It is best, indicates that model is perfect, have no error；It is worst with 0, indicates that error is infinitely great.

In Granger CaFpngerusality test model, the number for indicating independent variable is not limited.Also that is, Granger CaFpngerusality test model The leading effect of dependent variable of an independent variable pair can be analyzed, claims the Granger CaFpngerusality test model of unitary at this time, such as walks Rapid 160；The leading effect that dependent variable of multiple independents variable pair can also be analyzed simultaneously, claims polynary Granger Causality to examine at this time Test model.If but we are it must be understood that there are serious synteny or the data discrete mistakes of independent variable between multiple independents variable Height, polynary Granger CaFpngerusality test model often collapse in calculating process.It is event, in a step 160, preferably, being first By the patent index data (131) of other patent index (PI) to the financial index data of a other financial index (FI) (132) the Granger CaFpngerusality test model calculation of unitary is imposed, could be dug from 45 patent indexes (PI) in step 170 Other patent index (PI) of the leading effect with conspicuousness, referred to as patent leading indicators (172) are excavated, and excludes other necks First effect does not have the patent index (PI) of conspicuousness, each the patent leading indicators (172) excavated at this time refer to finance The leading effect for marking (FI) all has conspicuousness.

In step 170, the setting of first threshold (171) is crucial, and first threshold (171) is set too stringent, can It can not excavate the patent leading indicators (172) with conspicuousness；First threshold (171) is set too loose, may be too many The insufficient patent index of conspicuousness (PI) is also mistaken as patent leading indicators (172) and excavates.If first time sequence Column operations program (161) takes the Granger CaFpngerusality test model of unitary, it has been observed that obtained by model is examined with the common F of appropriateness P_value, general P_value<0.1, it is acceptable for model, reach 90% confidence interval；If P_value<0.05, it is mould Type is good, reaches 95% confidence interval；If P_value<0.005, it is that model is splendid, reaches 99.5% confidence interval.This When we can first to set first threshold (171) as 0.1, it is first preliminary understand can excavate how many patent leading indicators (172) Reach 90% confidence interval, if quantity is few, sets first threshold (171) as 0.1；It, can be again if there are many quantity First threshold (171) is set as 0.05 or 0.005, you can excavating crucial, leading effect has the patent of splendid conspicuousness Leading indicators (172).

In the case where first threshold (171) is set in 0.05, after excavating patent leading indicators (172), if The time leading phase (L) set in step 160 is 1 year, observes each patent entity (PE) its patent leading indicators (172) and exists The numerical value in this year and can reach as patent entity (PE) in the prediction result of the financial index numerical value in next year 95% confidence interval.If first threshold (171) is set in 0.005, and the time leading phase (L) set in step 160 is It 3 years, after excavating patent leading indicators (172), observes each patent entity (PE) its patent leading indicators (172) and exists The numerical value in this year, the prediction result of financial index data that can be as patent entity (PE) after 3 years, and can reach 99.5% confidence interval.

Below by using the listed company of Shanghai exchange as patent entity (PE), the first preferred embodiment is described in further detail Implementation process.

Shanghai exchange shared 951 listed companies by end in 2012, and the parent of patent entity (PE) is 951, Its industry distribution and proportion are as shown in Figure 2.

The present invention is the prediction model for establishing patent index (PI) to the primacy of financial index (FI), it is therefore necessary to be considered Subsidiary's structure of listed company.If the finance combination of subsidiary calculates together to parent company, the patent index of subsidiary (PI) parent company must be also merged into calculate together.So 951 listed companies must all investigate its subsidiary's structure.By After investigation, it has been found that in 951 listed companies, subsidiary's number average is 13.6, median 9.0.By subsidiary The difference of average value and median, can learn subsidiary that most of listed company possessed estimate at 10 hereinafter, but its In have a small number of listed companies that there is larger numbers of subsidiary, therefore drawn high average value.By investigation, possess most subsidiaries Listed company, it has up to 174 subsidiaries.

