CN106815266A - Judgement document's search method and device - Google Patents
Judgement document's search method and device Download PDFInfo
- Publication number
- CN106815266A CN106815266A CN201510869926.9A CN201510869926A CN106815266A CN 106815266 A CN106815266 A CN 106815266A CN 201510869926 A CN201510869926 A CN 201510869926A CN 106815266 A CN106815266 A CN 106815266A
- Authority
- CN
- China
- Prior art keywords
- judgement document
- document
- judgement
- relevance score
- parameters
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/34—Browsing; Visualisation therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/31—Indexing; Data structures therefor; Storage structures
- G06F16/313—Selection or weighting of terms for indexing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This application discloses a kind of judgement document's search method and device.Wherein, the method includes：Judgement document's set that acquisition is matched with term；Every the relevance score and complexity of judgement document in judgement document's set are calculated, wherein, relevance score is used to represent the matching degree of judgement document and term, and complexity is used to represent the complexity of judgement document；Respectively according to every amendment relevance score of judgement document in relevance score and complicated dynamic behaviour judgement document set, multiple amendment relevance scores corresponding with many judgement documents in judgement document's set respectively are obtained；Judgement document in being gathered judgement document according to multiple amendment relevance score is ranked up, and obtains ranking results；And show the judgement document in judgement document's set according to ranking results.By the application, accuracy relatively low technical problem when the retrieval result in correlation technique to judgement document is ranked up is solved.
Description
Technical field
The application is related to data processing field, in particular to a kind of judgement document's search method and device.
Background technology
Since 2014, under the requirement of the Supreme People's Court, people's courts at different levels start progressively big on the internet
The open legally effective judgement document in scale ground.Largely judgement document has been come into force to laws such as judge, lawyer, jurisprudential study personnel
For the legal practice of worker, with reference and researching value very high.Because judgement document is long article notebook data,
Therefore information retrieval technique is used widely in this regard.
However, traditional judgement document's retrieval only stops at hit, that is, return complete with user search matching criteria
Portion's document.If qualified document quantity has many pieces, it is impossible to complete in former pages of displayings of result of page searching,
Current technology can only be shown the sequence of priority according to correlation to these documents.This retrieval mode only with
The correlation that keyword is weighed have ignored other characteristics that judgement document includes as the benchmark of sequence, also have ignored
The real demand of user.Not Case Law System is national for China, and judge needs strictly to specify according to law during judge
Carry out deduction of arguing, thus user wish to be those argue fully, the case that has a major impact.
Accuracy relatively low problem when being ranked up for the retrieval result in correlation technique to judgement document, at present not yet
Propose effective solution.
The content of the invention
The main purpose of the application is to provide a kind of judgement document's search method and device, right in correlation technique to solve
Accuracy relatively low problem when the retrieval result of judgement document is ranked up.
To achieve these goals, according to the one side of the application, there is provided a kind of judgement document's search method.Should
Method includes：Judgement document's set that acquisition is matched with term；Calculate every judgement document in judgement document's set
Relevance score and complexity, wherein, relevance score is used to represent the matching degree of judgement document and term, complicated
Spend the complexity for representing judgement document；Respectively according in relevance score and complicated dynamic behaviour judgement document set
Every amendment relevance score of judgement document, obtains corresponding many with many judgement documents in judgement document's set respectively
Individual amendment relevance score；Judgement document in being gathered judgement document according to multiple amendment relevance score is ranked up,
Obtain ranking results；And show the judgement document in judgement document's set according to ranking results.
Further, judgement document's set includes the first judgement document, calculates every judge's text in judgement document's set
The complexity of book includes：The index parameter of the first judgement document is obtained, wherein, index parameter includes following parameter extremely
It is one of few：The length of the first judgement document, the lawsuit gold for being applicable law number and the first judgement document of the first judgement document
Volume；And the complexity of the first judgement document is calculated according to index parameter.
Further, the complexity for calculating the first judgement document according to index parameter includes：Obtain index parameter in each
The weight of parameter；And the value according to parameters in index parameter and the judgement document of weight calculation first of parameters
Complexity.
Further, the method also includes：According to following preset rules, weight is set：Calculate multiple default judgement documents
Every relevance score of judgement document in set, wherein, the default judgement document's collection of multiple is combined into default with multiple respectively
Multiple judgement documents set of training word matching；Calculate every complexity of judgement document in multiple default judgement document's set
Degree, wherein, the initial weight computation complexity of value and parameters according to parameters in index parameter, every ginseng
It is corresponding consistent with the index parameter that uses of judgement document for calculating, and in index parameter identical parameters initial weight phase
Deng；Relevance score and corresponding complicated dynamic behaviour according to every judgement document in the default judgement documents' set of multiple is more
Every amendment relevance score of judgement document in individual default judgement document's set, and determine each default judgement document's collection
The judgement document of the first forward predetermined number of relevance score sequence is corrected in conjunction；Collected according to each default judgement document
The judgement document for closing corresponding first predetermined number sorts according to the sequence and benchmark of amendment relevance score, corrects each
The initial weight of parameter；And respectively using the initial weight of revised parameters as parameters weight, phase
The revised weight of same parameter is equal.
