CN109146727A - Patent value assessment method and system based on AI - Google Patents
Patent value assessment method and system based on AI Download PDFInfo
- Publication number
- CN109146727A CN109146727A CN201810862832.2A CN201810862832A CN109146727A CN 109146727 A CN109146727 A CN 109146727A CN 201810862832 A CN201810862832 A CN 201810862832A CN 109146727 A CN109146727 A CN 109146727A
- Authority
- CN
- China
- Prior art keywords
- value
- learning
- transaction data
- initial assessment
- valve estimating
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 15
- 238000005516 engineering process Methods 0.000 claims description 22
- 230000008569 process Effects 0.000 claims description 13
- 230000008859 change Effects 0.000 claims description 8
- 238000011161 development Methods 0.000 claims description 8
- 230000001939 inductive effect Effects 0.000 claims description 8
- 238000013135 deep learning Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 abstract description 10
- 238000012545 processing Methods 0.000 abstract description 5
- 230000006870 function Effects 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 4
- 230000001186 cumulative effect Effects 0.000 description 4
- 238000013507 mapping Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 230000001419 dependent effect Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000003672 processing method Methods 0.000 description 3
- 238000013475 authorization Methods 0.000 description 2
- 210000004556 brain Anatomy 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 238000012552 review Methods 0.000 description 2
- 230000006641 stabilisation Effects 0.000 description 2
- 238000011105 stabilization Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000004883 computer application Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000012854 evaluation process Methods 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000008450 motivation Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/18—Legal services
- G06Q50/184—Intellectual property management
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Technology Law (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides patent value assessment methods and system based on AI, are related to processing data information technical field.A kind of patent value assessment method based on artificial intelligence includes the following steps: that setting has the patent valve estimating model of learning functionality；The data information for acquiring patent to be assessed is assessed based on patent valve estimating model above-mentioned, obtains the initial assessment value of the patent；The patent is acquired in the transaction data being listed in afterwards, the transaction data is fed back into aforementioned patent appraisal Model；The patent valve estimating model is based on transaction data and is adjusted to aforementioned initial assessment value, obtains the ideal value of the patent.Invention not only avoids the possible evaluation errors of manual type, also save the human cost of patent valve estimating, have taken into account the flexibility and accuracy of data processing.
Description
Technical field
The present invention relates to processing data information technical fields.
Background technique
It applies for a patent and after patented power, both can protect the invention achievement of oneself, scientific achievement is prevented to be lost, obtain
Monopoly profits make up Innovation Input, while being also beneficial to scientific and technological progress and economic development.It is rapid with countries in the world economy
Development, various industrial field also increases significantly the demand of patented technology, but in actual running, industrial circle is still difficult to obtain
Obtain its patented technology definitely needed.Such as the manufacturer for needing to introduce new-type patented technology, look for new-type special
Sharp technology is undoubtedly the operating cost for being difficult to overcome；For another example for general so-called technology importer, Chang Wufa is rapid
Patented technology needed for effectively obtaining also is a kind of serious damage for the development of overall national power.
In patent transaction, the importance for assessing the value of patent is self-evident.Currently, the value assessment of patent is still main
It is carried out using traditional manual type, common appraisal procedure such as cost-or-market method, Market valuation method, income present value method etc..Wherein
It is related to often relating to multiple evaluation indexes, workload of operation is big, can consume biggish human resources；Simultaneously as depending on
Manual operation is likely to occur subjective deviation in evaluation process, influences true estimated value.
Summary of the invention
It is an object of the invention to: overcome the deficiencies of the prior art and provide a kind of patent value assessment method based on AI
And system.The present invention carries out value assessment to patent by artificial intelligence (AI), and patent valve estimating model can be based on transaction
Data information is adjusted the initial assessment value of patent, to obtain the ideal value of the patent.The present invention not only avoids
Manual type possible evaluation error, also saves the human cost of patent valve estimating, has taken into account data processing
Flexibility and accuracy.
To realize above-mentioned target, the present invention provides following technical solutions.
A kind of patent value assessment method based on artificial intelligence includes the following steps: that setting has the special of learning functionality
Sharp appraisal Model；The data information for acquiring patent to be assessed is assessed based on patent valve estimating model above-mentioned, is obtained
The initial assessment of the patent is taken to be worth；The patent is acquired in the transaction data being listed in afterwards, by the transaction data
Feed back to aforementioned patent appraisal Model；The patent valve estimating model is based on transaction data to aforementioned initial assessment
Value is adjusted, and obtains the ideal value of the patent.