In terms of the financial index (FI) of listed company, we tentatively select net assets income ratio ROE (Rate of Return on Common Shareholder ' s Equity) as representing, carry out subsequent analysis.At patent index (PI) Aspect, we are using 45 patent indexes (PI) such as P1~P45 above-mentioned.

Aspect is selected in effective sample, we set from 2008 to 2012 year this financial index (FI) 5 years There must be a data, and end in 2012 has at least 50 patent sums (including disclosure of the invention, invention mandates, utility model, outside See the totallings such as design), 951 listed companies are qualified by screening finally obtains 375, i.e. patent entity (PE) Sample is 375, and industry distribution and proportion are as shown in Figure 3.

375 patent entities (PE) are collected to the data (131) of its 2008 to 2012 45 patent indexes (PI) And after the data (132) of financial index (FI), first panel data (141) are formed, to carry out subsequent analysis.

Referring to FIG. 4, for 375 patent entities (PE) of the invention, 5 times time (2008,2009,2010, 2011,2012) the financial index data of the patent index data (131) and financial index (FI) of 45 patent indexes (PI) (132) a part of content of constituted first panel data (141).

After generating first panel data (141), it is necessary to data therein, the data either as independent variable or conduct The data of dependent variable analyze its data distribution situation, check whether and meet normal distribution.Only meet the data of normal distribution, Independent variable and the relational model of dependent variable are just easily established, otherwise most probably because error causes model to collapse very much greatly.

Pass through Kolmogorov-Smirnov check problems, it has been found that in first panel data (141), financial index (FI) financial index data (132) are substantially substantially presented normal distribution, and the patent index of each patent index (PI) Normal distribution is not presented in data (131).Therefore, we to the patent index data (131) of each patent index (PI) again Imposing Box-Cox conversion programs makes it that normal distribution substantially be presented.

Now, our patent index data (131) and financial index to the patent index (PI) of all normal distribution (FI) financial index data (132), and Z score translation operation program (152) is imposed again, keep the desired value of each data (average Value) all it is 0, standard deviation is all 1.First panel data (141) are all converted to the of normal distribution and criterion score at this time Two panel datas (151).

In second panel data (151), independent variable is 45 patent indexes of 375 patent entities (PE) in 5 years (PI), dependent variable is financial index (FI) of 375 patent entities (PE) in 5 years：ROE, now we use unitary Glan Outstanding Causality Test model (Granger Causality Test Model) examines neck of each independent variable to dependent variable successively First relationship.

A critically important parameter is in Granger CaFpngerusality test model (Granger Causality Test Model) The setting of time leading phase (L).In the present embodiment, we be set separately the time leading phase (L) be 1 year, 2 years, 3 years, 4 years, come Observe the precondition of the patent index (PI) to financial index (FI).

When the time set the leading phase (L) as 1 year when, the independent variable that uses is 2008,2009,2010,2011 this 4 years Patent index (PI)；Corresponding dependent variable is 2009,2010,2011,2012 this 4 years financial index (FI).Wherein, Patent index (PI) in 2008 is to match to 2009 financial index (FI), patent index (PI) in 2009 be match to 2010 financial index (FI), patent index (PI) in 2010 are to match to 2011 financial index (FI), special in 2011 Sharp index (PI) is to match to 2012 financial index (FI), is so just able to verify that patent index (PI) leads financial index (FI) The conspicuousness of first effect.At this time in second panel data (151), there are 375 × 4=1500 data that can use.

When the time set the leading phase (L) as 2 years when, the independent variable that uses is 2008,2009,2010 this 3 years patent Index (PI)；Corresponding dependent variable is respectively 2010,2011,2012 this 3 years financial index (FI), at this time second panel number According in (151), there are 375 × 3=1125 data that can use.