Further, the judgement document of corresponding first predetermined number is gathered according to amendment according to each default judgement document
The sequence of relevance score and benchmark sort, and the initial weight for correcting parameters includes：Calculate each default judge's text
The judgement document that corresponding first predetermined number is closed in book fair is different from benchmark sequence according to the sequence of amendment relevance score
Rate value, obtains multiple rate values；Judge whether multiple rate values are respectively less than predetermined threshold value；Judging multiple ratios
When there is the rate value more than predetermined threshold value in value, the initial weight of parameters is corrected；And judging multiple ratios
When rate value is respectively less than predetermined threshold value, terminate the amendment to the initial weight of parameters.
To achieve these goals, according to the another aspect of the application, there is provided a kind of judgement document retrieves device.Should
Device includes：Acquiring unit, for obtaining the judgement document matched with term set；First computing unit, is used for
Every the relevance score and complexity of judgement document in judgement document's set are calculated, wherein, relevance score is used for table
Show the matching degree of judgement document and term, complexity is used to represent the complexity of judgement document；Second computing unit,
For respectively according to every amendment correlation of judgement document point in relevance score and complicated dynamic behaviour judgement document set
Value, obtains multiple amendment relevance scores corresponding with many judgement documents in judgement document's set respectively；Sequencing unit,
It is ranked up for the judgement document in being gathered judgement document according to multiple amendment relevance score, obtains ranking results；
And display unit, for showing the judgement document in judgement document's set according to ranking results.
Further, judgement document's set includes the first judgement document, and the first computing unit includes：Acquisition module,
Index parameter for obtaining the first judgement document, wherein, index parameter includes at least one of following parameter：First
The length of judgement document, the lawsuit amount of money for being applicable law number and the first judgement document of the first judgement document；And first
Computing module, the complexity for calculating the first judgement document according to index parameter.
Further, the first computing module includes：Acquisition submodule, obtains the weight of parameters in index parameter；
And calculating sub module, judge for the value according to parameters in index parameter and the weight calculation of parameters first
The complexity of document.
Further, the device also includes：3rd computing unit, it is every during multiple default judgement documents gather for calculating
The relevance score of piece judgement document, wherein, the default judgement documents' collection of multiple is combined into presets training words with multiple respectively
The multiple judgement documents set matched somebody with somebody；4th computing unit, for calculating every judge during multiple default judgement documents gather
The complexity of document, wherein, the initial weight of value and parameters according to parameters in index parameter calculates complicated
Degree, the every index parameter that judgement document uses correspondence for participating in calculating is consistent, and identical parameters in index parameter
Initial weight is equal；5th computing unit, for according to every phase of judgement document in the default judgement document's set of multiple
Every amendment correlation of judgement document point in closing property score value and the multiple default judgement document's set of corresponding complicated dynamic behaviour
Value, and determine that each default judgement document corrects the judge of the first forward predetermined number of relevance score sequence in gathering
Document；Amending unit, the judgement document for gathering corresponding first predetermined number according to each default judgement document presses
Sequence and benchmark sequence according to amendment relevance score, correct the initial weight of parameters；And determining unit, use
In respectively using the initial weight of revised parameters as parameters weight, the revised weight of identical parameters
It is equal.
Further, amending unit includes：Second computing module, for calculating each default judgement document's set correspondence
The first predetermined number judgement document according to the sequence rate value different from benchmark sequence of amendment relevance score, obtain
To multiple rate values；Judge module, for judging whether multiple rate values are respectively less than predetermined threshold value；And correcting module,
For when there is the rate value more than predetermined threshold value in judging multiple rate values, correcting the initial weight of parameters,
When multiple rate value respectively less than predetermined threshold values are judged, terminate the amendment to the initial weight of parameters.
The application is gathered by obtaining the judgement document matched with term；Calculate every judge's text in judgement document's set
The relevance score and complexity of book, wherein, relevance score is used to represent the matching degree of judgement document and term,
Complexity is used to represent the complexity of judgement document；Collected according to relevance score and complicated dynamic behaviour judgement document respectively
Every amendment relevance score of judgement document in conjunction, obtains corresponding with many judgement documents in judgement document's set respectively
Multiple amendment relevance scores；Judgement document in being gathered judgement document according to multiple amendment relevance score is carried out
Sequence, obtains ranking results；And the judgement document in judgement document's set is shown according to ranking results, solve phase
Accuracy relatively low problem when retrieval result in the technology of pass to judgement document is ranked up, and then raising has been reached to cutting out
Sentence the effect of the accuracy of the retrieval result sequence of document.
Brief description of the drawings
The accompanying drawing for constituting the part of the application is used for providing further understanding of the present application, the schematic reality of the application
Apply example and its illustrate for explaining the application, do not constitute the improper restriction to the application.In the accompanying drawings：
Fig. 1 is the flow chart of the judgement document's search method according to the embodiment of the present application；
Fig. 2 is the flow chart that weight is set according to preset rules according to the embodiment of the present application；And
Fig. 3 is the schematic diagram that device is retrieved according to the judgement document of the embodiment of the present application.
