CN101872345A - Logical reasoning mechanism being suitable for legal expert system - Google Patents
Logical reasoning mechanism being suitable for legal expert system Download PDFInfo
- Publication number
- CN101872345A CN101872345A CN200910059074A CN200910059074A CN101872345A CN 101872345 A CN101872345 A CN 101872345A CN 200910059074 A CN200910059074 A CN 200910059074A CN 200910059074 A CN200910059074 A CN 200910059074A CN 101872345 A CN101872345 A CN 101872345A
- Authority
- CN
- China
- Prior art keywords
- module
- reasoning
- fuzzy
- legal
- expert system
- 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
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a logical reasoning mechanism being suitable for a legal expert system, effectively overcomes the malady that the current legal expert system is based on regulation, and solves the problems that the majority of legal expert systems can not apply judging knowledge for reasoning, only provides simple explanation through regulation feedback, the problems that uncertainty how to stimulate the intuition, experience and reasoning result and other problems are expressed dissatisfactorily. The reasoning mechanism introduces a correlation theory and a method of fuzzy reasoning, abandons simple positive and negative two-value features of traditional logic, regards the logical world into gray level with continuous changes, allows a proposition to be this or that, and exists partial positive and partial negative. The invention provides a more natural and more reasonable operation environment, can describe target system behavior adopting a fuzzy logic technology, imitates people to judge and control behavior technology by cognition and experience, can transform some natural languages into algorithms of computer operation to achieve the purpose of using a simple program to process complex systems which can not be processed or is difficult to be processed by traditional control method. The fuzzy logic correlation theory and method are applied in the legal expert system, can well overcome the shortcomings of knowledge expression, reasoning mechanism and knowledge acquisition and the like of the traditional method in the legal expert system, are more convenient to express numerous uncompleted and inaccurate knowledge and relationship in legal fields, more appropriately describe uncertainty in legal reasoning, makes up for defects of traditional reasoning method, and are suitable for large-scale popularization.
Description
Affiliated technical field
The present invention relates to the design of the logical inference machine of artificial intelligence legal system, be specially adapted to the reasoning from logic of law expert system under complex situations.
Background technology
The law expert system is an important achievement using at legal field of artificial intelligence technology in recent decades, it not only can provide effective legal consultation and help for common suitable person, can also alleviate professionals' such as judge, lawyer labour intensity, improve the efficient and the quality of law work, bring huge social benefit and economic benefit.
At present, existing law expert system can integrated use knowledge various method for expressing, assisted retrieval, the electronics lawyer of rules and legal precedent, electronics judge's the legal aid and the auxiliary measurement of penalty have been realized, a part of brainwork of having liberated the bench and bar to a certain extent.But, because its own particularity of legal field and the limitation of the theory and technology that the law expert system is relied on when construction, there is a large amount of problems in current law expert system on inference mechanism, be mainly reflected in, a lot of knowledge of expert are experimental when exercising common sense sex knowledge or empirical thinking, the expert is in thinking, be not all to be the process that follows rational thinking, their great majority can not use the judgement sex knowledge to carry out reasoning, can only feed back by rule simplicity of explanation is provided, solving how to simulate the intuitive of Templar's reasoning, what show on the problems such as uncertainty of empirical and The reasoning results is unsatisfactory, and it is real intelligent therefore also to be far from being.
Thinking of the present invention is to introduce the correlation theory of fuzzy logic inference and the limitation that method overcomes formal logic inference.Fuzzy logic is a kind of multi valued logic based on fuzzy set theory, it has abandoned the binary feature of the simple affirmation and negation of traditional logic, the logic world regarded as have the continually varying gray scale, allow a proposition to be this or that, exist part certainly and part negate.It provides a kind of more natural, more rational operating environment, handles ubiquitous, coarse fuzzy message for the apish mode of thinking of computing machine bigger possibility is provided.Adopt fuzzy logic technology can use the behavior of natural language description goal systems, technology with consciousness and judgement of Empirical Mode apery and control behavior, can also be converted to some natural language the calculation rule of computer operation, with reach with simple program handle those with traditional control method can't or the purpose of the complication system of intractable.The correlation theory of fuzzy logic and method are applied to solve a lot of deficiencies of classic method at the aspects such as knowledge representation, inference mechanism and knowledge acquisition of law expert system in the law expert system well, represent numerous incomplete, coarse knowledge of legal field and relation more easily; Describe uncertain property in the analogy of law more appropriately, remedy the defective of traditional inference method; Accelerate search speed and improve the 26S Proteasome Structure and Function of whole law expert system.