When the time set the leading phase (L) as 3 years when, the independent variable that uses is 2008,2009 this 2 years patent index (PI)；Corresponding dependent variable is respectively 2011,2012 this 2 years financial index (FI), at this time in second panel data (151), There are 375 × 2=750 data that can use.

When the time set the leading phase (L) as 4 years when, patent index (PI) that the independent variable that uses is 2008；Accordingly Dependent variable be financial index (FI) in 2012, at this time in second panel data (151), only 375 data can use.

First time Sequence Operation Theory program (161) we use unitary Granger CaFpngerusality test model (Granger Causality Test Model), with appropriateness, we are examined model using F, and model uses three kind of first threshold with appropriateness P_value It is worth (171)：

(1)P_value<0.1, independent variable patent index (PI) has the precondition of dependent variable financial index (FI) can The conspicuousness of receiving, up to 90% confidence interval；

(2)P_value<0.05, independent variable patent index (PI) has the precondition of dependent variable financial index (FI) Good conspicuousness, up to 95% confidence interval；

(3)P_value<0.005, independent variable patent index (PI) has the precondition of dependent variable financial index (FI) Splendid conspicuousness, up to 99.5% confidence interval.

After operation by second panel data (151) by unitary Granger CaFpngerusality test model, we successfully have found it In have certain patent indexes (PI) that there is conspicuousness, referred to as patent leading indicators to the primacy of financial index (FI) really (172).Fig. 5 to Fig. 8 be respectively the leading phase be 1 year, 2 years, 3 years, 4 years and with conspicuousness patent leading indicators (172), * represents P-value in figure<0.1；* represents P-value<0.05；* * represent P-value<0.005.

Fig. 5 to Fig. 8 provides a good forecast function, if we compare 1 year later portfolio performance of concern, sees Examine the numerical value change for the patent leading indicators (172) that the leading phase shown in Fig. 5 is 1 year；After if we compare concern 2 years Portfolio performance, then the leading phase shown in observation chart 6 be 2 years patent leading indicators (172) numerical value change；If concern 3 years Or 4 years portfolio performances, then the numerical value change of observation chart 7 or patent leading indicators shown in Fig. 8 (172) respectively.

If we are more concerned with wherein most crucial while 1 year leading, 2 years, 3 years, 4 years patent leading indicators (172), then find out while appearing in the patent leading indicators (172) of Fig. 5 to Fig. 8.Such as beneath 5 patent leading indicators (172), it is exactly while all there is conspicuousness person leading 1 year, 2 years leading, leading 3 years and leading 4 years.

P12 (current utility model patent number)

P23 (current utility model patent specification total page number)

P29 (the claim sum of current utility model patent)

P35 (the exclusive rights sum of current utility model patent)

P41 (the attached drawing number of current utility model patent specification)

This five patent leading indicators (172) of P12, P23, P29, P35, P41 belong to the patent index of utility class (PI), reason is that the listed company of Shanghai exchange with manufacturing industry (industry code C) is most, and includes to the present embodiment Also it is most with manufacturing industry in 375 patent entities (PE) of effective sample, and manufacturing patent is accounted for utility model in the past It is most of, thus the patent index (PI) of utility class be just easy to take off during the foundation of prediction model naturally it is clever and Go out.And this result also reveals another information simultaneously：Even if many experts think the degree of innovation of utility model patent it is relatively low, Value is not high, but is to predict that the performance of its financial index (FI) has conspicuousness for the listed company of Shanghai Stock Exchange Forecasting tool.