Further, patent valve estimating model includes according to the learning strategy that transaction data adjusts initial assessment value
One of rote learning, deduction calligraphy learning, analogical learning, inductive learning and deep learning or various ways.
Further, the transaction data includes the attention rate of user's residence time in patent auction process and patent
Grade；According to residence time length and/or attention rate grade, weighted value is set, initial assessment value is adjusted.
Further, the transaction data includes that history of the patent in auction process is changed hands number, according to changing hands
Weighted value is arranged in number, to adjust initial assessment value.
Further, in the appraisal Model, to the value of patent in terms of technology dimension and law dimension two
It is assessed；The index of the technology dimension includes advance, substitutability, fields development trend and/or enforcement difficulty
Grade；The index of the law dimension include Patent right requirement item number, dependent patent claims range, patent stability,
Evading property, protection, the infringement property sentenced and/or patent distribution.
Further, the data information with the patent with other patents to mark characteristic is acquired, other patents are obtained
Estimated value information after the initial assessment of patent value is adjusted.
The invention also discloses a kind of patent valve estimating systems based on artificial intelligence, comprise the following structure:
Initialization module the patent valve estimating model with learning functionality is arranged, and tests patent valve estimating
The accuracy of model after the accuracy meets accuracy threshold requirement, starts the patent valve estimating model；
Initial assessment module, connection initialization module, to acquire the data information of patent to be assessed, based on above-mentioned special
Sharp appraisal Model is assessed, and the initial assessment value of the patent is obtained；
Transaction Information acquisition module, to acquire the patent in the transaction data being listed in afterwards, by the number of deals
It is believed that breath feeds back to aforementioned patent appraisal Model；
Module is adjusted, initial assessment module and Transaction Information acquisition module are connected, to pass through according to transaction data
Patent valve estimating model is adjusted aforementioned initial assessment value, obtains the ideal value of the patent.
Further, the patent valve estimating model adjusts the learning strategy of initial assessment value according to transaction data
Including one of rote learning, deduction calligraphy learning, analogical learning, inductive learning and deep learning or various ways.
Further, the transaction data includes the attention rate of user's residence time in patent auction process and patent
Grade；Patent valve estimating model is arranged weighted value and is worth to initial assessment according to residence time length and/or attention rate grade
It is adjusted.
Further, the transaction data includes that history of the patent in auction process is changed hands number；Patent value
Assessment models are worth according to number setting weighted value is changed hands with adjusting initial assessment.
The present invention due to using the technology described above, compared with prior art, for example and without limitation, has below
Advantage and good effect: value assessment is carried out to patent by artificial intelligence (AI), provided with the patent valence with learning functionality
It is worth assessment models, can be adjusted based on initial assessment value of the transaction data to patent, to obtain the patent
Ideal value.Invention not only avoids the possible evaluation errors of manual type, also save the people of patent valve estimating
Power cost has taken into account the flexibility and accuracy of data processing.
Detailed description of the invention
Fig. 1 is the flow chart of the patent value assessment method provided in an embodiment of the present invention based on AI.
Fig. 2 is the learning strategy figure of patent valve estimating model provided in an embodiment of the present invention.
Fig. 3 is the indication information figure of patent valve estimating model provided in an embodiment of the present invention.
Fig. 4 is the examples of interfaces figure of patent auction system provided in an embodiment of the present invention.
Fig. 5 is the function structure chart of the patent valve estimating system provided in an embodiment of the present invention based on AI.
Figure label is as follows:
Transaction interface 100, title bar 110, toolbar 120, information display area 130, scrolling bar 140；
Patent valve estimating system 200, initialization module 210, initial assessment module 220, Transaction Information acquisition module
230, adjust module 240.
Specific embodiment
Below in conjunction with the drawings and specific embodiments to the patent value assessment method and system provided by the invention based on AI
It is described in further detail.It should be noted that the combination of technical characteristic or technical characteristic described in following embodiments is not
Should be considered as it is isolated, they can be combined with each other to reaching superior technique effect.In the attached of following embodiments
In figure, the identical label that each attached drawing occurs represents identical feature or component, can be apply to different embodiments.
It should be noted that this specification structure depicted in this specification institute accompanying drawings, ratio, size etc., only to cooperate explanation
The revealed content of book is not limited to invent enforceable restriction item so that those skilled in the art understands and reads
Part, the modification of any structure, the change of proportionate relationship or the adjustment of size are not influencing the effect of invention can be generated and institute's energy
Under the purpose reached, it should all fall in the range of the revealed technology contents of invention can cover.Preferred implementation side of the invention
The range of formula includes other realization, wherein sequence shown or discussed can not be pressed, including is pressed according to related function
Basic mode simultaneously or in the opposite order, to execute function, this should be by the skill of the embodiment of the present invention technical field
Art personnel are understood.