Summary of the invention
System at first carries out the case coupling to the case information feature according to the initial information of user's input, and it fails to match then turns to fuzzy reasoning, and reasoning finishes then to provide appropriate conclusion on the basis of comprehensive two kinds of The reasoning results.
The case matching module comprises 5 modules such as user/system interaction module, case matching degree computing module, explanation function module, conclusion judge module and process control module, this 5 modules cooperates jointly, co-ordination, finish the case matching feature together, the working mechanism of each module as shown in Figure 1:
Wherein, user/system interaction module mainly is service being provided alternately and interaction data being sent to a module of inference machine for user and system.After the start-up system, enter the case inputting interface, the case categorizing selection that the user provides from system, and submission data, user/system interaction module sends case matching degree computing module to after receiving data, and sends user's input record to the explanation function module, writes log information.
Case matching degree computing module is the chief component of reasoning by cases machine, it is under the control of process control module, carry out the computing of case matching degree according to user/user data that the system interaction module provides, and send the result to the conclusion judge module and carry out rationality and judge.At first, the degree of membership to self case of each feature in the storehouse is calculated according to the record in the case library by system, finish the input service of detailed case the user after, system calculates the degree of membership of different characteristic to each case in the case library according to user's input feature vector, utilize the degree of membership contrast of feature in the degree of membership of user's input feature vector and the case library to calculate the hamming distance, the matching degree of hamming relative distance and input feature vector and different cases, the case of selecting matching degree maximum wherein sends the conclusion judge module to as the conclusion case and carries out rationality and judge, and the detail operations of reasoning by cases is passed to the explanation function module, and write log information by process control module.
The conclusion judge module is that The reasoning results is carried out the module that rationality is differentiated, provides the reasoning conclusion whether reasonable discriminating work according to the threshold value of calculating to the user.After case is judged end, case matching degree computing module is responsible for the reasoning by cases conclusion is passed to the conclusion judge module, the conclusion judge module is then judged The reasoning results according to the matching degree threshold value that system provides, and thinks then that more than or equal to threshold value reasoning is reasonable, finishes reasoning; Otherwise the reasoning failure sends the fuzzy reasoning application to process control module, and the control inference machine carries out fuzzy reasoning, and by process control module the detail operations of conclusion rationality judgement is passed to the explanation function module, and writes log information.
Process control module is mainly realized the flow process control of inference machine, and the control case coupling and the related operation of fuzzy reasoning hereinafter, is responsible at last the detail operations result of each flow process is passed to the explanation function module.In the case reasoning process, the computing flow process of process control module major control feature degree of membership, hamming distance, hamming relative distance and final case matching degree; After the reasoning of case coupling finished, according to the discriminating conclusion of conclusion judge module, whether process control module carried out fuzzy reasoning with control system, after the reasoning by cases failure, then control system is proceeded fuzzy reasoning, and detail operations is passed to the explanation function module, and writes log information.
The explanation function module mainly is responsible for the explanation function of system and writing of daily record.When beginning to diagnose, the explanation function module is responsible for the detailed case history of user's input is write daily record, in case analysis and fuzzy analysis process hereinafter, the explanation function module will be responsible for the calculating process of case and fuzzy reasoning and reasoning conclusion are write daily record, and last diagnostic result is recorded in the daily record.
The fuzzy reasoning module is except process control module, explanation function module, also comprise three big modules such as inference machine interactive module, fuzzy membership computing module, system's conclusion judge module, this 5 modules cooperates jointly, finish the fuzzy reasoning function together, the working mechanism of fuzzy reasoning module as shown in Figure 2.