Herein it has to be emphasized that when the effective sample for establishing model, i.e., when patent entity (PE) changes, finally show one's talent Patent leading indicators (172) will change according to sample properties.For example, when the effective sample for establishing model uses Shenzhen The listed company of exchange, the patent leading indicators (172) shown one's talent necessarily are partially different than to be listed with Shanghai exchange The patent leading indicators (172) that company is excavated by sample.Namely because patent leading indicators (172) can be according to sample spy Property and change, the method that first preferred embodiment of the invention is proposed has more universality, is suitable for various sample clusters Body.We can excavate leading suitable for the patent of information industry listed company using information industry listed company as sample populations Index (172)；It is sample populations that Bio-pharmaceutical Industry listed company, which can also be directed to, excavates and is suitable for Bio-pharmaceutical Industry listing The patent leading indicators (172) of company；It can be more directed to chemical engineering of materials industry listed company, individually excavate and be suitable for chemical engineering of materials The patent leading indicators (172) of industry listed company.

The above-mentioned explanation done using the listed company of Shanghai exchange by patent entity (PE) and verification, financial index (FI) Using rate of return on equity ROE, however we are it has to be emphasized that rate of return on equity ROE explanations merely for convenience, The method of first preferred embodiment is applicable in various existing financial index (FI).

It is a kind of constructing method of the leading equation of patent referring to FIG. 9, for the second preferred embodiment of proposition of the invention (500), the leading equation of this patent (501) is predicting the financial information of patent entity (PE).The leading equation of patent (501) a leading score of patent (502), the finance letter of the leading patent entity (PE) of the leading score of patent (502) can be generated There are one the preset time leading phases (L) for breath.The step of constructing method (500) of the leading equation of this patent includes：

Step 510：The constructing method (100) of patent leading indicators according to the first preferred embodiment obtains second Panel data (151) and multiple patent leading indicators (172).

Step 520：Based on multiple patent leading indicators (172), third panel is formed from second panel data (151) screening Data (521).

Step 530：The second time series operation program (531) based on the time leading phase (L), the second time series are provided The independent variable of operation program (531) is all patent leading indicators (172) of third panel data (521), the second time series The dependent variable of operation program (531) is the financial index (FI) of third panel data (521).

Step 540：Second threshold (541) is set, the second time series operation program (531) and time leading phase are passed through (L), operation third panel data (521) obtain from screening in multiple patent leading indicators (172) and meet second threshold (541) Multiple patent core index (542) and generate the leading equation of patent (501), the leading equation of patent (501) substantially by Multiple patent core index (542) and its corresponding weight coefficient (543) are formed.

The main purpose of second preferred embodiment is to be directed to have the case where multiple patent leading indicators (172), will be multiple special Sharp leading indicators (172) are combined as a leading equation of patent (501), thereby are used for predicting the finance performance of enterprise, will more It is quick, convenient.The leading equation of data value importation patent possessed by patent core index (542) patent entity (PE) (501), the leading score of patent (502) is generated.The higher person of the leading score of patent (502) represents patent entity (PE) and is led in the time In advance (L) later corresponding financial index (FI) numerical value it is higher；The lower person of the leading score of patent (502) represents patent reality Body (PE) the time leading phase (L) later corresponding financial index (FI) numerical value it is lower.Due to business finance index (FI) Height directly express the quality of its business performance, the higher person of numerical value of financial index (FI), performance better, more has There is investment value.Since the leading score of patent (502) represents patent entity (PE) in time leading phase (L) the corresponding wealth of institute later The numerical value of index of being engaged in (FI), therefore the high or low of the leading score of patent (502) is observed, just it can be picked out from patent entity (PE) Object with investment potential.

In step 530, the second time series operation program (531) is polynary Granger CaFpngerusality test model, from change Amount is the obtained multiple patent leading indicators (172) of the first preferred embodiment step 170, it is therefore an objective to by the leading finger of multiple patents Mark (172) is combined, and the leading effect to financial index (FI) after multiple patent leading indicators (172) are combined is excavated.