Embodiment
Shown in Figure 1, a kind of patent value assessment method based on artificial intelligence includes the following steps:
The patent valve estimating model with learning functionality is arranged in S100.
In the present embodiment, the patent valve estimating model adjusts the study that initial assessment is worth according to transaction data
Strategy may include one of rote learning, deduction calligraphy learning, analogical learning, inductive learning and deep learning or a variety of sides
Formula, it is shown in Figure 2.
The rote learning refers to that learner without any reasoning or other Knowledge conversions, directly draws environment and mentioned
The information of confession.This kind of learning system is directly by prior primary concern is that how to index the knowledge of storage and be used
The program finish, constructed learns, and learner does not make any work, or by directly receiving the set fact and data
Learnt, any reasoning is not made to input information.
The deductive learning refers to by reasoning, from axiom, derives conclusion by logical conversion.This study
Method may include macro operation study, knowledge edition and chunking technology.
The analogical learning refers to using the knowledge similitude in two different fields (source domain, aiming field), can pass through
Analogy derives the corresponding knowledge of aiming field from the knowledge (including similar feature and other properties) of source domain, to realize
It practises.Analogical learning system can be such that an existing computer application system is changed into be adapted to new field, original to complete
The similar function of not designing.
The inductive learning obtains the concept by induction after being to provide some examples or counter-example of certain concept
General description.This study needs depth reasoning, because environment does not provide general conceptual description (such as axiom).Inductive learning
Be it is most basic, development also more mature learning method is widely studied and is applied in artificial intelligence field.
The deep learning is the method for carrying out representative learning based on data.Its motivation is to establish, simulates human brain progress
The neural network of analytic learning, it imitates the mechanism of human brain to explain data, such as image, sound and text.Major technique ratio
If any adaptive learning algorithm, convolutional neural networks, Recognition with Recurrent Neural Network, recurrent neural network, deep neural network and depth
Stack network etc..Its application includes computer vision technique in the prior art, speech recognition technology, natural language processing skill
Art, machine translation etc..
S200 acquires the data information of patent to be assessed, is assessed based on patent valve estimating model above-mentioned, obtains
The initial assessment of the patent is worth.
The mode for obtaining the data information of patent to be assessed may come from the patent data file of user's active upload,
It is also possible to the associated patent database of user search, is obtained from the patent data file stored in the patent database to be evaluated
Estimate the data information of patent.
The patent database is that is pre-established store the database of patent data, patent data status attribute.Institute
Patent data status attribute is stated, may include patent name, patent type, legal information, applicant, master in the present embodiment
Classification number, transferable information, the reference information such as number information and abstract.Preferably, the patent type may include patent application,
Invent authorization, utility model, design；The legal information may include examine in, effectively, failure, transfer, review, in vain
Declaration.
The retrieval of user can be realized by retrieval module.Usually, the retrieval module may include keyword
Retrieval module, IPC code retrieval module and applicant's retrieval module.Preferably, the key search module includes title
Abstract retrieval, title retrieval, abstract retrieval, claim retrieval, title abstract claim retrieval and full-text search.
It after the data information for obtaining patent to be assessed, is assessed based on aforementioned patent appraisal Model, obtains this specially
The initial assessment value of benefit.
It is shown in Figure 3, it, can be in terms of technology dimension and law dimension two to special in the appraisal Model
Benefit or the value of patent packet are assessed.
The index of the technology dimension includes advance, substitutability, fields development trend and/or enforcement difficulty etc.
Grade.
The index of the law dimension includes Patent right requirement item number, dependent patent claims range, patent stabilization
Property, evading property, protection, the infringement property sentenced and/or patent distribution.
It should be noted that in the appraisal Model, it can also be from market dimension to the valence of patent or patent packet
Value is assessed.The market dimension index includes that product corresponding to patent is right in the economic scale of patent right country and country
The support dynamics of patent fields.
Using each index of the appraisal Model, the value of the patent or patent packet is assessed.As an example rather than
Limitation, the method for assessing the value of the patent by taking index " Patent right requirement item number " as an example can be such that
Mapping between default Patent right requirement item number and patent value grade, patent value grade and the magnitude of value is closed
System.By way of example and not limitation, for example the value grade of claim item number 2 or less (including 2) is 5 etc., claim
Value grade of the item number between 3 to 6 is 4 etc., and value grade of the claim item number between 7 to 10 is 3 etc.,
Value grade of the claim item number between 11 to 15 is 2 etc., claim item number 16 or more (including 16)
Being worth grade is 1 etc..And for aforementioned patent value class 5 etc., 4 etc., 3 etc., 2 etc., 1 etc., counter value amount is respectively 1
Ten thousand, 50,000,100,000,150,000,200,000.
