CN109191293A - A kind of artificial intelligence service system and method based on intelligent contract and logical card - Google Patents
A kind of artificial intelligence service system and method based on intelligent contract and logical card Download PDFInfo
- Publication number
- CN109191293A CN109191293A CN201810901731.1A CN201810901731A CN109191293A CN 109191293 A CN109191293 A CN 109191293A CN 201810901731 A CN201810901731 A CN 201810901731A CN 109191293 A CN109191293 A CN 109191293A
- Authority
- CN
- China
- Prior art keywords
- rule
- expert
- consultation
- data
- artificial intelligence
- 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.)
- Granted
Links
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 61
- 238000000034 method Methods 0.000 title claims abstract description 45
- 230000007246 mechanism Effects 0.000 claims abstract description 10
- 238000012795 verification Methods 0.000 claims description 19
- 238000007689 inspection Methods 0.000 claims description 13
- 238000012360 testing method Methods 0.000 claims description 9
- 238000001514 detection method Methods 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 6
- 238000009825 accumulation Methods 0.000 claims description 4
- 235000013399 edible fruits Nutrition 0.000 claims description 3
- 230000009885 systemic effect Effects 0.000 claims description 3
- 239000000654 additive Substances 0.000 claims 1
- 230000000996 additive effect Effects 0.000 claims 1
- 238000004422 calculation algorithm Methods 0.000 abstract description 24
- 238000012423 maintenance Methods 0.000 abstract description 3
- 238000012549 training Methods 0.000 description 9
- 201000010099 disease Diseases 0.000 description 8
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 8
- 208000024891 symptom Diseases 0.000 description 8
- 238000012550 audit Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 238000013499 data model Methods 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 208000004998 Abdominal Pain Diseases 0.000 description 4
- 230000006399 behavior Effects 0.000 description 4
- 238000013480 data collection Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 4
- 210000004072 lung Anatomy 0.000 description 4
- 230000000505 pernicious effect Effects 0.000 description 4
- 206010063659 Aversion Diseases 0.000 description 3
- 206010011224 Cough Diseases 0.000 description 3
- 208000002193 Pain Diseases 0.000 description 3
- 206010037660 Pyrexia Diseases 0.000 description 3
- 208000037656 Respiratory Sounds Diseases 0.000 description 3
- 210000004727 amygdala Anatomy 0.000 description 3
- 230000002146 bilateral effect Effects 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 208000015181 infectious disease Diseases 0.000 description 3
- 230000003902 lesion Effects 0.000 description 3
- 208000020816 lung neoplasm Diseases 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 206010037833 rales Diseases 0.000 description 3
- 208000002881 Colic Diseases 0.000 description 2
- 102000001554 Hemoglobins Human genes 0.000 description 2
- 108010054147 Hemoglobins Proteins 0.000 description 2
- 206010021703 Indifference Diseases 0.000 description 2
- 206010035664 Pneumonia Diseases 0.000 description 2
- 239000002221 antipyretic Substances 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 210000003651 basophil Anatomy 0.000 description 2
- 210000001772 blood platelet Anatomy 0.000 description 2
- 230000036760 body temperature Effects 0.000 description 2
- 239000003795 chemical substances by application Substances 0.000 description 2
- 238000001816 cooling Methods 0.000 description 2
- 210000003979 eosinophil Anatomy 0.000 description 2
- 230000005284 excitation Effects 0.000 description 2
- 230000006698 induction Effects 0.000 description 2
- 230000002045 lasting effect Effects 0.000 description 2
- 210000004698 lymphocyte Anatomy 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000003058 natural language processing Methods 0.000 description 2
- 210000004493 neutrocyte Anatomy 0.000 description 2
- 238000003909 pattern recognition Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000029058 respiratory gaseous exchange Effects 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 206010067715 Gastrointestinal sounds abnormal Diseases 0.000 description 1
- 208000035211 Heart Murmurs Diseases 0.000 description 1
- 206010020751 Hypersensitivity Diseases 0.000 description 1
- 241000202934 Mycoplasma pneumoniae Species 0.000 description 1
- 206010028980 Neoplasm Diseases 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 206010057190 Respiratory tract infections Diseases 0.000 description 1
- 206010047700 Vomiting Diseases 0.000 description 1
- 230000007815 allergy Effects 0.000 description 1
- 230000001754 anti-pyretic effect Effects 0.000 description 1
- 230000003143 atherosclerotic effect Effects 0.000 description 1
- 238000002555 auscultation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 230000036772 blood pressure Effects 0.000 description 1
- 206010006451 bronchitis Diseases 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 206010012601 diabetes mellitus Diseases 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000035622 drinking Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 239000005337 ground glass Substances 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 230000013016 learning Effects 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 210000003800 pharynx Anatomy 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 210000002345 respiratory system Anatomy 0.000 description 1
- GOLXNESZZPUPJE-UHFFFAOYSA-N spiromesifen Chemical compound CC1=CC(C)=CC(C)=C1C(C(O1)=O)=C(OC(=O)CC(C)(C)C)C11CCCC1 GOLXNESZZPUPJE-UHFFFAOYSA-N 0.000 description 1
- 238000013526 transfer learning Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 238000002562 urinalysis Methods 0.000 description 1
- 230000008673 vomiting Effects 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
-
- 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
- G06Q20/00—Payment architectures, schemes or protocols
- G06Q20/08—Payment architectures
- G06Q20/14—Payment architectures specially adapted for billing systems
-
- 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
- G06Q30/00—Commerce
- G06Q30/018—Certifying business or products
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Technology Law (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of based on intelligent contract and the logical artificial intelligence method of servicing demonstrate,proved and system.The artificial intelligent Service method includes the following steps: to decide through consultation expert into that rule or data-driven model upload to assessment system, and is verified;The expert being verified is decided through consultation that rule or the corresponding assessment system of data-driven model bind intelligent contract, and block chain is written and is issued to artificial intelligence service system;When input data calls artificial intelligence service to be predicted, the expert being applicable in artificial intelligence service system invocation pattern library decides through consultation that rule or data-driven model obtain prediction result;After user selects prediction result, logical card is paid to artificial intelligence service system according to intelligent contract.Block chain technical application to professional domain is realized the shared of expert data and expertise, logical card mechanism is effectively guaranteed the legitimate rights and interests of industry specialists, industry data contributor, algorithm personnel and system maintenance personnel by this method.
Description
Technical field
The artificial intelligence service system based on intelligent contract and logical card (Token) that the present invention relates to a kind of, while being related to one
Artificial intelligence method of servicing of the kind based on intelligent contract and logical card, belongs to field of artificial intelligence.
Background technique
Currently, in every field such as medical, financial, unmanned, electric business, the service of AI driving is all flourishing.AI
(Artificial Intelligence, artificial intelligence) is generally divided into Liang great mainstream branch, and rule-based method and data are driven
Dynamic method.In the method for data-driven, and there is theoretical, space vector similarity the cluster scheduling theory branch with the Bayes of probability
The explanatory stronger classical statistics pattern-recognition of support, and (include based on General Neural Network with obvious black box subcharacter
The algorithm of the minimum energy function such as Boltzmann machine) development deep learning network such as CNN, LSTM etc..
Currently, many applications are often also to have merged several classes.For example, knowledge mapping is mainly shown as rule-based mould
Type, but the rule or state transition function of many knowledge mappings, combine interpretable probability and Clustering Model.
Either rule-based method, or interpretable probabilistic model (Bayesian Net etc.), or rely more on
In models such as deep learning, intensified learning and the transfer learnings of mass data driving discovery mode, closed loop in practice is required,
Constantly evolve.But it cannot apply and share in the artificial intelligence field of the professional domains such as medical treatment, data and knowledge at present.