But at this point, we must be understood that another key concept, polynary Granger CaFpngerusality test model are not multiple one First Granger CaFpngerusality test model generates the simple totalling of result, a other when combining multiple patent leading indicators (172) Patent leading indicators (172) can change to the conspicuousness of the leading effect of financial index (FI), even in certain patents The leading effect of leading indicators (172) can become not significantly instead.Therefore in step 530, preferably, can be with further operating Dependent variable deletes program item by item.Also that is, all patent leading indicators (172) are first included in polynary Granger CaFpngerusality test model In dependent variable, it is worst or even do not have aobvious to delete conspicuousness the P_value of each patent leading indicators (172) after observation is examined The patent leading indicators (172) of work property, then reform polynary Granger CaFpngerusality test model, then observe each patent neck after inspection The P_value of first index (172), then to delete conspicuousness worst or even do not have the patent leading indicators (172) of conspicuousness, repeats this Process finally leaves the higher patent leading indicators (172) of significant property and referred to as patent core index (542).At this point, polynary lattice Blue outstanding person's Causality Test model can integrate all patent core index (542) and generate the leading equation of patent (501), and patent is leading Equation (501) is substantially made of multiple patent core index (542) and its corresponding weight coefficient (543).

We illustrate by taking leading 1 year shown in Fig. 5 patent leading indicators (172) as an example again, these patent leading indicators (172) all it is for predicting that the finance of patent entity (PE) show.Wherein there are the patent leading indicators (172) of part that can also be used for Unitary patent is assessed, as the intensity index of assessment unitary patent, such as：

P6：Disclosure of the invention patent is averaged the patent service life

P7：Utility model patent is averaged the patent service life

P9：Invention granted patent is averaged the patent service life

P10：Invention granted patent is averaged the period under review

P38：The exclusive rights average of current utility model patent

If we want the patents leading indicators such as leading 1 year P6, P7, P9, P10, P15, P38 (172) forming patent Leading equation (501), first, we by second panel data (151), 375 patent entities (PE) from 2008 extremely Financial index (FI) in 2012：ROE and 6 patent leading indicators (172)：P6, P7, P9, P10, P15, P38 are picked out, Form third panel data (521).

We use polynary Granger CaFpngerusality test model to now, set second threshold (541)：P_value<0.05, the The independent variable once analyzed is 6 patent leading indicators (172)：P6, P7, P9, P10, P15, P38, dependent variable ROE, analysis As a result such as Figure 10 A.It can therefrom find, the leading significant effect of each patent leading indicators (172) changes, and P15 becomes Obtain worst, P_value=0.7833.

When second of analysis, we reject P15 worst P_value, and independent variable uses P6, P7, P9, P10, P38 etc. 5 A patent leading indicators (172), analysis result such as Figure 10 B are.Wherein P9 is worst, P_value=0.4188.

When third time is analyzed, we reject P9 worst P_value, and independent variable is special using P6, P7, P10, P38 etc. 4 Sharp leading indicators (172), analysis result such as Figure 10 C, wherein each patent leading indicators (172) meet second threshold (541), P_value is both less than 0.05, and prediction model reaches 95% confidence interval.4 leading fingers of patent shown in Figure 10 C It marks (172), we are defined as patent core index (542) at this time.

The basic operations model of polynary Granger CaFpngerusality test model is Multivariable regressive analysis model, therefore above-mentioned third time Analysis generates a composite equation formula other than excavating patent core index (542), and we term it patent necks First equation (501), wherein

The leading equation of patent (501)=w6 × P6+w7 × P7+w10 × P10+w38 × P38

Wherein w6, w7, w10, w38 are corresponding respectively to patent core index (542)：The weight system of P6, P7, P10, P38 Number (543) represents the susceptibility of primacy prediction of its corresponding patent core index (542) to financial index (FI).This reality Example is applied by actual operation, w6=0.1236, w7=0.0236, w10=0.0596, w38=0.0247, patent leading side at this time The R square values of formula (501) reach 0.9065, and R square values reach 0.8750 after adjustment, are good models.In the present embodiment, The value highest of w6, almost the 5 of w38 times indicate P6 (disclosure of the invention patent is averaged the patent service life) to the pre- of financial index (FI) It surveys most sensitive.When (exclusive rights of current utility model patent are average by P6 (disclosure of the invention patent is averaged the patent service life) and P38 Number) numerical value all only change a unit the case where, P6 (disclosure of the invention patent is averaged the patent service life) is to financial index (FI) institute Caused by variation, be exactly P38 (the exclusive rights average of current utility model patent) to the change caused by financial index (FI) 5 times of momentum.