The claim item number for obtaining patent to be assessed obtains its corresponding magnitude of value according to the claim item number, than
If the claim item number of a patent is 8, then its magnitude of value is 100,000.If using only " patent in appraisal Model
Evaluation index of claim item number ", then aforementioned 100,000 magnitude of value is the initial assessment value of the patent.
It, by way of example and not limitation, can be first to every patent in patent packet for the patent packet including multinomial patent
It is individually evaluated, carries out cumulative summation again after obtaining the estimated value of each part patent, the magnitude of value summation of acquisition is special as this
The estimated value of Li Bao.
For the appraisal Model including multiple evaluation indexes, weight can be set for multiple evaluation indexes, it will be each
After the corresponding magnitude of value of item index is multiplied by corresponding weight, then carry out cumulative summation, the estimated value as the patent.Certainly,
By way of example and not limitation, user can select to fit above-mentioned data processing method according to demand when appraisal Model is arranged
Close the data processing method of oneself needs, various data processing methods such as weighted sum, weighted average, square after sum etc. it is equal
It can refer to the prior art, details are not described herein.
S300 acquires the patent in the transaction data being listed in afterwards, the transaction data is fed back to aforementioned
Patent valve estimating model.
The transaction data refers to the information relevant to the patent that is listed acquired after patent is listed.As an example
Rather than it limits, for example may include user's access times, user's residence time in patent auction process, attention rate of patent etc.
Grade, patent are changed hands the information such as number, patent historical trading price.
S400, the patent valve estimating model are based on transaction data and are adjusted to aforementioned initial assessment value,
Obtain the ideal value of the patent.
In the present embodiment, the transaction data may include user's residence time and patent in patent auction process
Attention rate grade；According to residence time length and/or attention rate grade, weighted value is set, initial assessment value is adjusted
It is whole.
When weighted value is arranged, the residence time is longer, and the weighted value is bigger；And/or the attention rate grade is got over
Height, the weighted value are bigger.
The attention rate grade can obtain by the following method: the sight information of user is acquired by camera settings,
Obtain sight target patent of interest；Count preceding aim patent concern number obtained；Based on preset concern number
With the mapping relations between attention rate grade, obtains the attention rate grade of the patent and export.
By way of example and not limitation, by taking attention rate grade as an example, for example it is based on the collected concern number of certain target patent
It is 63 times, corresponding attention rate grade is 3 stars (five-star rating system).And according to attention rate grade, 1 star, 2 stars, 3 stars
Grade, 4 stars and the corresponding weighted value of 5 stars are followed successively by 1.03,1.06,1.10,1.15 and 1.20, then preceding aim patent is corresponding
Weighted value be 1.1.If the initial assessment value that patent valve estimating model generates is 100,000, according to transaction data tune
Ideal value after whole is 110,000.
In another implementation method of the present embodiment, the transaction data can also include the patent in auction process
History change hands number, according to number setting weighted value of changing hands, to adjust initial assessment value.
By way of example and not limitation, for example, based on certain collected history of target patent change hands number (transaction count) be 8
It is secondary, it is directed to different numbers of changing hands, 0 time, 1-3 times, 4-6 times, 8-10 times, 10 times or more corresponding weighted value is followed successively by 1,
1.05,1.10,1.15 and 1.20, then the corresponding weighted value of preceding aim patent is 1.15.If patent valve estimating model generates
Initial assessment value be 100,000, then according to transaction data ideal value adjusted be 11.5 ten thousand.
In another implementation method of the present embodiment, the number with the patent with other patents to mark characteristic can also be acquired
It is believed that breath, is adjusted the initial assessment value of the patent after obtaining the estimated value information of other patents.
The other patents having to mark characteristic, refer in patent valve estimating, can be used as the patent to be assessed
The mark post patent of reference object.Preferably, have to mark characteristic patent be and patent technical field to be assessed, the row
Industry and the same or similar patent of subject name claimed.