A large amount of knowledge that some experts possess cannot be widely applied, and the localization of data and the privatization of knowledge cause differently
Field technique level is irregular.And expert also expends time energy without power and is divided the experience of many years accumulation on a large scale
It enjoys, because cannot return accordingly.
So it is a kind of not only can guarantee that knowledge and the safety of experience are shared and can guarantee contributor's acquisition it is corresponding return newly
AI energizes the method for professional domain service, becomes urgent need.
Token is proved as equity existing for a kind of digital form, can be generated and be circulated in the system of centralization, can also
To generate and circulate such as block chain (public chain or alliance's chain) in the system of decentralization.Token combination block chain, can be by adding
Close algorithm (true, anti-tamper, protection privacy) and distributed account book determine the uniqueness of the true and false and assets, and by being total to
Know algorithm to circulate.
Block chain technology is a kind of novel decentralization agreement, can the relevant data of secure storage, information can not forge
With distort, and have stronger land parcel change trace, the transaction and information data access certification on block chain are by all on block chain
Node is completed jointly, and the common recognition algorithm constantly brought forth new ideas guarantees its consistency.An account public (in chain) can be safeguarded on block chain
This, the data of all users on memory block chain, public account book is located at any node in memory block as it can be seen that anti-
Only data are distorted.But at present block chain technology AI energize professional domain service application seldom.
Summary of the invention
In view of the deficiencies of the prior art, primary technical problem to be solved by this invention is to provide a kind of based on intelligence conjunction
About and the artificial intelligence service system of logical card.
Another technical problem to be solved by this invention provides a kind of artificial intelligence service based on intelligent contract and logical card
Method.
For achieving the above object, the present invention uses following technical solutions:
According to a first aspect of the embodiments of the present invention, a kind of artificial intelligence service system based on intelligent contract and logical card is provided
System, including knowledge uploading module, knowledge verification module, release module, prediction module and payment module;
Wherein, the knowledge uploading module is used to deciding through consultation expert into that rule or data-driven model upload to assessment system;
The knowledge verification module is used for the expert quotient verify, and will be verified for transmitting knowledge uploading module
Set pattern is then or data-driven model is sent to the release module;
The release module is used to deciding through consultation the expert being verified into rule or the corresponding assessment system of data-driven model
Intelligent contract is bound, and block chain is written and is issued to artificial intelligence service system;
When user input data calls artificial intelligence service to be predicted, the prediction module is issued according to release module
Expert decide through consultation that rule or data-driven model are predicted, and prediction result is exported to user;
The payment module is according to the prediction result of the prediction module selected with user, according to intelligent contract from user account
It pays logical card to artificial intelligence service system, and Transaction Information publication is carried out by the release module.
Wherein more preferably, the artificial intelligence service system based on intelligent contract and logical card, further includes collision detection mould
Block;
The collision detection module is used to decide through consultation the expert of upload rule or data-driven model carry out conflict inspection, sentence
Whether disconnected be that new expert decides through consultation rule or new data driving model.
According to a second aspect of the embodiments of the present invention, a kind of artificial intelligence service side based on intelligent contract and logical card is provided
Method includes the following steps:
Expert is decided through consultation that rule or data-driven model upload to assessment system, and carries out assessment verifying;
The expert that assessment is verified is decided through consultation into rule or data-driven model as authenticated mode and is stored in mode
Library, and it is published to corresponding AI service system;Revenue Sharing Mechanism is bound into intelligent contract simultaneously, block chain is written;
When input data calls AI service to be predicted, the expert being applicable in AI service system invocation pattern library decides through consultation rule
Then or data-driven model obtains prediction result, and exports;
After user selects some rule or model and its corresponding prediction result, input, output and the choosing of entire data
The behavior record selected is counted into block chain, and Token is paid to each contributor of the mode according to intelligent contract.
Wherein more preferably, expert decides through consultation that rule or data-driven model upload to after assessment system, before being verified,
Further include following steps:
Rule or data-driven model, which carry out conflict inspection, is decided through consultation to the expert of upload, judges whether to be newly-increased effective special
Family decides through consultation rule or increases effective data-driven model newly.
Wherein more preferably, rule, which carries out conflict inspection, is decided through consultation to the expert of upload, judges whether it is to increase effective expert newly
It decides through consultation rule, includes the following steps:
S101 judges to decide through consultation that rule repeats or part is duplicate old with the presence or absence of the new expert with upload in AI service system
Expert decides through consultation rule, if it is present turning to step S102;Otherwise, step S103 is turned to;Wherein, it repeats or part repeatedly refers to
The restrictive condition of applicable input data is identical or partially overlaps;
S102 judges that new expert decides through consultation whether rule is updated compared with old expert decides through consultation rule, if it is,
Turn to step S103;Otherwise, refuse the upload that new expert decides through consultation rule;
S103 decides through consultation that rule carries out expert and assesses verifying to the new expert of upload.
Wherein more preferably, judge that expert decides through consultation whether rule updates and decides through consultation whether the input of rule is added newly including new expert
Data object, new expert decide through consultation whether rule further segments type to the legacy data object of input, new expert decides through consultation rule
Input object whether increase attribute, new expert decides through consultation whether rule modifies attribute, new expert decides through consultation rule whether by old expert
Decide through consultation rule carry out further division, the one or more that change of some condition or parameter in rule.
Wherein more preferably, the expert of described pair of upload decides through consultation that rule carries out assessment verifying, includes the following steps:
According to the expert of upload decide through consultation rule belonging to field determine the Committee of Experts member that is audited;
Verification threshold is set, when the expert that verifying uploads decides through consultation that the Committee of Experts membership of rules compliance is greater than verifying
When threshold value, the expert of upload decides through consultation rules compliance, and assessment is verified.
Wherein more preferably, the data-driven model uploaded to algorithm team carries out conflict inspection, judges whether to be newly-increased number
According to driving model, include the following steps:
S111, by original data-driven model and new data driving model simultaneously to the labeled data of systems stay accumulation
Collection carries out Diseases diagnosis, and marks the consistency of new data-driven model and every Geju City model one by one;
S112 then refuses new data when there are an existing old model, output and completely the same new model output
The publication of driving model;Otherwise, then step S113 is turned to;
S113 carries out assessment verifying to the new data driving model of upload.
Wherein more preferably, the new data driving model includes carrying out model based on old labeled data collection existing in system
The new data driving model that optimization training is obtained new data driving model and obtained based on the labeled data training increased newly in system
One kind or two kinds of combination.
Wherein more preferably, to the new data driving model of upload, assessment verifying is carried out, is included the following steps:
To misrepresent deliberately, rate of failing to report multiplied by corresponding cost function obtains final expected shortfall, if new data driving model
Final expected shortfall be less than original data-driven model final expected shortfall, and lower than systemic presupposition model it is pre-
Loss function threshold value is surveyed, then assessment is verified.
Wherein more preferably, assessing the newly-increased expert being verified decides through consultation rule or data-driven model as authenticated mould
Formula is stored in pattern base, and is published to corresponding AI service system;Revenue Sharing Mechanism is bound into intelligent contract, write area simultaneously
Block chain.
Wherein more preferably, it when input data calls AI service to be predicted, is applicable in AI service system invocation pattern library
Expert decide through consultation that rule or data-driven model obtain prediction result, and exported, included the following steps:
When input data calls AI service to be predicted, AI service system finds corresponding expert according to the input data
Decide through consultation rule and data-driven model;
It decides through consultation that rule and data-driven model judge the labeled data of input according to the expert found, is predicted
As a result;
The mode (expert decides through consultation rule and data-driven model) and prediction result of selection are subjected to output displaying.