1 is please referred to Fig.1, is a kind of method of assessment enterprise investment potentiality for the third preferred embodiment of proposition of the invention (600), include the following steps：

Step 610：Collect the patent information (612) of multiple enterprises (611)；

Step 620：One leading equation of patent (501) obtained by aforementioned second preferred embodiment, patent are provided Leading equation (501) is substantially made of multiple patent core index (542) and its corresponding weight coefficient (543)；

Step 630：Patent information (612) based on each enterprise (611) calculates each corresponding patent core index of enterprise (542) data (631)；

Step 640：The data (631) of patent core index (542) based on each enterprise (611), pass through patent leading side The calculating of formula (501) and the leading score of patent (502) for generating each enterprise (611)；

Step 650：The leading score of patent (502) of each enterprise (611) is arranged by a collator (651) Sequence, ranking results (652) represent the investment potential of enterprise (611).

The above-mentioned higher person of the leading score of patent (502), represent enterprise (611) the time leading phase (L) later it is corresponding The numerical value of financial index (FI) is higher；The lower person of the leading score of patent (502), represent enterprise (611) the time leading phase (L) with Afterwards corresponding financial index (FI) numerical value it is lower.Since the numerical value height of enterprise (611) financial index (FI) is directly expressed The quality of its business performance, the higher person of numerical value of financial index (FI), performance better, more have investment value；Wealth The lower person of numerical value of index of being engaged in (FI), performance is poorer, does not more have investment value.Due to the leading score of patent (502) represent enterprise (611) the time leading phase (L) later corresponding financial index (FI) numerical value, therefore by enterprise (611) The height ranking of the possessed leading score of patent (502), just can be from wherein picking out the object with investment potential.

2 are please referred to Fig.1, is a kind of calculating of assessment enterprise investment potentiality for the 4th preferred embodiment of proposition of the invention Machine system (700), including patent information harvester (710), index computation device (720), the leading score computing device of patent (730) and score collator (740).

Wherein, patent information harvester (710) collects the patent information (612) of multiple enterprises.

Patent information (612) of the index computation device (720) based on each enterprise calculates and generates the corresponding patent of enterprise The data (631) of heart index (542).

The leading score computing device (730) of patent is led to according to the data (631) of the patent core index (542) of each enterprise The leading equation of patent (501) that aforementioned second preferred embodiment is obtained is crossed, leading point of the patent of each enterprise is calculated and generate Number (502), the leading equation of this patent (501) is substantially by multiple patent core index (542) and its corresponding weight coefficient (543) it is formed.

The leading score of multiple patents (502) is ranked up by score collator (740), and ranking results (652) represent The sequence of the investment potential of multiple enterprises.

It is described above, the special personage of correlative technology field should be appreciated that and implement.The foregoing is merely this simultaneously The preferred embodiment of invention is not limited to the interest field of the present invention.It is any to be completed based on disclosed content Equivalent change or modification, should be included in the covering scope of claims.

Claims

1. a kind of constructing method of the leading equation of patent, the leading equation of the patent is predicting the finance letter of patent entity Breath, the leading equation of the patent generate a leading score of patent, the finance letter of the leading patent entity of the leading score of the patent There are one preset time leading phase, the constructing methods of the leading equation of the patent to be characterized in that the following steps for breath：

(1) multiple patent entities are set and to describe the multiple patent indexes and financial index of the respectively patent entity, respectively should Patent index is obtained by the patent information operation of the respectively patent entity, the patent index include with it is sum and current calculating, It is special to describe the index of Chinese invention publication, the index of description Chinese invention granted patent, the Chinese utility model of description The index of profit, the index with the Chinese design patent of description, which is stock price indices, wherein current calculate refers to handle The data operation of patent index is limited in some specific time interval；