By way of example and not limitation, for example patent to be assessed is related to a kind of technology that streaming media files are handled,
The subject name of protection is files in stream media compress technique, is generated based on the patent valve estimating model with learning functionality first
Beginning estimated value is 100,000.By searching for other patents relevant to the patent, a kind of discovery patent " files in stream media compression biography
Transmission method " is to have the target patent to mark characteristic with it, since the target patent has very high patent attention rate and changes hands
Number is more, and the current ideal value of the target patent (estimated value adjusted) is 200,000.It is then aforementioned that there is learning functionality
Patent valve estimating model, according to it is aforementioned to mark patent ideal value, after analogical learning strategy, to the patent just
Beginning estimated value is adjusted, and the ideal value for obtaining the patent is 160,000.
After obtaining the ideal value of patent, ideal value can be loaded into the corresponding column of patent auction window
Output.The ideal value-mono- such as aforementioned 16 ten thousand-one for obtaining patent is loaded into patent auction as the ideal value of the patent
It is exported in the corresponding column of window.
The patent auction window is the information input window of user oriented setting, while being also the transaction interface of user,
It can export relevant information based on preset output item index, shown in Figure 4.
By way of example and not limitation, the patent auction window can be using the transaction interface in Fig. 4.The transaction interface
100 include title bar 110, toolbar 120, information display area 130 and scrolling bar 140.The information of transaction interface output can be with
Listed code, subject name, technical field including patent, patent number, ideal value, the present bidding of patent, present bidding
With the valence ratio of ideal value, auction remaining number of days (listed residue number of days) information etc..
It is shown in Figure 5, it is another embodiment of the present invention, discloses a kind of patent valve estimating based on artificial intelligence
System.
The patent valve estimating system 200 includes the initialization module 210 of communication connection, and initial assessment module 220 is handed over
Easy information acquisition module 230 and adjustment module 240.
The initialization module 210 the patent valve estimating model with learning functionality is arranged, and tests patent valence
It is worth the accuracy of assessment models.After the accuracy meets accuracy threshold requirement, start the patent valve estimating model.It is described
Accuracy threshold value, by way of example and not limitation, such as 80%.When it is implemented, can be based on patent valve estimating model to
Patent through trading carries out value assessment, and acquisition estimated value is compared with the real trade of patent value, if the two
Difference 20% hereinafter, i.e. assessment accuracy be not less than 80%, then estimate the patent valve estimating model be valid model,
The starting patent valve estimating model can be used.
In the present embodiment, the patent valve estimating model adjusts the study that initial assessment is worth according to transaction data
Strategy may include one of rote learning, deduction calligraphy learning, analogical learning, inductive learning and deep learning or a variety of sides
Formula.
The initial assessment module 220, connection initialization module 210, to acquire the data information of patent to be assessed, base
It is assessed in patent valve estimating model above-mentioned, obtains the initial assessment value of the patent.
The mode for obtaining the data information of patent to be assessed may come from the patent data file of user's active upload,
It is also possible to the associated patent database of user search, is obtained from the patent data file stored in the patent database to be evaluated
Estimate the data information of patent.The patent database be pre-establish store patent data, patent data status attribute
Database.The patent data status attribute, in the present embodiment, may include patent name, patent type, legal information,
Applicant, Main classification number, transferable information, the reference information such as number information and abstract.Preferably, the patent type may include
Patent application, invention authorization, utility model, design；The legal information may include examine in, effectively, failure, transfer the possession of,
Review, invalid declaration.
It, can be in terms of technology dimension and law dimension two to patent or patent packet in the appraisal Model
Value is assessed.The index of the technology dimension includes advance, substitutability, fields development trend and/or implementation
Grade of difficulty.The index of the law dimension includes Patent right requirement item number, dependent patent claims range, patent stabilization
Property, evading property, protection, the infringement property sentenced and/or patent distribution.
Using each index of the appraisal Model, the value of the patent or patent packet is assessed.As an example rather than
Limitation, the method for assessing the value of the patent by taking index " Patent right requirement item number " as an example can be such that
Mapping between default Patent right requirement item number and patent value grade, patent value grade and the magnitude of value is closed
System.By way of example and not limitation, for example the value grade of claim item number 2 or less (including 2) is 5 etc., claim
Value grade of the item number between 3 to 6 is 4 etc., and value grade of the claim item number between 7 to 10 is 3 etc.,
Value grade of the claim item number between 11 to 15 is 2 etc., claim item number 16 or more (including 16)
Being worth grade is 1 etc..And for aforementioned patent value class 5 etc., 4 etc., 3 etc., 2 etc., 1 etc., counter value amount is respectively 1
Ten thousand, 50,000,100,000,150,000,200,000.