Wherein more preferably, when the prediction result of selection, which corresponds to multiple experts, decides through consultation rule or data-driven model, AI
The expert for choosing minimum transaction value is decided through consultation that rule or data-driven model match by service system.
Wherein more preferably, the artificial intelligence method of servicing, further includes following steps:
When determining prediction result when the error occurs, fed back to AI service system.Block chain will be written in AI service system,
And it is written to mark database.
Wherein more preferably, when the number of user feedback system prediction result mistake is more than specific quantity threshold value, to these
Prediction result data set is verified, and judges that the corresponding expert decides through consultation rule or data-driven model is anti-in these users
Mistake or user itself misjudgment occur on the data example of feedback.
Wherein more preferably, when user feedback system prediction result mistake is correct feedback, based on intelligent contract, to the use
Give Token reward in family.
Wherein more preferably, the artificial intelligence method of servicing further includes following steps:
Decide through consultation that Token is distributed to different types of contributor by rule according to intelligent contract and the expert pre-established,
And it is recorded into the accounting nodes of block chain.
Artificial intelligence service system and method provided by the present invention realize block chain technical application to professional domain
Expert data and expertise it is shared, logical card mechanism is effectively guaranteed industry specialists, industry data contributor, algorithm personnel
With the legitimate rights and interests of system maintenance personnel.
Detailed description of the invention
Fig. 1 is the structural schematic diagram of the artificial intelligence service system provided by the present invention based on intelligent contract and logical card;
Fig. 2 is the flow chart of the artificial intelligence method of servicing provided by the present invention based on intelligent contract and logical card.
Specific embodiment
Detailed specific description is carried out to technology contents of the invention in the following with reference to the drawings and specific embodiments.
The artificial intelligence service system based on intelligent contract and logical card that present invention firstly provides a kind of.As shown in Figure 1, should
System includes at least knowledge uploading module, knowledge verification module, release module, prediction module and payment module.
Wherein, knowledge uploading module is used to decide through consultation expert that rule or data-driven model (knowledge) to upload to assessment system
System;Expert decides through consultation that rule is that the expert in candidate rule library is decided through consultation that rule uploads to assessment system and comments for the Committee of Experts by expert
Estimate, and is verified;AI service system can be just published to after being verified;Data-driven model is that algorithm team services according to AI
The data that System History operation data and expert upload carry out algorithm modeling, data-driven model are obtained, by data-driven mould
Type uploads to assessment system and carries out assessment verifying, and AI service system is published to after being verified.Detailed verification process is subsequent
It is described in detail.
The expert being verified is decided through consultation rule for verify by what knowledge uploading module transmitted by knowledge verification module
Then or data-driven model is sent to release module;Release module is used to deciding through consultation the expert being verified into rule or data-driven
The corresponding assessment system of model binds intelligent contract, and block chain is written and issues to AI service system.
When user input data calls AI service to be predicted, prediction module is decided through consultation according to the expert that release module is issued
Rule or data-driven model are predicted, and prediction result is exported to user;Payment module is pre- according to selecting with user
The prediction result for surveying module, pays Token to AI service system from user account according to intelligent contract;And Transaction Information is sent out
It gives release module and carries out Transaction Information publication.
Wherein, in embodiment provided by the present invention, the artificial intelligence service system based on intelligent contract and logical card is also
Including collision detection module.When establishing AI service system, need expert that expert is decided through consultation rule or data-driven mould
Type uploads to the assessment system of block chain.In embodiment provided by the present invention, expert is decided through consultation rule by knowledge uploading module
Or data-driven model uploads to after assessment system, knowledge verification module decides through consultation rule or data-driven model to expert is uploaded
Before being verified, it is also necessary to which collision detection module decides through consultation that rule or data-driven model carry out conflict inspection to the expert of upload
It tests, judges whether it is that new expert decides through consultation rule or new data driving model.Guarantee that the expert decides through consultation rule or data-driven mould
Type decides through consultation that rule or data-driven model do not conflict with existing expert in AI service system.Phase is not present i.e. in AI service system
Same expert decides through consultation rule or data-driven model.
As shown in Fig. 2, the artificial intelligence method of servicing provided by the present invention based on intelligent contract and logical card, including it is as follows
Step: it firstly, expert is decided through consultation that rule or data-driven model upload to assessment system, and is verified;By what is be verified
Expert decides through consultation that rule or the corresponding assessment system of data-driven model bind intelligent contract, and block chain is written to AI service system
Publication;Then, when user input data calls AI service to be predicted, the expert in AI service system invocation pattern library is decided through consultation
Rule or data-driven model obtain prediction result, export to user;After user selects prediction result, according to intelligent contract
Pay Token to AI service system;Finally, block chain is written in Transaction Information, and block chain is sent out to AI service system.Under
Detailed specific description is carried out in face of this process.
Expert is decided through consultation that rule or data-driven model upload to assessment system, and is verified by S1.
When establishing AI service system, need expert that expert is decided through consultation that rule or data-driven model upload to area
The assessment system of block chain.In embodiment provided by the present invention, expert decides through consultation that rule or data-driven model upload to assessment
Further include following steps before being verified after system:
Rule or data-driven model, which carry out conflict inspection, is decided through consultation to the expert of upload, judges whether it is that new expert decides through consultation rule
Then or new data driving model.
In embodiment provided by the present invention, expert decides through consultation that rule or data-driven model upload to expert and decide through consultation rule
After assessment system, before expert is decided through consultation that rule or data-driven model are verified, rule can be decided through consultation to the expert of loading
Or data-driven model carries out conflict inspection, guarantee the expert decide through consultation rule or data-driven model with it is existing in AI service system
Expert decide through consultation rule or data-driven model do not conflict.I.e. in AI service system there is no identical expert decide through consultation rule or
Data-driven model.
Wherein, rule, which carries out conflict inspection, is decided through consultation to the expert of upload, judges whether it is that new expert decides through consultation rule, it is specific to wrap
Include following steps:
S101 judges to decide through consultation that rule repeats or part is duplicate old with the presence or absence of the new expert with upload in AI service system
Expert decides through consultation rule, if it is present turning to step S102;Otherwise, step S103 is turned to;Wherein, it repeats or part repeatedly refers to
The restrictive condition of applicable input data is identical or partially overlaps.
S102 judges that new expert decides through consultation whether rule is updated compared with old expert decides through consultation rule, if it is,
Turn to step S103;Otherwise, refuse the upload that new expert decides through consultation rule.
Wherein, in embodiment provided by the present invention, compared with old expert decides through consultation rule, judge that new expert decides through consultation rule
Whether updated, including new expert decides through consultation that rule addition new object, new expert decide through consultation that rule further segments old object
Type, new expert decide through consultation that rule objects increase attribute, new expert decides through consultation that rules modification attribute, new expert decide through consultation rule by old expert
Decide through consultation rule carry out further division, the one or more that change of some condition or parameter in rule.
Specifically, judging that new expert decides through consultation whether rule is updated compared with old expert decides through consultation rule, comprising:
A) it adds new object or type or object increase attribute or modification attribute is further segmented to old object.Such as: base
Because of the invention of detection technique, this kind of new object of various description types in addition to increasing gene is done also by many tumor diseases
Thinner parting.
CT examination, the tubercle of discovery, the further parting of meeting are solid nodules, ground glass shadow, half solid nodules.Further,
The attribute of tubercle includes various image group features, such as HU average value, standard deviation, minimum HU value, highest HU value, kurtosis, side
Edge clarity, burrs on edges degree etc..