Specific embodiment
It should be noted that in the case where not conflicting, the feature in embodiment and embodiment in the application can phase
Mutually combination.Describe the application in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
In order that those skilled in the art more fully understand application scheme, below in conjunction with the embodiment of the present application
Accompanying drawing, is clearly and completely described to the technical scheme in the embodiment of the present application, it is clear that described embodiment
The only embodiment of the application part, rather than whole embodiments.Based on the embodiment in the application, ability
The every other embodiment that domain those of ordinary skill is obtained under the premise of creative work is not made, should all belong to
The scope of the application protection.
It should be noted that term " first ", " in the description and claims of this application and above-mentioned accompanying drawing
Two " it is etc. for distinguishing similar object, without for describing specific order or precedence.It should be appreciated that this
The data that sample is used can be exchanged in the appropriate case, so as to embodiments herein described herein.Additionally, term
" comprising " and " having " and their any deformation, it is intended that covering is non-exclusive to be included, for example, comprising
The process of series of steps or unit, method, system, product or equipment are not necessarily limited to those steps clearly listed
Rapid or unit, but may include not listing clearly or intrinsic for these processes, method, product or equipment
Other steps or unit.
According to the embodiment of the present application, there is provided a kind of judgement document's search method, Fig. 1 is according to the embodiment of the present application
The flow chart of judgement document's search method.As shown in figure 1, the method includes steps S102 to step S110：
Step S102：Judgement document's set that acquisition is matched with term.
The term of the embodiment of the present application is the keyword of the retrieval for judgement document, and term can be one,
Can be multiple.Carry out document retrieve when, generally first receive user input term, and according to the term from
The judgement document that acquisition is matched with the term in judgement document's database.It is alternatively possible to obtain comprising the term
Judgement document, or obtain comprising the term partial content judgement document, for example, term for close
With invalid, then can obtain the composition judgement document set of the judgement document comprising avoidance of contract, or acquisition is included
The judgement document of avoidance of contract gathers with while collectively constituting judgement document comprising contract and invalid judgement document.
Step S104：Every the relevance score and complexity of judgement document in judgement document's set are calculated, wherein, phase
Closing property score value is used to represent the matching degree of judgement document and term, and complexity is used to represent the complexity of judgement document.
The relevance score of the embodiment of the present application is used to weigh the matching degree of judgement document and term, and relevance score is got over
Greatly, then explanation judgement document is higher with the matching degree of term, and relevance score is smaller, then illustrate judgement document with inspection
The matching degree of rope word is lower.It is alternatively possible to by count term occurs in judgement document number of times and integrity degree come
Relevance score is calculated, wherein, the number of times that term occurs in judgement document is more and integrity degree is higher, then the judge
The relevance score of document is higher.
Generally, those argue fully, the case that has a major impact has reference higher and grinds for a user
Study carefully value, therefore, the embodiment of the present application has considered every phase of judgement document when judgement document's retrieval is carried out
Closing property score value and complexity return to retrieval result, improving the specific aim of the judgement document of return.The embodiment of the present application
Complexity be used to weigh the complexity of judgement document, for example, the duplication degree and significance level of judgement document's merit,
Specifically, the complexity of judgement document can be weighed by some index parameters, for example, the case-involving amount of money, law of application
Bar number, judgement document's length etc..In actual conditions, applicable index parameter can be according to demand selected, for example, can
The complexity of judgement document is calculated to select some index parameter, it is also possible to which the multiple index parameters of selection are counted jointly
Calculate the complexity of judgement document.
Step S106：Repaiied according to every judgement document in relevance score and complicated dynamic behaviour judgement document set respectively
Positive correlation score value, obtains multiple amendment relevance scores corresponding with many judgement documents in judgement document's set respectively.
For example, the relevance score of judgement document can be multiplied with complexity obtains the amendment correlation of the judgement document
Score value.In the embodiment of the present application, every judgement document both corresponds to an amendment correlation point in judgement document's set
Value, wherein, its relevance score and complicated dynamic behaviour are repaiied according to the corresponding amendment relevance score of judgement document
Positive correlation score value.
Step S108：Judgement document in being gathered judgement document according to multiple amendment relevance score is ranked up, and obtains
To ranking results.
Step S110：The judgement document in judgement document's set is shown according to ranking results.
Specifically, can will correct the big judgement document of relevance score to include the anterior locations in Search Results, from
And it is easy to user intuitively to check the judgement document that those merits are complex, reference is larger.