(2) collection period is set, which is made of time interval and time issue, which is not small In two integer；

(3) respectively the patent entity is in the collection period for collection, each corresponding patent index numerical value of time interval institute and finance Index value；

(4) multiple patent index numerical value of multiple patent entity and multiple financial index numerical value are formed into first panel number According to；

(5) the first panel data are formed to the second panel data of normal distribution and criterion score by conversion, In this be converted to Z score operation；

(6) it sets the time leading phase and the first time Sequence Operation Theory program based on the time leading phase is provided, time neck Include at least one time interval in advance, the independent variable of the first time Sequence Operation Theory program is the one of the second panel data A patent index, the dependent variable of the first time Sequence Operation Theory program are the financial index of the second panel data, wherein The first time Sequence Operation Theory program is unitary Granger CaFpngerusality test model；

(7) first threshold is set, the first time Sequence Operation Theory program and the time leading phase, operation second face are gradually used Plate data screen from multiple patent index and obtain at least one patent leading indicators for meeting the first threshold；

(8) multiple patent leading indicators are based on, third panel data is formed from the second panel data screening；

(9) provide the second time series operation program based on the time leading phase, the second time series operation program from Variable is all patent leading indicators of the third panel data, the dependent variable of the second time series operation program be this The financial index of three panel datas, wherein the second time series operation program are polynary Granger CaFpngerusality test model；

(10) second threshold is set, the second time series operation program and the time leading phase, the operation third panel are passed through Data screen from multiple patent leading indicators and obtain the multiple patent core index for meeting the second threshold and generate this specially The leading equation of profit, the leading equation of the patent are made of multiple patent core index and its corresponding weight coefficient.

2. the constructing method of the leading equation of patent according to claim 1, wherein multiple patent entity is that listing is public Department.

3. the constructing method of the leading equation of patent according to claim 1, the wherein time interval are made of following Group select：Week, the moon, season, half a year, Yi Jinian.

4. the constructing method of the leading equation of patent according to claim 1, wherein between the step (4) and (5), into one Step includes normal distribution-test program, examines any one the patent index numerical value and the financial index numerical value.

5. the constructing method of the leading equation of patent according to claim 1, the wherein first threshold are not more than 0.1.

6. a kind of method of assessment enterprise investment potentiality, it is characterised in that：

(1) patent information of multiple enterprises is collected；

(2) a leading equation of patent is provided, the leading equation of the patent is by the leading equation of patent described in claim 1 Constructing method obtained, the leading equation of the patent is made of multiple patent core index and its corresponding weight coefficient；

(3) patent information based on the respectively enterprise calculates the data of the respectively corresponding patent core index of the enterprise；

(4) data of the patent core index based on the respectively enterprise generate the respectively enterprise by the leading equation of the patent calculates The leading score of patent of industry；

(5) the leading score of multiple patent is ranked up by a collator, which represents multiple enterprise Investment potential.

7. a kind of computer system of assessment enterprise investment potentiality, including patent information harvester, index computation device, patent Leading score computing device and score collator, it is characterised in that：

The patent information harvester collects the patent information of multiple enterprises；

Patent information of the index computation device based on the respectively enterprise calculates the number for generating the corresponding patent core index of the enterprise According to；

The leading score computing device of the patent is leading by a patent according to the data of the patent core index of the respectively enterprise Equation calculates and generates the leading score of patent of the respectively enterprise, and the leading equation of the patent is by patent described in claim 1 The constructing method of leading equation is obtained, and the leading equation of the patent is by multiple patent core index and its corresponding weight system It is several to be formed；

The leading score of multiple patent is ranked up and generates ranking results by the score collator, which represents should The investment potential of multiple enterprises.