Diabetes: several class patients with type Ⅰ DM can be divided into, and can further be segmented.
B) it adds or corrects new expert and decide through consultation rule.It corrects new expert and decides through consultation rule, generally decide through consultation old expert
Regular further subdivision, allows it to be more precisely suitable for narrower situation.
For example, meeting patients' description, symptom combination description (A1, A2 ... ... A8);Sign (B1, B2, B3) description
State, originally respectively with probability P 1, P2, P3 corresponds to disease D1, D2, D3.It is replaced by more accurately (A1, A2 ... ... later
A8);Sign (B1, B2, B3) and biochemical physical examination item (C1, C2, C3, C4), correspond to disease D1;And (A1, A2 ... ... A8), body
(B1, B2, B3, B5) and biochemical physical examination item (C1, C2, C3, C4) are levied, disease D2 is corresponded to.It is illustrated below described:
Infant, female, 11 years old, Han nationality, the permanent village XX, the past constitution was fair, no repeated respiratory infections history, no allergies,
Free from infection contact history.
Reasoning based on old rule: symptom (fever, aversion to cold, cough);(39 DEG C of body temperature, breathe 32 beats/min, pharynx to sign
Portion is slightly red, and bilateral amygdala is without obvious enlargement) --- " infection of the upper respiratory tract, acute bronchitis, Eaton agent pneumonia ...
Rule is decided through consultation by adding or correcting new expert, after old expert is decided through consultation regular further subdivision, increases sign-
The various indexs of dry and wet rale and blood routine, routine urinalysis in biochemical investigation inspection, are more precisely suitable for the case where more refining.
Based on one of the Induction matrix for updating rule: symptom (fever, aversion to cold, cough: relatively light);Sign (39 DEG C of body temperature-
Effectively, 32 beats/min of breathing-is not heard and dry and wet rale, and pharyngeal slightly red, bilateral amygdala is without obvious enlargement for antipyretic cooling),
Test rating (leucocyte 4.4x109/L, hemoglobin 128g/L, total number of blood platelet 191x109/L, neutrophil leucocyte percentage
63%, lymphocyte percentage 26.6%, eosinophils ratio 1.10%, basophil ratio 0.0%) --- " on exhale
Inhale road infection.
Based on one of the Induction matrix for updating rule: symptom (fever: interruption -1 week, aversion to cold, cough: paroxysmal-stimulation
Property);(cooling of 39 DEG C-antipyretic of body temperature breathes the attenuating of 32 beats/min-right lung breath sound, left lung without substantially changeing to sign
Breath sound is clear, does not hear and dry and wet rale, and pharyngeal slightly red, bilateral amygdala is without obvious enlargement), test rating (leucocyte
4.4x109/L, hemoglobin 128g/L, total number of blood platelet 191x109/L, neutrophil leucocyte percentage 63%, lymphocyte hundred
Score 26.6%, eosinophils ratio 1.10%, basophil ratio 0.0%, mycoplasma pneumoniae IgM antibody
(+)) --- " Eaton agent pneumonia
Illustrate:
Patients' description: age (section), gender, nationality, permanent residence, morbidity ground, medical history etc..
Single symptom describes example: abdominal pain-position: pain around the navel;Character: colic pain;The frequency: paroxysm type;Inducement: after drinking.
Symptom combination is the combination of above-mentioned symptom description single in detail, such as (abdominal pain: navel week-paroxysm-colic pain;Vomiting: atherosclerotic
Object-after meal-night) be two symptoms combination.
Sign includes temperature pulse respiration breathing, blood pressure, also includes that auscultation obtains various sound sign (heart murmurs, borborygmus
Sound etc.), touch the reaction etc. of pressing.
S103 decides through consultation that rule carries out expert's verifying to the new expert of upload.Wherein more preferably, as Committee of Experts member couple
When expert decides through consultation that rule carries out audit modification, by deciding through consultation that rule is modified to the expert of upload, establishes new expert and decide through consultation
Sub-branch's rule, sub-branch's rule establishes the special of branch's sub-rule by becoming the new rule being verified after verifying
Family is the creator of the new assessment system.
After deciding through consultation that rule or data-driven model carry out conflict inspection to the expert of upload, expert is decided through consultation into rule or number
It is verified according to driving model.Wherein, the expert of upload decides through consultation that rule can pass through NLP (Natural Language for machine
Processing, natural language processing) parsing profession teaches book or various guide experts decide through consultation regular foundation, and it can also be expert
It is established based on knowledge and experience.After expert decides through consultation that rule uploads, regular institute need to be decided through consultation according to expert through AI service system
The Committee of Experts member in the system that the field of category is assigned automatically carries out audit amendment, after being verified, can bind area
Then the intelligent contract of block chain setting is issued.In embodiment provided by the present invention, the Committee of Experts in system at
Member decides through consultation that rule is verified to the expert of upload, specifically comprises the following steps:
According to the expert of upload decide through consultation rule belonging to field determine the Committee of Experts member that is audited;In the present invention
In provided embodiment, the expert of dosis refracta inside industry is needed to audit jointly.
Verification threshold is set, in the Committee of Experts member that the expert that verifying uploads decides through consultation rules compliance, is approved after examination
When the membership of the rule is greater than verification threshold, the expert of upload decides through consultation rules compliance;
Otherwise, refuse the expert and decide through consultation that rule is issued to AI service system.
In embodiment provided by the present invention, when the expert that Committee of Experts member uploads expert decides through consultation regular progress
When audit, by deciding through consultation that rule is modified established sub-branch rule to the expert of upload, then the sub- expert quotient of the branch of the establishment
Set pattern is then equal to newly-built expert and decides through consultation rule, also needs other Committee of Experts members and verifies.What expert decided through consultation
The verification process of branch's sub-rule decides through consultation that the verification process of rule is identical with the expert of above-mentioned upload, just repeats no more herein.
After verifying, which possesses new assessment system, and establishing and establishing the expert of branch's sub-rule is that this is new
Regular owner.
Wherein, conflict inspection is carried out to the data-driven model of upload, judges whether to be new data driving model, it is specific to wrap
Include following steps:
S111, by original data-driven model and new data driving model simultaneously to the labeled data of systems stay accumulation
Collection carries out Diseases diagnosis, obtains misrepresenting deliberately rate of failing to report;
New data driving model, with newly-increased data and different.In general, (effective) newly-increased data, i.e., newly-increased
The data of some model predictions not pair, and increased some mark attributes (label) newly, new data driving model can be brought.But
Even if not increasing data newly, there is also the possibility of the further model optimization of algorithm personnel.This newly-increased more preferably model also needs
It is reviewed receiving.
In embodiment provided by the present invention, new data driving model can be for based on old mark existing in system
Data set carries out model optimization, and training obtains new data driving model;Or it based on data are increased in system newly, trains and obtains
The new data driving model obtained.
In embodiment provided by the present invention, newly-increased data include newly-increased mark, the practical hair to data with existing sample
It is raw to predict one of the data not being consistent with original data-driven model (the existing data-driven model of AI service system)
Or two kinds.
Wherein, to the newly-increased mark of data with existing sample the case where, for example there are: the Lung neoplasm mark of a large amount of chest lung CT, such as
Fruit is further labelled with good pernicious grade and tubercle type, then belongs to the newly-increased mark of data with existing sample.