The embodiment of the present application is gathered by obtaining the judgement document matched with term；Calculate every in judgement document's set
The relevance score and complexity of judgement document, wherein, relevance score be used to representing judgement document and term
With degree, complexity is used to represent the complexity of judgement document；Judged according to relevance score and complicated dynamic behaviour respectively
Every amendment relevance score of judgement document in document set, obtains literary with many judges in judgement document's set respectively
The corresponding multiple amendment relevance scores of book；Judge's text in being gathered judgement document according to multiple amendment relevance score
Book is ranked up, and obtains ranking results；And show the judgement document in judgement document's set according to ranking results.This
Application embodiment is modified by the complexity of judgement document to the relevance score of judgement document, and according to amendment phase
Closing property score value is ranked up display to retrieval result, compared to being carried out to retrieval result according only to correlation in the prior art
Sequence, more agrees with the Search Requirement of user, when solving the retrieval result in correlation technique to judgement document and being ranked up
The relatively low problem of accuracy, and then reached the effect for improving the accuracy to the retrieval result sequence of judgement document.
Preferably, judgement document's set includes the first judgement document, calculates every judgement document in judgement document's set
Complexity include：The index parameter of the first judgement document is obtained, wherein, index parameter includes following parameter at least
One of：The length of the first judgement document, the lawsuit amount of money for being applicable law number and the first judgement document of the first judgement document；
And the complexity of the first judgement document is calculated according to index parameter.
First judgement document can be any one judgement document in judgement document's set.The length of the first judgement document can
The length of part is discussed with the length transcript, or first judgement document's legal principle that are the first judgement document, specifically,
Its length can be used as by counting number of words.The applicable law number of the first judgement document can be extracted by presupposed information and advised
The law sequence number then extracted in judgement document is counted, it is alternatively possible to by the identical law sequence number of different laws and
The different law sequence numbers of identical law are counted.During the lawsuit amount of money of the first judgement document refers to the first judgement document
The expense being related to, for example, legitimation fee, damages, fine etc..Alternatively, subtract if being related to expense in litigation
Half situation, extracts before the amount of money should be restored to and halve.For example, the partial content of a judgement document is as follows：
" think that plaintiff shuts out, and certain is married at a registry nearly 7 years with defendant Liu, and certain goodwill as between spouses is established after marriage in the court.
Though plaintiff advocates that defendant runs away from home but do not prove with evidence, not approve.Plaintiff sues for a divorce and does not possess legal condition,
It is unassisted.According to《Marriage Law of the People's Republic of China》32nd article,《Civil Procedure Law of The People's Republic 0f China》
144th article of regulation, adjudicates as follows：
Inaccurate plaintiff shuts out, and certain divorces with defendant Liu.
300 yuan of legitimation fee, by plaintiff.
Such as refuse to obey this judgement, petition for appeal can be submitted to the court in 15 days from court verdict is sent to, and work as by other side
The number of thing people proposes copy, appeals in intermediate people's court of Weifang City of Shandong Province.”
For this judgement document, its length is 216, and it is 2 (i.e. to be applicable law number《The People's Republic of China (PRC) weds
Relation by marriage method》32nd article and《Civil Procedure Law of The People's Republic 0f China》144th article), the lawsuit amount of money is
300。
Preferably, the complexity for calculating the first judgement document according to index parameter includes：Obtain each ginseng in index parameter
Several weights；And value according to parameters in index parameter and the judgement document of weight calculation first of parameters
Complexity.
Specifically, parameters can be the different index parameter of any two in the index parameter of the embodiment of the present application,
For example, the different index parameter being applicable more than law number and the lawsuit amount of money, or any two, for example, suitable
With law number, the lawsuit amount of money and document length.The value of parameters refer to for specific judgement document, each
The value of parameter, for example, for judgement document 1, the value that its is applicable law number is A1, and the value of its lawsuit amount of money is
B1, for judgement document 2, the value that its is applicable law number is A2, and the value of its lawsuit amount of money is B2.Parameters
Weight can be rule of thumb default value, or according to the preset rules value that obtains of training.
It is for instance possible to use equation below calculates the complexity of the first judgement document：
C=ln (pLL+1) ln (pMM+1) ln (pNN+1), wherein, C represents the complexity of the first judgement document,
L, M, N represent the length of the first judgement document, the applicable law number of the first judgement document and the first judgement document respectively
The lawsuit amount of money, pL, pM, pN represent the length of the first judgement document, the applicable law of the first judgement document respectively
The weight of the lawsuit amount of money of number and the first judgement document.
Preferably, in order to further improve the accuracy that retrieval result sorts, the application is judged in reception term
Before document retrieval, weight can also be set according to following preset rules：Calculate every in multiple default judgement document's set
The relevance score of judgement document, wherein, the default judgement document's collection of multiple is combined into and is matched with the default training word of multiple respectively
Multiple judgement documents set；Every complexity of judgement document in multiple default judgement document's set is calculated, wherein,
The initial weight computation complexity of value and parameters according to parameters in index parameter, every sanction for participating in calculating
Sentence that the index parameter correspondence that document uses is consistent, and in index parameter identical parameters initial weight it is equal；According to many
The relevance score and corresponding complicated dynamic behaviour multiple of every judgement document preset judge in individual default judgement document's set
Every amendment relevance score of judgement document in document set, and determine to correct phase in each default judgement document's set
The judgement document of the first forward predetermined number of closing property score value sequence；According to each default judgement document set corresponding the
The judgement document of one predetermined number sorts according to the sequence and benchmark of amendment relevance score, corrects the initial of parameters
Weight；And respectively using the initial weight of revised parameters as parameters weight, identical parameters amendment
Weight afterwards is equal.