The claim item number for obtaining patent to be assessed obtains its corresponding magnitude of value according to the claim item number, than
If the claim item number of a patent is 8, then its magnitude of value is 100,000.If using only " patent in appraisal Model
Evaluation index of claim item number ", then aforementioned 100,000 magnitude of value is the initial assessment value of the patent.
It, by way of example and not limitation, can be first to every patent in patent packet for the patent packet including multinomial patent
It is individually evaluated, carries out cumulative summation again after obtaining the estimated value of each part patent, the magnitude of value summation of acquisition is special as this
The estimated value of Li Bao.
For the appraisal Model including multiple evaluation indexes, weight can be set for multiple evaluation indexes, it will be each
After the corresponding magnitude of value of item index is multiplied by corresponding weight, then carry out cumulative summation, the estimated value as the patent.
The Transaction Information acquisition module 230, to acquire the patent after listed in transaction data, will be described
Transaction data feeds back to aforementioned patent appraisal Model.The transaction data refers to and acquires after patent is listed
Information relevant to the patent that is listed.It by way of example and not limitation, for example may include user's access times, patent auction process
In user's residence time, patent attention rate grade, patent change hands the information such as number, patent historical trading price.
The adjustment module 240 connects initial assessment module 220 and Transaction Information acquisition module 230, to according to transaction
Data information is adjusted aforementioned initial assessment value by patent valve estimating model, obtains the ideal value of the patent.
In the present embodiment, the transaction data may include user's residence time and patent in patent auction process
Attention rate grade；According to residence time length and/or attention rate grade, weighted value is set, initial assessment value is adjusted
It is whole.When weighted value is arranged, the residence time is longer, and the weighted value is bigger；And/or the attention rate higher grade, institute
It is bigger to state weighted value.
The attention rate grade can obtain by the following method: the sight information of user is acquired by camera settings,
Obtain sight target patent of interest；Count preceding aim patent concern number obtained；Based on preset concern number
With the mapping relations between attention rate grade, obtains the attention rate grade of the patent and export.
By way of example and not limitation, by taking attention rate grade as an example, for example it is based on the collected concern number of certain target patent
It is 63 times, corresponding attention rate grade is 3 stars (five-star rating system).And according to attention rate grade, 1 star, 2 stars, 3 stars
Grade, 4 stars and the corresponding weighted value of 5 stars are followed successively by 1.03,1.06,1.10,1.15 and 1.20, then preceding aim patent is corresponding
Weighted value be 1.1.If the initial assessment value that patent valve estimating model generates is 100,000, according to transaction data tune
Ideal value after whole is 110,000.
In another implementation method of the present embodiment, the transaction data can also include the patent in auction process
History change hands number, according to number setting weighted value of changing hands, to adjust initial assessment value.
By way of example and not limitation, for example, based on certain collected history of target patent change hands number (transaction count) be 8
It is secondary, it is directed to different numbers of changing hands, 0 time, 1-3 times, 4-6 times, 8-10 times, 10 times or more corresponding weighted value is followed successively by 1,
1.05,1.10,1.15 and 1.20, then the corresponding weighted value of preceding aim patent is 1.15.If patent valve estimating model generates
Initial assessment value be 100,000, then according to transaction data ideal value adjusted be 11.5 ten thousand.
In another implementation method of the present embodiment, the number with the patent with other patents to mark characteristic can also be acquired
It is believed that breath, is adjusted the initial assessment value of the patent after obtaining the estimated value information of other patents.
The other patents having to mark characteristic, refer in patent valve estimating, can be used as the patent to be assessed
The mark post patent of reference object.Preferably, have to mark characteristic patent be and patent technical field to be assessed, the row
Industry and the same or similar patent of subject name claimed.
In the above description, although all components of the various aspects of present disclosure can be interpreted to be assembled or by
It is operatively connected as a unit or module, but present disclosure is not intended to for its own to be limited to these aspects.But
Within the scope of the protection of goal of present disclosure, each component can selectively and operatively be merged with arbitrary number.This
Each of a little components part itself can also be implemented as hardware, while various components can partly merge or selectively total
Body merges and is implemented as with the computer program for executing the program module of the function of hardware equivalents.It is this to construct
The code or code segment of program can easily be exported by those skilled in the art.This computer program can be stored in calculating
In machine readable medium, the various aspects to realize present disclosure can be run.Computer-readable medium may include magnetic note
Recording medium, optical record medium and carrier media.