And in newly-increased data, it actually occurs and predicts the data not being consistent with legacy data driving model.For example there are: a large amount of chest lungs
The Lung neoplasm of CT marks, and after model foundation, can automatically process to input picture, identifies tubercle, and predicts the good pernicious etc. of tubercle
Grade and tubercle type.Intersected the different of the good pernicious judgement of tubercle and model prediction for judging to build consensus by doctor if encountered
It causes, then these mark sample needs are especially paid close attention to, and are further analyzed as sample, are improved with providing the evolution of model.
When based on the new data driving model for increasing data training acquisition in system newly, AI algorithm personnel are according to comprising newly-increased
The sample data set training new data driving model of the mark of data;Wherein, AI algorithm personnel are according to the mark comprising increasing data newly
Note sample data set training new data driving model process, with it is subsequent data-driven model is uploaded into assessment system when, into
The process of row verifying is similar, i.e., AI algorithm personnel establish various models, according to the sample data set of the mark comprising increasing data newly
Carry out closed test;Setting model accuracy threshold value is obtained when the test accuracy of the model of foundation reaches model accuracy threshold value
The data-driven model of update, behind be described in detail again, just do not repeating herein.
S112, processor judge whether there is an existing old model, and output exports completely the same with new model.Such as
Fruit has, then refuses the publication of new data driving model.If there is no such model, then step S113 is turned to.
S113, processor carry out assessment verifying to the new data driving model of upload.AI service system forbids plagiarizing expert
Decide through consultation rule and data-driven model.All experts for being proved to output and input result indifference decide through consultation rule or data-driven
Model, will according to arrive first first well-behaved, the subsequent founder of automatic shield.But same expert decides through consultation rule, data-driven model,
If new data driving model realizes speed faster, system is automatically based upon the expert being divided into mechanism and decides through consultation rule, gives more
New person is certain to be divided into, and is paid by Token.
In embodiment provided by the present invention, expert decides through consultation that regular founder or data-driven model founder enjoy area
" class patent " treatment in block chain is forbidden any plagiarism within a certain period of time and is copied.It is after the period, then any new
It builds expert and decides through consultation rule or data-driven model, can copy and run at a low price.
It when data-driven model is uploaded to assessment system, is verified, is included the following steps:
The misrepresenting deliberately of the data-driven model, rate of failing to report are obtained the final loss phase multiplied by corresponding cost function by assessment system
Prestige value, if the final expected shortfall of new data driving model is less than the final loss expectation of original data-driven model
Value, and be lower than the model prediction loss function threshold value of systemic presupposition, then it is deposited new data driving model as authenticated mode
It is stored in pattern base, and is published to corresponding AI service system;Revenue Sharing Mechanism is bound into intelligent contract simultaneously, block is written
Chain.
Data-driven model is misrepresented deliberately, is failed to report based on the sample data set in mark database.
In general, the sample data set that mark is completed is uploaded to assessment system by industry specialists;If necessary to construct number
According to driving model, a large amount of historical data is needed to be analyzed and processed.It needs expert to be first labeled to training data, that is, obtains
The mark sample data set that a collection of industry specialists are approved.
AI algorithm personnel are based on labeled data, establish model and adjust ginseng, are sealed according to the sample data set that mark is completed
Close test;Wherein, the model that AI algorithm personnel establish can be theoretical, space vector similarity the cluster with the Bayes of probability
The model of the explanatory stronger classical statistics pattern-recognition building of scheduling theory support, is also possible to special with obvious flight data recorder
The model of the buildings such as the deep learning network based on General Neural Network development of sign.The method for constructing model is not done any
Limitation.
Such as: the CT medical image sample of expert's mark, i.e. doctor in the picture, mark lesion (such as Lung neoplasm) in three-dimensional
The lesion type of spatial position and the position in CT image, lesion attribute (good pernicious grade etc.).AI algorithm personnel are according to mark
The various models of CT medical image sample training being poured in, when the test accuracy of the model of foundation reaches model accuracy threshold value,
Pass through verifying.
Mode-the expert being verified is decided through consultation that rule or data-driven model-are stored in pattern base, and is published to by S2
Corresponding AI service system;Revenue Sharing Mechanism is bound into intelligent contract simultaneously, block chain is written.
In embodiment provided by the present invention, the transaction value of intelligent contract is divided into according to the contribution to AI service system
Degree determines.Expert decides through consultation creator's (containing regenerator) of rule, becomes the owner of assessment system, will possess the expert and decide through consultation
The virtual dividend power of rule.I.e. when the expert decides through consultation that rule is called, AI service system need to pay expert and decide through consultation rule
Creator's certain proportion or certain amount of Token.Other Token, can be used as the assessment system should pay it is whole
The tax revenue of a AI service system, the O&M of audit, amendment and AI service system for compensating the rule are promoted.
Decide through consultation that the price of rule invocation, AI service system can give expert's virtual dividend for deciding through consultation rule about each expert
Weigh one market guiding price of owner.Specifically how to fix a price, decides through consultation that regular owner determines by expert.Market guiding price, is based on
The expert decides through consultation that rule other similar experts corresponding in AI service system decide through consultation the calling price of rule, services also based on AI
The corresponding cost of labor for carrying out corresponding expert and deciding through consultation rule judgement outside system.
And the foundation about data-driven model, need exerting jointly for initial labeled data collection and both AI algorithm personnel
Power.Therefore, the creator of a data-driven model is two groups of personnel.One side is contribution data person collective, occupies certain part
Volume.The personal virtual of the share, which is shared out bonus, to be weighed, and by its contribution proportion, (i.e. everyone is contributed and weighted contributions value of audit data accounts for entirety
Mark the ratio of the weighted contributions total value of sample data).Another party is that the algorithm of algorithm personnel is contributed.Data-driven model calls
Price, decides through consultation that rule invocation price is similar to expert, is not just repeating herein.
S3 is applicable in special when user input data calls AI service to be predicted in AI service system invocation pattern library
Family decides through consultation rule or data-driven model obtains prediction result, exports to user.
All experts issued by AI service system decide through consultation rule or data-driven model, will be integrated into service.
When user to AI service system input certain class data (image, text, time series data, vector etc.) call AI service predict,
When identification, AI service system will merge all kinds of (system determine may correct) experts, and to decide through consultation that rule or data-driven model obtain pre-
It surveys as a result, specifically comprising the following steps: to user to export
When user input data calls AI service to be predicted, AI service system is found corresponding according to the input data
Expert decides through consultation rule and data-driven model.
It decides through consultation that rule and data-driven model judge the labeled data of input according to the expert found, is predicted
As a result;Wherein, in embodiment provided by the present invention, prediction result shows as misrepresenting deliberately rate of failing to report or final expected shortfall
Statistics calculate.The statistics by the expert decide through consultation labeled data collection that rule or new data driving model accumulate systems stay into
Row Diseases diagnosis (wherein each data contain final mark), obtains disease and misrepresents deliberately rate of failing to report and expected shortfall.
The prediction result that user selects is subjected to output displaying.Wherein, displaying includes: to show the Token power of prediction result
Beneficial price, disease misrepresent deliberately rate of failing to report, expected shortfall etc., and can be ranked up according to the output data of these dimensions.
In embodiment provided by the present invention, when user's selection is using some prediction result, it is possible to be different
Data-driven model or expert decide through consultation rule generation same output result (i.e. may different data-driven model or
Expert decides through consultation rule, provides same result in some specific input data), AI service system will choose wherein lowest price automatically
Expert decide through consultation rule or data-driven model, matched, and give the corresponding assessment system of the data-driven model with right
The Token excitation answered.
Token is paid AI service system after user selects prediction result by S4 automatically.Only when user's (AI service
Caller) pay equivalent Token after, can just consult prediction result and details.