Default training word is the keyword for above-mentioned weight training, and the embodiment of the present application pre-sets one group for weight
The keyword of training, and the corresponding default judgement document set of each default training word is obtained respectively, wherein, preset and cut out
Each judgement document is the judgement document that corresponding default training word is matched in sentencing document set, for example, one
The default training word of group includes default training word a and default training word b, the judge that acquisition is matched with default training word a
The default judgement document set a of document composition, the default judge's text of judgement document's composition that acquisition is matched with default training word b
B is closed in book fair.
After multiple default judgement document's set are obtained, every judgement document in multiple default judgement document's set is calculated
Relevance score and complexity.Specifically, the computational methods of the relevance score of judgement document and complexity ibid,
Will not be repeated here, it is necessary to illustrate, for different judgement documents, the index parameter of selection is identical, identical finger
The initial weight for marking parameter is also identical, for example, for judgement document 1, the index parameter that it is chosen is to be applicable law number
With the lawsuit amount of money, wherein, be applicable the initial weight of law number for x, the initial weight of the lawsuit amount of money is y, then for cutting out
Sentence document 2, the index parameter that it is chosen is applicable the initial power of law number also to be applicable law number and the lawsuit amount of money
Weight is x, and the initial weight of the lawsuit amount of money is y, and the initial weight of parameters can be interval interior any one (0,1)
Individual numerical value.
After every the relevance score and complexity of judgement document in obtaining each default judgement document's set, calculate
Every amendment relevance score of judgement document, for example, the relevance score of judgement document can be multiplied with complexity
Obtain its amendment relevance score.Every amendment correlation of judgement document in multiple default judgement document's set are obtained
After score value, determine to correct the first forward predetermined number of relevance score sequence in each default judgement document's set
Judgement document.
For example, there are 100 judgement documents in default judgement document's set a, exist in default judgement document's set b
130 judgement documents, respectively according to 100 amendment relevance scores of judgement document in default judgement document's set a
It is ranked up and is ranked up according to the amendment relevance score of 130 judgement documents in default judgement document's set b,
And 10 judgement documents for sorting forward in default judgement document's set a and default judgement document's set b are determined respectively.
The benchmark of the embodiment of the present application is ordered as the sequence of user preset, above-mentioned according to amendment relevance score for weighing
The accuracy of the ranking results of sequence, specifically, can be sorted and base platoon by counting according to amendment relevance score
Sequence identical ratio or different ratios determine the accuracy of ranking results sorted according to amendment relevance score.
For example, 10 judge's texts that the amendment relevance score that default training word " avoidance of contract " is returned sorts forward
Book A1 to A10, A1 to A10 are ordered as according to amendment relevance score：
A1>A2>A3>A4>A5>A6>A7>A8>A9>The benchmark of A10, A1 to A10 is ordered as
A2>A3>A1>A5>A6>A4>A7>A8>A10>A9.According to arrangement principle, 10 judgement document A1 to A10
In the presence of 45 kinds of sortords, for example, A2>A3, A8>A10 etc., as seen from the above, A1 to A10 is according to repairing
There are 41 identicals sequences during positive correlation score value sorts and benchmark sorts, there are 4 different sequences, i.e. A2
It is different with the sequence of A4 with A1, A5 with A4, A6 with A1, A3, therefore, sorted according to amendment relevance score
It is 91% with benchmark sequence identical ratio, different ratios is 9%.
Preferably, the judgement document of corresponding first predetermined number is gathered according to each default judgement document according to amendment phase
The sequence and benchmark sequence of closing property score value, the initial weight for correcting parameters include：Calculate each default judgement document
Gather sequence and benchmark sequence different ratios of the judgement document of corresponding first predetermined number according to amendment relevance score
Rate value, obtains multiple rate values；Judge whether multiple rate values are respectively less than predetermined threshold value；Judging multiple rate values
When middle presence is more than the rate value of predetermined threshold value, the initial weight of parameters is corrected；And judging multiple ratios
When value is respectively less than predetermined threshold value, terminate the amendment to the initial weight of parameters.
Count each default judgement document respectively in the manner described above and gather the judgement document of corresponding first predetermined number and press
According to the sequence for the correcting relevance score rate value different with benchmark sequence, and it is compared with predetermined threshold value, wherein,
Predetermined threshold value can be configured according to actual conditions, for example, 5%, 8% etc..
For example, predetermined threshold value is 5%, 10 forward according to relevance ranking is corrected in default judgement document's set a
Judgement document A1 to A10, A1 to A10 are different with the sequence of its benchmark according to the ranking results of amendment relevance ranking
Ratio is 12%, i.e. rate value a is 12%；It is forward according to amendment relevance ranking in default judgement document's set b
10 judgement documents B1 to B10, B1 to B10 sort not according to the ranking results and its benchmark of amendment relevance ranking
With ratio for 20%, i.e. rate value b be 20%, it can thus be seen that rate value a and rate value b be all higher than it is pre-
If threshold value, now need to be modified the initial weight of above-mentioned parameters, for example, amendment above-mentioned formula in pL,
Tri- initial weights of pM, pN.