In addition, as the term of " comprising ", " including " and " having " should default it is being interpreted as including property or open
, rather than exclusive or closure, unless it is explicitly defined as opposite meaning.All technologies, science and technology or its other party
The term in face all meets meaning understood by one of ordinary skill in the art, unless it is defined as opposite meaning.It is looked in dictionary
To public term should not be idealized very much or impractically explained very much under the background of the relevant technologies document, unless the disclosure
Content is clearly defined as like that.
Although describing the exemplary aspect of present disclosure for purposes of illustration, those skilled in the art should
, it is realized that foregoing description is only the description to present pre-ferred embodiments, not to any restriction of the scope of the invention, the present invention
The range of preferred embodiment include other realization, wherein can not by it is described go out or the sequence that discusses execute function.
Any change, the modification that the those of ordinary skill in field of the present invention does according to the disclosure above content, belong to claims
Protection scope.
Claims (10)
1. a kind of patent value assessment method based on artificial intelligence, it is characterised in that include the following steps:
Patent valve estimating model with learning functionality is set；
The data information for acquiring patent to be assessed is assessed based on patent valve estimating model above-mentioned, obtains the patent
Initial assessment value；
The patent is acquired in the transaction data being listed in afterwards, the transaction data is fed back into aforementioned patent value and is commented
Estimate model；
The patent valve estimating model is based on transaction data and is adjusted to aforementioned initial assessment value, obtains the patent
Ideal value.
2. according to the method described in claim 1, it is characterized by: patent valve estimating model is adjusted according to transaction data
The learning strategy of initial assessment value includes in rote learning, deduction calligraphy learning, analogical learning, inductive learning and deep learning
One or more modes.
3. according to the method described in claim 1, it is characterized by: the transaction data includes in patent auction process
The attention rate grade of user's residence time and patent；According to residence time length and/or attention rate grade, weighted value is set to first
Beginning estimated value is adjusted.
4. according to the method described in claim 1, it is characterized by: the transaction data includes the patent in auction process
In history change hands number, according to number setting weighted value of changing hands, to adjust initial assessment value.
5. according to the method described in claim 1, it is characterized by: in the appraisal Model, from technology dimension and method
Two aspects of rule dimension assess the value of patent；The index of the technology dimension includes advance, substitutability, affiliated
Field development trend and/or enforcement difficulty grade；The index of the law dimension includes Patent right requirement item number, patent independence
Scope of the claims, patent stability, evading property, protection, the infringement property sentenced and/or patent distribution.
6. according to the method described in claim 1, it is characterized by: acquisition has with the patent to other patents of mark characteristic
Data information is adjusted the initial assessment value of the patent after obtaining the estimated value information of other patents.
7. a kind of patent valve estimating system based on artificial intelligence, characterized by comprising:
Initialization module the patent valve estimating model with learning functionality is arranged, and tests patent valve estimating model
Accuracy, after the accuracy meets accuracy threshold requirement, start the patent valve estimating model；
Initial assessment module, connection initialization module are based on patent valence above-mentioned to acquire the data information of patent to be assessed
Value assessment models are assessed, and the initial assessment value of the patent is obtained；
Transaction Information acquisition module, to acquire the patent after listed in transaction data, by the number of deals it is believed that
Breath feeds back to aforementioned patent appraisal Model；
Module is adjusted, initial assessment module and Transaction Information acquisition module are connected, to pass through patent according to transaction data
Appraisal Model is adjusted aforementioned initial assessment value, obtains the ideal value of the patent.
8. system according to claim 1, it is characterised in that: patent valve estimating model is adjusted according to transaction data
The learning strategy of initial assessment value includes in rote learning, deduction calligraphy learning, analogical learning, inductive learning and deep learning
One or more modes.
9. system according to claim 1, it is characterised in that: the transaction data includes in patent auction process
The attention rate grade of user's residence time and patent；Patent valve estimating model is according to residence time length and/or attention rate etc.
Grade, setting weighted value are adjusted initial assessment value.