If user selects one of them in multiple prediction results of push, according to the intelligence being arranged on block chain
Contract decides through consultation that the price of rule or data-driven model is corresponding to the payment of AI service system with the corresponding expert of prediction result
Token。
User specifically uses certain expert and decides through consultation regular or some data-driven using in whole industry AI service
One conclusion of model, then corresponding the paid Token of this time service is distributed by system, and specific Token is distributed to the section
The rule of point or various types of contributors of model;Remaining Token, system retain for total system lasting O&M,
The expenditure of the work such as market development, it can also be used to which the rule for the AI task that future increases newly is established and verified or labeled data is collected
And the advance payment of verifying.
In embodiment provided by the present invention, it is proposed that expert decides through consultation rule verification person and expert decide through consultation that rule is established and more
The interests of new person not can be carried out binding.If the expert for person's verifying that expert decides through consultation rule verification decides through consultation that mistake occurs in rule, will
Mistake expert decides through consultation that all verifiers of rule carry out error punishment.The Token wallet of mistake expert decides through consultation rule verification person
Certain expense will be detained, which is greater than the corresponding Token obtained of contribution of the verifying.The specific punishment amount of money is according to reality
Situation is set.
In embodiment provided by the present invention, after paying Token to AI service system according to intelligent contract, also
Include the following steps:
Transaction Information is written block chain, and is distributed to each accounting nodes on block chain by S5.
Further include following steps in embodiment provided by the present invention:
S6, when discriminating test result when the error occurs, fed back to AI service system.
When user feedback system prediction result mistake, block chain will be stored in after mark, and be forwarded to mark database.
When the number of feedback error is more than specific quantity threshold value, system will verify these prediction result data sets, and judgement should
Corresponding expert decides through consultation rule or data-driven model is mistake to occur on the data example of these user feedbacks, or use
Family itself misjudgment.When user feedback system prediction result mistake is correct feedback, based on intelligent contract, which is given
Give Token reward.
In embodiment provided by the present invention, after paying Token to AI service system according to intelligent contract, also
Include the following steps:
Rule is decided through consultation according to intelligent contract and the expert pre-established, according to a specific ratio or quantity point by Token
Dispensing node expert decides through consultation the different types of contributor of rule or data model.
Wherein, expert decides through consultation that regular contributor is broadly divided into the person of establising or updating and expert quotient that expert decides through consultation rule
The verifier of set pattern then.
Wherein, data model contributor is broadly divided into the algorithm model creator of data model (after regenerator is considered as update
The creator of model) and labeled data supplier.
User decides through consultation that rule or some data are driven using certain expert in whole industry AI service system, is specifically used
One conclusion of mover model, the then Token that this time service correspondence need to be paid, i.e., decide through consultation rule or data-driven using the expert
Model need to pay corresponding Virtual Token, the expert that the Token paid gives the node in specific proportions decide through consultation rule or
The contributor of data model, remaining Token, system retain the work such as the lasting O&M for this AI service system, market development
Expenditure, it can also be used to the rule of future newly-increased AI task is established and verifying or labeled data is collected and the advance payment of verifying
?.
Expert decides through consultation rule verification, and person must not decide through consultation the interests binding of regular contributor with expert.All errors validity persons,
Its Token wallet will be deducted the Token of acquirement corresponding to contribution than the verifying multiplied by the amount of money of punishment multiple.Specifically punish
It penalizes multiple to be set by platform, is greater than 1 in principle.
In embodiment provided by the present invention, the role for the distribution Token being related to is specifically included that
1) in terms of expert decides through consultation rule: deciding through consultation that the regular presenter of rule or expert decide through consultation rule more including new expert
New person also decides through consultation the rule verification person of rule comprising audit expert;
2) in terms of data-driven model: comprising providing the data set provider of labeled data, the data for auditing labeled data are tested
Card person, the model provider component or regenerator that algorithm (model) proposes.Labeled data supplier can be the user of AI service,
Mark feedback is provided in use, is also possible to that AI service is not used, it is simple that labeled data is provided and obtains remuneration person.
These roles according to percentage contribution obtain corresponding encryption currency or it is called obtain encryption currency be divided into power.
Wherein, node expert decides through consultation that the contributor of rule or data model includes that regular presenter, model provider component and data provide
Person.
Above-mentioned all rules or model contributory behaviour (the different type contribution comprising different role), rule and model
It calls price and all users to call the behavior, feedback behavior and Token trading activity of AI data processing, is recorded into
The accounting nodes of block chain.
Specifically, the foundation of data-driven model, needs initial labeled data collection (data set provider) and algorithm personnel
The joint efforts of both (model provider component).Therefore, the creator of a data-driven model is two groups of personnel.One side is number
According to contributor group, certain share is occupied.The personal virtual of the share, which is shared out bonus, to be weighed, and by its contribution proportion, i.e., everyone contributes and examines
The weighted contributions value of Nuclear Data accounts for the ratio of the weighted contributions total value of whole mark sample data.Another party is the calculation of algorithm personnel
Method contribution.
This AI service system is the service system of a closed loop, and user uses AI service system, if adopting system offer
Prediction result, will automatically form one verifying closed loop.In addition system rewards the problem of user's submission system, BUG and mistake
Prediction example (and being aided with user's checking or the subjective correct output result thought), feeds back excellent to complete the system of data closed loop
Change.System will reward these feedbacks, give the certain Token reward of user.
AI service system forbids plagiarism expert to decide through consultation rule and data-driven model.It is all to be proved to output and input result
The expert of indifference decide through consultation rule or data-driven model, will according to arrive first first well-behaved, the subsequent founder of automatic shield.But it is same
The expert of sample decides through consultation rule, data-driven model, if newly realizing that this realizes speed faster, system, which is automatically based upon, is divided into mechanism
Expert decides through consultation rule, and it is certain to give regenerator (data set provider of i.e. new regular presenter, new model provider component and Xin)
Be divided into.Regular founder or model founder enjoy " class patent " treatment, forbid any plagiarism within a certain period of time and imitate
System.After the period, then any newly-built expert decides through consultation rule or model, can copy and run at a low price.
Any primary expert decides through consultation regular examination, training data original provider and verifier and subsequent discovery
Model or expert decide through consultation that rule goes wrong and provide counter-example person, and according to the Token price of corresponding with service, (price can be with platform
Bid or negotiate to determine in internal market), give corresponding Token excitation.
In conclusion artificial intelligence service system provided by the present invention and method, by the way that block chain technical application is arrived
Professional domain, realization expert data and expertise are shared, and logical card mechanism is effectively guaranteed industry specialists, industry data tribute
The legitimate rights and interests of contributor, algorithm personnel and system maintenance personnel.
The embodiment of the invention also provides a kind of computer readable storage mediums.Here computer readable storage medium is deposited
Contain one or more program.Wherein, computer readable storage medium may include volatile memory, such as arbitrary access
Memory;Memory also may include nonvolatile memory, such as read-only memory, flash memory, hard disk or solid-state are hard
Disk;Memory can also include the combination of the memory of mentioned kind.Described in the computer readable storage medium one or
Multiple programs can be executed by one or more processor, to realize the above-mentioned artificial intelligence service based on intelligent contract and logical card
The part steps or Overall Steps of method.
The artificial intelligence service system provided by the present invention based on intelligent contract and logical card and method are carried out above
Detailed description.For those of ordinary skill in the art, it is done under the premise of without departing substantially from true spirit
Any obvious change, will all constitute the infringement weighed to the invention patent, corresponding legal liabilities will be undertaken.