The inference machine interactive module is the data transmission channel between reasoning by cases machine and the indistinct logic computer, and main being responsible for is transferred to indistinct logic computer so that proceed reasoning with user's input feature vector information and correlation values.After reasoning by cases finishes, if the conclusion diagnostic module is thought the reasoning by cases failure through differentiating, then the inference machine interactive module is responsible for data such as case information with user's input and is passed to the fuzzy membership computing module accurately and carry out fuzzy reasoning under the domination of process control module, and finally by process control module relevant information is passed to the ambiguous interpretation module.
The fuzzy membership computing module is the nucleus module of indistinct logic computer, it is under the control of process control module, the user data that provides of machine interactive module blurs the computing of total degree of membership by inference, and sends the result to the conclusion judge module and carry out rationality and judge.Before the fuzzy reasoning, system goes out the weights of each feature to affiliated insect pest according to the feature calculation in the fuzzy characteristics storehouse earlier; When reasoning by cases is failed, under the control of process control module, the inference machine interactive module will be the detailed case of user input then is passed to the fuzzy membership computing module, computing module then goes out the weights of user's input feature vector to each case according to the characteristic information and the information calculations in the fuzzy characteristics storehouse of user's input, and the feature that calculates each case is always weighed, finally the weights of each case are obtained the user with total power and import the total degree of membership of case characteristic for each case according to the trial feature, be the fuzzy reasoning result of input feature vector, selecting wherein the conclusion as fuzzy reasoning of total degree of membership maximum to be transferred to the conclusion judge module carries out rationality and judges, and the detailed process of fuzzy reasoning is sent to the ambiguous interpretation module, the writing system daily record by process control module.
System's conclusion judge module is after inference machine is finished reasoning work, in conjunction with the The reasoning results of reasoning by cases machine and indistinct logic computer, and the comprehensive reasoning conclusion that provides system.After fuzzy reasoning finishes, whether system will effective according to the total degree of membership threshold decision fuzzy reasoning result who calculates, if effectively, then the matching degree of total degree of membership of fuzzy reasoning result with the case coupling compared, select the most believable conclusion final conclusion the most, and the most auxiliary conclusion of another conclusion; If fuzzy reasoning is also failed, then equally according to the confidence level comparative result, select the most believable conclusion final conclusion the most, and the most auxiliary conclusion of another conclusion, and prompting reasoning failure, the suggestion user revises case library, and finally the detail operations of system's conclusion rationality being judged by process control module passes to the ambiguous interpretation functional module, and writes log information.
Description of drawings
Fig. 1 is a case matching module working mechanism
Fig. 2 is a fuzzy reasoning module working mechanism
Fig. 3 overall system design
Fig. 4 inference machine workflow
Embodiment
Logical inference machine is made up of batch processing, realizes the reasoning of problem is found the solution.Comprise following work:
(1) mates according to predetermined policy known fact in selective rule and the integrated data base from rule base.So-called coupling is meant the precondition of rule and the known fact in the integrated data base is compared, if both consistent or approximate and satisfied degrees of approximation of predesignating become then that the match is successful, respective rule can be used; Otherwise it is unsuccessful to become coupling, and respective rule is not useable for current reasoning;
(2), then become and clash if the rule that the match is successful has many.At this moment, inference machine must be employed the program of carrying out certain Strategy of Conflict Resolution, therefrom selects a rule that the match is successful and carries out.
(3) carrying out a certain when regular,, then these conclusions are being added in the integrated data base if consequent that should rule is one or more conclusions; If these operations are then carried out in the one or more operations of consequent of rule successively.
(4), when carrying out each bar rule, also to check the uncertainty of conclusion by certain algorithm for uncertain knowledge.
(5) check the condition that inference machine moves at any time, when satisfying termination condition, stop the operation of inference machine.Usually, the work period of an inference machine is made up of pattern match, selection and execution three phases, as shown in Figure 4.The groundwork in each stage is summarized as follows in work period
1. coupling (unification) (promptly from knowledge base article one rule beginning) strictly all rules in the scanning rule storehouse successively top down compares all elements of working storage and the former piece of strictly all rules one by one, the rule that satisfies condition with search.The match is successful if all conditions in regular former piece is all with the current fact of working storage, then this rule put in the conflict set, carries out the detection of next bar rule then, and strictly all rules is all detected in rule base.