Specifically, the embodiment of the present application is after the initial weight to above-mentioned parameters is modified, by after amendment
The initial weight of parameters recalculate every complexity of judgement document, weight in multiple default judgement documents' set
Above-mentioned steps are performed again, until each default judgement document gathers the judgement document of corresponding first predetermined number according to repairing
The sequence of the positive correlation score value rate value different from benchmark sequence is respectively less than predetermined threshold value, now, will finally correct
The initial weight of the parameters for arriving as parameters weight.
It should be noted that the amendment to the initial weight of above-mentioned parameters can adjust each according to default step-length
The value of the initial weight of parameter, or manually adjust parameters initial weight value.
Fig. 2 is the flow chart that weight is set according to preset rules according to the embodiment of the present application.As shown in Fig. 2 according to
Preset rules set weight and comprise the following steps：
Step S202：Obtain one group of judgement document.
It is alternatively possible to above-mentioned judgement document is stored to internal memory or hard disc data storehouse or data by its unique mark correspondence
In structure.
Step S204：This group of judgement document is traveled through, every length transcript L of judgement document is extracted, is applicable law number N
With lawsuit amount of money M.
Specifically, this group of judgement document is traveled through, following operation is performed to every judgement document：One, extract this document
Length transcript, be designated as L；Two, the judge for extracting this document is applicable law, and record is applicable the total N of law,
Wherein, multiple laws of same portion's law are considered as a plurality of；Three, the lawsuit amount of money M of this document is extracted, wherein, relate to
And to expense in litigation halve situation when, the amount of money for being extracted should be restored to the level before halving.It should be noted that can
Stored into its field belonging to corresponding judgement document's record with by three index parameters of said extracted, corresponding to.
Step S206：Default three initial parameters pL, pN, pM are obtained, the complexity of current judgement document is calculated
Index C.
Specifically, complexity index C is complexity, in the ergodic process of step S204, by previously given three
Individual initial parameter (i.e. initial weight) pL, pN, pM, calculate the complexity index C of current judgement document, calculate
Formula is as follows：
C=ln (pLL+1) ln (pMM+1) ln (pNN+1), wherein, L, M, N are represented when lead official text respectively
The length of book, the lawsuit amount of money for being applicable law number and current judgement document of current judgement document, pL, pM, pN point
Length, the lawsuit amount of money for being applicable law number and current judgement document of current judgement document of current judgement document are not represented
Initial weight.After the complexity index C for obtaining current judgement document, also correspond to and store to current judgement document
In field belonging to record, wherein, current judgement document is current processed judgement document.
Step S208：Given one group of default training word, obtains the relevance score of its matching document, by its correlation point
Value is multiplied with its complexity index C, obtains modifying relevance score, is obtaining the modification correlation of all matching documents
After score value, according to modification relevance score from high to low, preceding 10 matchings document is pushed.
Specifically, illustrated by taking a default training word as an example below, for example, for default training word " contract without
Effect ", can carry out full-text search to the judgement document for storing, and obtain its matching document and (matched with default training word
Judgement document) relevance score, wherein, match document in preset training word occur number of times it is more, more complete,
Its relevance score is higher.To the corresponding every matching document of default training word, its relevance score is matched into text with this
The complexity index C of book is multiplied, the relevance score after being modified, and is designated as modification relevance score and (corrects phase
Closing property score value).
Above-mentioned treatment is performed in couple whole judgement documents matched with default training word, the modification of all matching documents is obtained
After relevance score, all matching documents are sorted from high to low according to modification relevance score, push wherein preceding 10
Piece matching document.Each the default training word preset to the group in training word is performed both by above-mentioned treatment, such that it is able to obtain
Each corresponding first 10 of default training word matches document.
Step S210：Document is matched to each corresponding first 10 of default training word, is ranked up by manual type,
And calculate the manual type sequence ratio accu% different from according to relevance score sequence is modified.
Specifically, document is matched for each corresponding first 10 of default training word in step S208, by artificial side
Formula is ranked up (i.e. benchmark sequence), for example, the amendment relevance score that default training word " avoidance of contract " is returned
Sequence forward 10 judgement documents A1 to A10, A1 to A10 are ordered as according to amendment relevance score：
A1>A2>A3>A4>A5>A6>A7>A8>A9>A10, A1 to A10 according to manual type be ranked up for
A2>A3>A1>A5>A6>A4>A7>A8>A10>A9。
Specifically, to whole corresponding preceding 10 matchings documents of default training word, calculate respectively manual type sequence with
According to the different ratio accu% of modification relevance score sequence.
Step S212：With the object function that is minimised as of accu%, gradient adjust three parameters pL, pN, pM just
Initial value, obtains three optimal solution pL ', pN ', the pM ' of parameter.
Wherein, the optimal solution pL ' of above three parameter, pN ', pM ' are the corresponding weight of These parameters parameter.