Claims (15)
1. a kind of artificial intelligence service system based on intelligent contract and logical card, it is characterised in that including knowledge uploading module, know
Know authentication module, release module, prediction module and payment module;
Wherein, the knowledge uploading module is used to deciding through consultation expert into that rule or data-driven model upload to assessment system;
The expert being verified is decided through consultation rule for verify by what knowledge uploading module transmitted by the knowledge verification module
Then or data-driven model is sent to the release module;
The release module is used to decide through consultation the expert being verified that rule or the corresponding assessment system of data-driven model to be bound
Intelligent contract, and block chain is written and is issued to artificial intelligence service system;
When user input data calls artificial intelligence service to be predicted, the prediction module is issued special according to release module
Family decides through consultation that rule or data-driven model are predicted, and prediction result is exported to user;
The payment module will lead to according to intelligent contract from user account according to the prediction result of the prediction module selected with user
Card pays artificial intelligence service system, and carries out Transaction Information publication by the release module.
2. artificial intelligence service system as described in claim 1, it is characterised in that further include collision detection module;
For deciding through consultation that rule or data-driven model carry out conflict inspection to the expert of upload, judgement is the collision detection module
No is that new expert decides through consultation rule or new data driving model.
3. a kind of artificial intelligence method of servicing based on intelligent contract and logical card, it is characterised in that include the following steps:
Expert is decided through consultation that rule or data-driven model upload to assessment system, and is verified;
The newly-increased data-driven model being verified is stored in pattern base, and is published to corresponding artificial intelligence service system;Together
When Revenue Sharing Mechanism bound into intelligent contract, block chain is written;
When input data calls artificial intelligence service, the expert being applicable in artificial intelligence service system invocation pattern library decides through consultation rule
Then or data-driven model obtains prediction result;
After user selects prediction result, logical card is paid to artificial intelligence service system according to intelligent contract.
4. artificial intelligence method of servicing as claimed in claim 3, it is characterised in that expert decides through consultation rule or data-driven model
It uploads to after assessment system, further includes following steps before being verified:
To the expert of upload decide through consultation rule or data-driven model carry out conflict inspection, judge whether be new expert decide through consultation rule or
Person's new data driving model.
5. artificial intelligence method of servicing as claimed in claim 4, which is characterized in that decide through consultation that rule is rushed to the expert of upload
It is prominent to examine, judge whether it is that new expert decides through consultation rule, includes the following steps:
S 101 judges to decide through consultation that rule repeats or part repeats with the presence or absence of the new expert with upload in artificial intelligence service system
Old expert decide through consultation rule, if it is present turn to step S 102;Otherwise, step S 103 is turned to;Wherein, it repeats or local
Repeatedly refer to that the restrictive condition of applicable input data is identical or partially overlaps;
S 102 judges that new expert decides through consultation whether rule is updated, if it is, turning to compared with old expert decides through consultation rule
Step S 103;Otherwise, refuse the upload that new expert decides through consultation rule;
S 103 decides through consultation that rule carries out expert's verifying to the new expert of upload.
6. artificial intelligence method of servicing as claimed in claim 5, it is characterised in that described pair upload expert decide through consultation rule into
Row expert verifying, includes the following steps:
According to the expert of upload decide through consultation rule belonging to field determine the Committee of Experts member that is audited;
Verification threshold is set, when the expert that verifying uploads decides through consultation that the Committee of Experts membership of rules compliance is greater than verification threshold
When, the expert of upload decides through consultation rules compliance, issues the new expert and decides through consultation rule;
Otherwise, refuse the expert and decide through consultation that rule is issued to artificial intelligence service system.
7. artificial intelligence method of servicing as claimed in claim 4, which is characterized in that rushed to the data-driven model of upload
It is prominent to examine, judge whether it is new data driving model, includes the following steps:
S 111, by original data-driven model and newly-increased data-driven model simultaneously to the mark number of systems stay accumulation
Prediction test is carried out according to collection, and marks the consistency of new data-driven model and every Geju City model one by one;
S 112, assessment system judge whether there is an existing old model, and output exports completely the same with new model;Such as
Fruit has, then refuses the publication of new data driving model;If there is no such model, then step S 113 is turned to;
S 113, assessment system carry out assessment verifying to the new data driving model of upload.
8. artificial intelligence method of servicing as claimed in claim 7, it is characterised in that described pair upload data-driven model into
Row assessment verifying, includes the following steps:
Assessment system is based on corresponding mark sample data set, calculates the misrepresenting deliberately of the data-driven model of upload, rate of failing to report;
Assessment system will misrepresent deliberately, rate of failing to report multiplied by corresponding cost function obtains final expected shortfall, if new data drives
The final expected shortfall of model is less than the final expected shortfall of original data-driven model, and is lower than the mould of systemic presupposition
Type predicts loss function threshold value, then new data driving model is verified.
9. artificial intelligence method of servicing as claimed in claim 5, it is characterised in that:
Judge that expert decides through consultation whether rule updates and decides through consultation whether the input of rule adds new data object, new special including new expert
Family decides through consultation whether rule further segments type to the legacy data object of input, new expert decides through consultation whether the input object of rule increases
Additive attribute, new expert decide through consultation whether rule modifies attribute, new expert decides through consultation whether rule by old expert decides through consultation that rule is carried out into one
The one or more that step divides, some condition or parameter change in rule.
10. artificial intelligence method of servicing as claimed in claim 5, it is characterised in that:
When the expert that Committee of Experts member uploads expert decides through consultation that rule is audited, by deciding through consultation rule to the expert of upload
It is then modified and establishes branch's sub-rule, the expert for establishing branch's sub-rule is the owner of the new rule.
11. artificial intelligence method of servicing as claimed in claim 3, it is characterised in that when input data calls artificial intelligence service
When being predicted, it is pre- that the expert being applicable in artificial intelligence service system invocation pattern library decides through consultation that rule or data-driven model obtain
Survey as a result, and exported, include the following steps:
When input data calls artificial intelligence service to be predicted, artificial intelligence service system is found pair according to the input data
The expert answered decides through consultation rule and data-driven model;
It decides through consultation that rule and data-driven model judge the data of input according to the expert found, obtains prediction result;
The prediction result of selection is subjected to output displaying.
12. artificial intelligence method of servicing as claimed in claim 3, it is characterised in that:
When the prediction result of selection, which corresponds to multiple experts, decides through consultation rule or data-driven model, artificial intelligence service system will
The expert for choosing minimum transaction value decides through consultation that rule or data-driven model match.
13. artificial intelligence method of servicing as claimed in claim 3, it is characterised in that:
When the number of user feedback system prediction result mistake be more than specific quantity threshold value when, to these prediction result data sets into
Row is verified, and judges that the corresponding expert decides through consultation rule or data-driven model is sent out on the data example of these user feedbacks
Raw mistake or user itself misjudgment.
14. artificial intelligence method of servicing as claimed in claim 13, it is characterised in that further include following steps:
When user feedback system prediction result mistake is correct feedback, based on intelligent contract, logical card reward is given to the user.