2. select the situation that (conflict resolution) many rules are mated simultaneously to be called conflict, at this moment will obtain the relative importance value of institute's conflict rule according to predetermined interpretational criteria, (triggering) which bar rule is selected in decision.The strategy of conflict resolution has multiple, for example:
(1) selects by the order that the match is successful: the preferential selection rule that the match is successful at first.
(2) according to priority select: the preferential the highest rule of selecting priority.
(3) select by the level of detail: preferentially select premise part to describe the most detailed rule.
(4) select by the order of carrying out: the rule that preferential selection is carried out recently.
(5) select by new fact: the preferential selection and the up-to-date relevant rule of the fact that is added in the working storage.
(6) whether used selection by rule: the preferential not used rule originally of selecting.Above strategy cuts both ways, and real system usually is used in combination diverse ways.
3. carry out (action)
The conclusion of selected rule is added to working storage, as the new fact; During operation, inference mechanism repeats the circulation of this three phases, according to the fact of knowledge in the rule base and working storage, constantly release unknown conclusion by known prerequisite, and record in the working storage, as new prerequisite or the true reasoning process that continues, up to releasing final conclusion.
Claims (3)
1. logical reasoning mechanism that is applicable to the law expert system, it is characterized in that: the employing fuzzy logic inference is a core.Its nucleus module comprises case matching module and fuzzy reasoning module.
2. case matching module as claimed in claim 1, it is characterized in that: the case matching module comprises 5 modules such as user/system interaction module, case matching degree computing module, explanation function module, conclusion judge module and process control module, this 5 modules cooperates jointly, co-ordination, finishes the case matching feature together.
3. fuzzy reasoning module as claimed in claim 1, it is characterized in that: the fuzzy reasoning module comprises beyond process control module, the explanation function module, inference machine interactive module, fuzzy membership computing module, system's conclusion judge module, these five modules are finished the fuzzy reasoning function together.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910059074A CN101872345A (en) | 2009-04-27 | 2009-04-27 | Logical reasoning mechanism being suitable for legal expert system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910059074A CN101872345A (en) | 2009-04-27 | 2009-04-27 | Logical reasoning mechanism being suitable for legal expert system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101872345A true CN101872345A (en) | 2010-10-27 |
Family
ID=42997210
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910059074A Pending CN101872345A (en) | 2009-04-27 | 2009-04-27 | Logical reasoning mechanism being suitable for legal expert system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101872345A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103782245A (en) * | 2011-07-12 | 2014-05-07 | 西门子公司 | Actuation of a technical system |
| CN104679780A (en) * | 2013-12-02 | 2015-06-03 | 方正信息产业控股有限公司 | Data processing method and device and server |
| TWI494890B (en) * | 2011-04-12 | 2015-08-01 | Univ Southern Taiwan | Support system of sentencing decision |
| TWI588771B (en) * | 2016-06-08 | 2017-06-21 | 南臺科技大學 | Thesis analysis system |
| CN107361743A (en) * | 2017-07-17 | 2017-11-21 | 广西犇云科技有限公司 | Traditional Chinese medical science intelligent diagnosis system and method |
| CN111539529A (en) * | 2020-04-15 | 2020-08-14 | 东莞证券股份有限公司 | Event reasoning method and device |
| CN111897960A (en) * | 2020-07-17 | 2020-11-06 | 南京擎盾信息科技有限公司 | Method, device, equipment and storage medium for reasoning between dynamic legal events |
| CN113836605A (en) * | 2021-09-08 | 2021-12-24 | 华中科技大学 | Intelligent management system for gear metal powder injection molding |
-
2009
- 2009-04-27 CN CN200910059074A patent/CN101872345A/en active Pending
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI494890B (en) * | 2011-04-12 | 2015-08-01 | Univ Southern Taiwan | Support system of sentencing decision |