Specifically, the corresponding accu% of each default training word can be compared with predetermined threshold value respectively, if each is pre-
If the corresponding accu% of training word is respectively less than predetermined threshold value, then without be modified to pL, pN, pM, current pL,
PN, pM are optimal solution, if there is the accu% more than predetermined threshold value in the corresponding accu% of each default training word,
Three parameters pL, pN, pM are then corrected, and every matching document is recalculated by revised pL, pN, pM
Complexity, that is, repeat above-mentioned steps S206 to step S212 until the corresponding accu% of each default training word
Respectively less than predetermined threshold value, now corresponding pL, pN, pM are optimal solution.
Step S214：According to pL ', pN ', pM ', whole judgement documents are traveled through again and its complexity index is calculated.
It should be noted that during user search, relevance score is multiplied with new complexity index, obtain
New amendment relevance score, and matching result is ranked up and represented according to the score value.
The embodiment of the present application is a kind of sortord of fusion judgement document's importance (i.e. complexity), and which is being passed
On the basis of system full-text search relevance ranking, the index for weighing judgement document's importance is added (for example, length, relating to
The case amount of money, it is applicable law quantity etc.), and combine and manually judge and ask the mode of object function optimal solution to adjust relevant parameter
Value, in the hope of optimum efficiency so that the ranking results of judgement document more agree with the Search Requirement of user.
It should be noted that can be in such as one group computer executable instructions the step of the flow of accompanying drawing is illustrated
Performed in computer system, and, although logical order is shown in flow charts, but in some cases, can
Shown or described step is performed with different from order herein.
According to the another aspect of the embodiment of the present application, there is provided a kind of judgement document retrieves device, judgement document's device
Can be used for performing judgement document's search method of the embodiment of the present application, judgement document's search method of the embodiment of the present application
Can also retrieve device by the judgement document of the embodiment of the present application to perform.
Fig. 3 is the schematic diagram that device is retrieved according to the judgement document of the embodiment of the present application, as shown in figure 3, the device bag
Include：Acquiring unit 10, the first computing unit 20, the second computing unit 30, sequencing unit 40 and display unit 50.
Acquiring unit 10, for obtaining the judgement document matched with term set.
The term of the embodiment of the present application is the keyword of the retrieval for judgement document, and term can be one,
Can be multiple.Carry out document retrieve when, generally first receive user input term, and according to the term from
The judgement document that acquisition is matched with the term in judgement document's database.It is alternatively possible to obtain comprising the term
Judgement document, or obtain comprising the term partial content judgement document, for example, term for close
With invalid, then can obtain the composition judgement document set of the judgement document comprising avoidance of contract, or acquisition is included
The judgement document of avoidance of contract gathers with while collectively constituting judgement document comprising contract and invalid judgement document.
First computing unit 20, for calculating every the relevance score and complexity of judgement document during judgement document gathers
Degree, wherein, relevance score is used to represent the matching degree of judgement document and term, and complexity is used to represent judge's text
The complexity of book.
The relevance score of the embodiment of the present application is used to weigh the matching degree of judgement document and term, and relevance score is got over
Greatly, then explanation judgement document is higher with the matching degree of term, and relevance score is smaller, then illustrate judgement document with inspection
The matching degree of rope word is lower.The complexity of the embodiment of the present application is used to weigh the complexity of judgement document, for example, cutting out
Sentence the duplication degree and significance level of document merit, specifically, judgement document can be weighed by some index parameters
Complexity, for example, the case-involving amount of money, being applicable law number, judgement document's length etc..
Second computing unit 30, for respectively according to every sanction in relevance score and complicated dynamic behaviour judgement document set
Sentence the amendment relevance score of document, obtain multiple amendments corresponding with many judgement documents in judgement document's set respectively
Relevance score.
Sequencing unit 40, arranges for the judgement document in being gathered judgement document according to multiple amendment relevance score
Sequence, obtains ranking results.
Display unit 50, for showing the judgement document in judgement document's set according to ranking results.
The embodiment of the present application is obtained the judgement document matched with term by acquiring unit 10 and gathered；First computing unit
20 calculate every the relevance score and complexity of judgement document in judgement document's set, wherein, relevance score is used for
The matching degree of judgement document and term is represented, complexity is used to represent the complexity of judgement document；Second calculates single
Unit 30 is respectively according to every amendment correlation of judgement document in relevance score and complicated dynamic behaviour judgement document set
Score value, obtains multiple amendment relevance scores corresponding with many judgement documents in judgement document's set respectively；Sequence is single
Judgement document during unit 40 gathers judgement document according to multiple amendment relevance score is ranked up, and obtains sequence knot
Really；And display unit 50 shows the judgement document in judgement document's set according to ranking results.The embodiment of the present application is led to
The complexity for crossing judgement document is modified to the relevance score of judgement document, and according to amendment relevance score to inspection
Hitch fruit is ranked up display, compared to being ranked up to retrieval result according only to correlation in the prior art, more agrees with
The Search Requirement of user, solves accuracy when the retrieval result in correlation technique to judgement document is ranked up relatively low
Problem, and then reached the effect for improving the accuracy to the retrieval result sequence of judgement document.