15. artificial intelligence method of servicing as claimed in claim 3, it is characterised in that further include following steps:
It decides through consultation that logical card is distributed to different types of contributor by rule according to intelligent contract and the expert pre-established, and records
Enter the accounting nodes of block chain.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810901731.1A CN109191293B (en) | 2018-08-09 | 2018-08-09 | Artificial intelligence service system and method based on intelligent contract and general evidence |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810901731.1A CN109191293B (en) | 2018-08-09 | 2018-08-09 | Artificial intelligence service system and method based on intelligent contract and general evidence |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109191293A true CN109191293A (en) | 2019-01-11 |
CN109191293B CN109191293B (en) | 2023-11-14 |
Family
ID=64921183
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810901731.1A Active CN109191293B (en) | 2018-08-09 | 2018-08-09 | Artificial intelligence service system and method based on intelligent contract and general evidence |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109191293B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110222721A (en) * | 2019-05-10 | 2019-09-10 | 深圳前海达闼云端智能科技有限公司 | Data processing method, device, block chain node and storage medium |
CN110276613A (en) * | 2019-06-20 | 2019-09-24 | 卓尔智联(武汉)研究院有限公司 | Data processing equipment, method and computer readable storage medium based on block chain |
CN110533419A (en) * | 2019-07-24 | 2019-12-03 | 阿里巴巴集团控股有限公司 | Sharing method, device and the server of air control model based on block chain |
CN111062705A (en) * | 2019-12-25 | 2020-04-24 | 东华理工大学 | Online English education big data platform |
CN111526180A (en) * | 2020-03-30 | 2020-08-11 | 中国建设银行股份有限公司 | Data sharing method and device |
CN111901753A (en) * | 2020-06-04 | 2020-11-06 | 中国联合网络通信集团有限公司 | Infectious disease close contact person determination method based on block chain and block chain link point |
CN117235782A (en) * | 2023-08-31 | 2023-12-15 | 北京可利邦信息技术股份有限公司 | Method, system and terminal for realizing privacy calculation data security based on alliance chain |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101093559A (en) * | 2007-06-12 | 2007-12-26 | 北京科技大学 | Method for constructing expert system based on knowledge discovery |
CN107194587A (en) * | 2017-05-24 | 2017-09-22 | 暨南大学 | It is a kind of to be circulated based on block chain and the art work of expert system and identification register method |
CN107240018A (en) * | 2017-07-25 | 2017-10-10 | 成都励睿德企业管理有限公司 | A kind of method and system for being used to pay tranaction costs in block chain network |
CN107563905A (en) * | 2017-07-20 | 2018-01-09 | 西安电子科技大学 | A kind of academic platform service system and method for building up based on block chain |
-
2018
- 2018-08-09 CN CN201810901731.1A patent/CN109191293B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101093559A (en) * | 2007-06-12 | 2007-12-26 | 北京科技大学 | Method for constructing expert system based on knowledge discovery |
CN107194587A (en) * | 2017-05-24 | 2017-09-22 | 暨南大学 | It is a kind of to be circulated based on block chain and the art work of expert system and identification register method |
CN107563905A (en) * | 2017-07-20 | 2018-01-09 | 西安电子科技大学 | A kind of academic platform service system and method for building up based on block chain |
CN107240018A (en) * | 2017-07-25 | 2017-10-10 | 成都励睿德企业管理有限公司 | A kind of method and system for being used to pay tranaction costs in block chain network |
Non-Patent Citations (2)
Title |
---|
卢力行: ""基于知识管理的专家资源管理平台的设计与实现"" * |
薛腾飞 等: ""基于区块链的医疗数据共享模型研究"" * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110222721A (en) * | 2019-05-10 | 2019-09-10 | 深圳前海达闼云端智能科技有限公司 | Data processing method, device, block chain node and storage medium |
CN110222721B (en) * | 2019-05-10 | 2021-07-30 | 达闼机器人有限公司 | Data processing method, data processing device, block chain node and storage medium |
CN110276613A (en) * | 2019-06-20 | 2019-09-24 | 卓尔智联(武汉)研究院有限公司 | Data processing equipment, method and computer readable storage medium based on block chain |
CN110533419A (en) * | 2019-07-24 | 2019-12-03 | 阿里巴巴集团控股有限公司 | Sharing method, device and the server of air control model based on block chain |
CN110533419B (en) * | 2019-07-24 | 2023-09-26 | 创新先进技术有限公司 | Block chain-based sharing method, device and server of wind control model |
CN111062705A (en) * | 2019-12-25 | 2020-04-24 | 东华理工大学 | Online English education big data platform |
CN111526180A (en) * | 2020-03-30 | 2020-08-11 | 中国建设银行股份有限公司 | Data sharing method and device |
CN111526180B (en) * | 2020-03-30 | 2022-05-24 | 中国建设银行股份有限公司 | Data sharing method and device |
CN111901753A (en) * | 2020-06-04 | 2020-11-06 | 中国联合网络通信集团有限公司 | Infectious disease close contact person determination method based on block chain and block chain link point |
CN117235782A (en) * | 2023-08-31 | 2023-12-15 | 北京可利邦信息技术股份有限公司 | Method, system and terminal for realizing privacy calculation data security based on alliance chain |
Also Published As
Publication number | Publication date |
---|---|
CN109191293B (en) | 2023-11-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109191293A (en) | A kind of artificial intelligence service system and method based on intelligent contract and logical card | |
US20230252017A1 (en) | Community data aggregation with automated followup | |
US11080793B2 (en) | Method of personalizing, individualizing, and automating the management of healthcare fraud-waste-abuse to unique individual healthcare providers | |
US9779407B2 (en) | Healthcare fraud preemption | |
Halpern et al. | Health and economic outcomes modeling practices: a suggested framework | |
US20150046181A1 (en) | Healthcare fraud protection and management | |
CN112735549A (en) | Data processing method and data processing system based on medical insurance data | |
Matuszak et al. | Performance/outcomes data and physician process challenges for practical big data efforts in radiation oncology | |
Trout et al. | A quality improvement initiative for COPD patients: A cost analysis | |
Durovich et al. | Designing a community-based population health model | |
CN110168659A (en) | For improving clinical effectiveness and reducing the nursing always nursing the CNA of expense and guiding | |
CN102893302A (en) | Clinical payment network system and methods | |
Burdett et al. | The efficacy of utility functions for multicriteria hospital case‐mix planning | |
Jakab | An empirical evaluation of the Kyrgyz health reform: does it work for the poor? | |
Fang et al. | A 2-additive Choquet integral-based multi-criterion decision-making method with complex linguistic information in drug value assessment | |
Kurnianingtyas et al. | Reforming premium amount in the Indonesian National Health Insurance System program using system dynamics | |
Xu et al. | What promotes medical overuse: Perspective on evolutionary game between administration and Medical Institutions | |
Feldhaus | Identifying health system priorities for equitable access to health services in low-and middle-income countries | |
US20240006061A1 (en) | Data currency system for digital human healthcare and medical data exchange, analytics, and applications | |
Lakkaraju | Human-Centric Machine Learning: Enabling Machine Learning for High-Stakes Decision-Making | |
WO2023181880A1 (en) | Information processing method, program, and information processing device | |
Jia | Embracing Machine Learning in Safety Assurance in Healthcare | |
US20240070627A1 (en) | Healthcare Blockchain Cryptosystem | |
Reynolds | Using Symbolic Data Analysis to Detect Fraud, Waste, and Abuse in Healthcare Insurance Claims Data | |
Callea et al. | PNS41 Estimating Administrative Costs Linked to Medical Devices Recalls from the Perspective of the Healthcare Provider |
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 | ||
CB02 | Change of applicant information |
Address after: Room 1101-61, No. 379 Tayuan Road, Suzhou High-tech Zone, Suzhou City, Jiangsu Province 215011 Applicant after: Suzhou Navitas Space-Time Information Technology Co.,Ltd. Address before: Room 1-403, Block B, Qilu Software Park Building, No. 1768 Xinluo Street, High tech Zone, Jinan City, Shandong Province, 250101 Applicant before: JINAN NAWEI INFORMATION TECHNOLOGY Co.,Ltd. |
|
CB02 | Change of applicant information | ||
GR01 | Patent grant | ||
GR01 | Patent grant |