10. system according to claim 1, it is characterised in that: the transaction data includes that the patent was being auctioned
History in journey is changed hands number；Patent valve estimating model is worth according to number setting weighted value is changed hands with adjusting initial assessment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810862832.2A CN109146727A (en) | 2018-08-01 | 2018-08-01 | Patent value assessment method and system based on AI |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810862832.2A CN109146727A (en) | 2018-08-01 | 2018-08-01 | Patent value assessment method and system based on AI |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109146727A true CN109146727A (en) | 2019-01-04 |
Family
ID=64799486
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810862832.2A Pending CN109146727A (en) | 2018-08-01 | 2018-08-01 | Patent value assessment method and system based on AI |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109146727A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110046872A (en) * | 2019-04-23 | 2019-07-23 | 北京恒冠网络数据处理有限公司 | A kind of intellectual property financial value analysis and management system based on big data |
| CN113128621A (en) * | 2021-05-12 | 2021-07-16 | 北京大学 | Data resource value evaluation report generation method and device |
| CN116596563A (en) * | 2023-07-17 | 2023-08-15 | 北京大学 | Digital asset valuation system based on multi-factor model |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107563793A (en) * | 2017-08-07 | 2018-01-09 | 深圳益强信息科技有限公司 | The method and system that brand value based on big data is assessed |
| CN107590755A (en) * | 2017-08-07 | 2018-01-16 | 深圳益强信息科技有限公司 | Patent value assessment method based on big data |
| CN108022189A (en) * | 2016-11-03 | 2018-05-11 | 西安科技大市场创新云服务股份有限公司 | A kind of patent of invention value calculation method and apparatus |
| CN108053266A (en) * | 2017-12-29 | 2018-05-18 | 国信优易数据有限公司 | A kind of patent value predictor method and device |
-
2018
- 2018-08-01 CN CN201810862832.2A patent/CN109146727A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108022189A (en) * | 2016-11-03 | 2018-05-11 | 西安科技大市场创新云服务股份有限公司 | A kind of patent of invention value calculation method and apparatus |
| CN107563793A (en) * | 2017-08-07 | 2018-01-09 | 深圳益强信息科技有限公司 | The method and system that brand value based on big data is assessed |
| CN107590755A (en) * | 2017-08-07 | 2018-01-16 | 深圳益强信息科技有限公司 | Patent value assessment method based on big data |
| CN108053266A (en) * | 2017-12-29 | 2018-05-18 | 国信优易数据有限公司 | A kind of patent value predictor method and device |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110046872A (en) * | 2019-04-23 | 2019-07-23 | 北京恒冠网络数据处理有限公司 | A kind of intellectual property financial value analysis and management system based on big data |
| CN113128621A (en) * | 2021-05-12 | 2021-07-16 | 北京大学 | Data resource value evaluation report generation method and device |
| CN116596563A (en) * | 2023-07-17 | 2023-08-15 | 北京大学 | Digital asset valuation system based on multi-factor model |
| CN116596563B (en) * | 2023-07-17 | 2023-09-12 | 北京大学 | Digital asset valuation system based on multi-factor model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Cao et al. | Attsum: Joint learning of focusing and summarization with neural attention | |
| Feng et al. | Enhanced sentiment labeling and implicit aspect identification by integration of deep convolution neural network and sequential algorithm | |
| CN109146727A (en) | Patent value assessment method and system based on AI | |
| CN106021364A (en) | Method and device for establishing picture search correlation prediction model, and picture search method and device | |
| Sam et al. | Understanding adoption of big data analytics in China: from organizational users perspective | |
| Trujillo | Patterns in a complex system: an empirical study of valuation in business bankruptcy cases | |
| KR20130001552A (en) | Method for classifying document by using ontology and apparatus therefor | |
| CN110489646A (en) | User's portrait construction method and terminal device | |
| Li et al. | A two-stage decision model for information filtering | |
| Zhu et al. | A neural generative model for joint learning topics and topic-specific word embeddings | |
| Anil et al. | Performance analysis of deep learning architectures for recommendation systems | |
| Liu et al. | Few-shot short-text classification with language representations and centroid similarity | |
| Duan et al. | A Resume Recommendation Algorithm Based on K-means++ and Part-of-speech TF-IDF | |
| Wang et al. | An early warning system for oil security in China | |
| Voorspoels et al. | Idealness and similarity in goal-derived categories: A computational examination | |
| Urbanczik et al. | A gradient learning rule for the tempotron | |
| Rowe et al. | Aligning tweets with events: Automation via semantics | |
| Hu et al. | Cognitive computation on consumer’s decision making of internet financial products based on neural activity data | |
| Pereira | Entity linking with multiple knowledge bases: An ontology modularization approach | |
| Xie et al. | Dynamic weight-based individual similarity calculation for information searching in social computing | |
| Chen et al. | Dgrl: text classification with deep graph residual learning | |
| Zhao et al. | Efcnn: A restricted convolutional neural network for expert finding | |
| Tian et al. | Multi-attribute scientific documents retrieval and ranking model based on GBDT and LR | |
| Osman et al. | Application of fuzzy logic for adaptive food recommendation | |
| Diaz-Garcia et al. | A fuzzy-based approach for cyberbullying analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190104 |
|
| RJ01 | Rejection of invention patent application after publication |