| CN103782245A (en) * | 2011-07-12 | 2014-05-07 | 西门子公司 | Actuation of a technical system |
| US9449275B2 (en) | 2011-07-12 | 2016-09-20 | Siemens Aktiengesellschaft | Actuation of a technical system based on solutions of relaxed abduction |
| CN103782245B (en) * | 2011-07-12 | 2016-12-21 | 西门子公司 | The manipulation of technological system |
| CN104679780A (en) * | 2013-12-02 | 2015-06-03 | 方正信息产业控股有限公司 | Data processing method and device and server |
| TWI588771B (en) * | 2016-06-08 | 2017-06-21 | 南臺科技大學 | Thesis analysis system |
| CN107361743A (en) * | 2017-07-17 | 2017-11-21 | 广西犇云科技有限公司 | Traditional Chinese medical science intelligent diagnosis system and method |
| CN111539529A (en) * | 2020-04-15 | 2020-08-14 | 东莞证券股份有限公司 | Event reasoning method and device |
| CN111539529B (en) * | 2020-04-15 | 2023-06-20 | 东莞证券股份有限公司 | Event reasoning method and device |
| CN111897960A (en) * | 2020-07-17 | 2020-11-06 | 南京擎盾信息科技有限公司 | Method, device, equipment and storage medium for reasoning between dynamic legal events |
| CN113836605A (en) * | 2021-09-08 | 2021-12-24 | 华中科技大学 | Intelligent management system for gear metal powder injection molding |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101872345A (en) | Logical reasoning mechanism being suitable for legal expert system | |
| Feng et al. | Reinforcement learning for relation classification from noisy data | |
| CN110609891B (en) | Visual dialog generation method based on context awareness graph neural network | |
| Alcalá et al. | A fast and scalable multiobjective genetic fuzzy system for linguistic fuzzy modeling in high-dimensional regression problems | |
| CN110097185B (en) | Optimization model method based on generation of countermeasure network and application | |
| CN111859978B (en) | Deep learning-based emotion text generation method | |
| CN110364251B (en) | Intelligent interactive diagnosis guide consultation system based on machine reading understanding | |
| Guh | A hybrid learning-based model for on-line detection and analysis of control chart patterns | |
| CN110647612A (en) | Visual conversation generation method based on double-visual attention network | |
| CN112000772B (en) | Sentence-to-semantic matching method based on semantic feature cube and oriented to intelligent question and answer | |
| CN109635108A (en) | A kind of remote supervisory entity relation extraction method based on human-computer interaction | |
| García et al. | Overview of the SLAVE learning algorithm: A review of its evolution and prospects | |
| Song et al. | Image captioning with context-aware auxiliary guidance | |
| CN114871000B (en) | Flotation dosing self-adaptive regulation and control method | |
| CN112000770A (en) | Intelligent question and answer oriented sentence-to-sentence matching method based on semantic feature map | |
| CN110727844A (en) | Online commented commodity feature viewpoint extraction method based on generation countermeasure network | |
| Huang et al. | Harnessing deep learning for population genetic inference | |
| Dessalene et al. | Therbligs in action: Video understanding through motion primitives | |
| CN113743461B (en) | Unmanned aerial vehicle cluster health degree assessment method and device | |
| CN107491735A (en) | The label and interactive relation combination learning method of human behavior identification | |
| CN110473636A (en) | Intelligent doctor's advice recommended method and system based on deep learning | |
| CN116910223B (en) | Intelligent question-answering data processing system based on pre-training model | |
| Qiao et al. | Cardinality estimator: processing SQL with a vertical scanning convolutional neural network | |
| Nam et al. | Selective prediction with long short-term memory using unit-wise batch standardization for time series health data sets: algorithm development and validation | |
| CN114239744B (en) | Individual processing effect evaluation method based on variational generation countermeasure network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| DD01 | Delivery of document by public notice |
Addressee: Liu Dongmei Document name: Notification of Publication of the Application for Invention |
|
| DD01 | Delivery of document by public notice |
Addressee: Liu Dongmei Document name: Notification of before Expiration of Request of Examination as to Substance |
|
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20101027 |