Preferably, judgement document's set includes the first judgement document, and the first computing unit 20 includes：Acquisition module,
Index parameter for obtaining the first judgement document, wherein, index parameter includes at least one of following parameter：First
The length of judgement document, the lawsuit amount of money for being applicable law number and the first judgement document of the first judgement document；And first
Computing module, the complexity for calculating the first judgement document according to index parameter.
Preferably, the first computing module includes：Acquisition submodule, obtains the weight of parameters in index parameter；With
And calculating sub module, judge text for the value according to parameters in index parameter and the weight calculation of parameters first
The complexity of book.
Further, the device also includes：3rd computing unit, it is every during multiple default judgement documents gather for calculating
The relevance score of piece judgement document, wherein, the default judgement documents' collection of multiple is combined into presets training words with multiple respectively
The multiple judgement documents set matched somebody with somebody；4th computing unit, for calculating every judge during multiple default judgement documents gather
The complexity of document, wherein, the initial weight of value and parameters according to parameters in index parameter calculates complicated
Degree, the every index parameter that judgement document uses correspondence for participating in calculating is consistent, and identical parameters in index parameter
Initial weight is equal；5th computing unit, for according to every phase of judgement document in the default judgement document's set of multiple
Every amendment correlation of judgement document point in closing property score value and the multiple default judgement document's set of corresponding complicated dynamic behaviour
Value, and determine that each default judgement document corrects the judge of the first forward predetermined number of relevance score sequence in gathering
Document；Amending unit, the judgement document for gathering corresponding first predetermined number according to each default judgement document presses
Sequence and benchmark sequence according to amendment relevance score, correct the initial weight of parameters；And determining unit, use
In respectively using the initial weight of revised parameters as parameters weight, the revised weight of identical parameters
It is equal.
Further, amending unit includes：Second computing module, for calculating each default judgement document's set correspondence
The first predetermined number judgement document according to the sequence rate value different from benchmark sequence of amendment relevance score, obtain
To multiple rate values；Judge module, for judging whether multiple rate values are respectively less than predetermined threshold value；And correcting module,
For when there is the rate value more than predetermined threshold value in judging multiple rate values, correcting the initial weight of parameters,
When multiple rate value respectively less than predetermined threshold values are judged, terminate the amendment to the initial weight of parameters.
Judgement document retrieval device includes processor and memory, above-mentioned acquiring unit, the first computing unit, the
Two computing units, sequencing unit and display unit etc. are stored in memory, by computing device as program unit
Said procedure unit in memory is stored to realize corresponding function.
Kernel is included in processor, is gone in memory to transfer corresponding program unit by kernel.Kernel can set one
Or more, the accurate retrieval to judgement document is completed by adjusting kernel parameter.
Memory potentially includes the volatile memory in computer-readable medium, random access memory (RAM) and/
Or the form, such as read-only storage (ROM) or flash memory (flash RAM) such as Nonvolatile memory, memory includes at least one
Individual storage chip.
Present invention also provides a kind of computer program product, when being performed on data processing equipment, it is adapted for carrying out just
The program code of beginningization there are as below methods step：Judgement document's set that acquisition is matched with term；Calculate the judge
The every relevance score and complexity of judgement document in document set, wherein, the relevance score is used to represent institute
The matching degree of judgement document and the term is stated, the complexity is used to represent the complexity of the judgement document；
Every amendment of judgement document in judgement document's set according to the relevance score and the complicated dynamic behaviour respectively
Relevance score, obtains multiple amendment correlations point corresponding with many judgement documents in judgement document set respectively
Value；Judgement document in being gathered the judgement document according to the multiple amendment relevance score is ranked up, and obtains
Ranking results；And show the judgement document in judgement document's set according to the ranking results.
Above-mentioned the embodiment of the present application sequence number is for illustration only, and the quality of embodiment is not represented.
In above-described embodiment of the application, the description to each embodiment all emphasizes particularly on different fields, and does not have in certain embodiment
The part of detailed description, may refer to the associated description of other embodiment.
In several embodiments provided herein, it should be understood that disclosed technology contents, can be by other
Mode realize.Wherein, device embodiment described above is only schematical, such as division of described unit,
Can be a kind of division of logic function, there can be other dividing mode when actually realizing, for example multiple units or component
Can combine or be desirably integrated into another system, or some features can be ignored, or do not perform.It is another, institute
Display or the coupling each other for discussing or direct-coupling or communication connection can be by some interfaces, unit or mould
The INDIRECT COUPLING of block or communication connection, can be electrical or other forms.
The unit that is illustrated as separating component can be or may not be it is physically separate, it is aobvious as unit
The part for showing can be or may not be physical location, you can with positioned at a place, or can also be distributed to
On multiple units.Some or all of unit therein can be according to the actual needs selected to realize this embodiment scheme
Purpose.
In addition, during each functional unit in the application each embodiment can be integrated in a processing unit, it is also possible to
It is that unit is individually physically present, it is also possible to which two or more units are integrated in a unit.It is above-mentioned integrated
Unit can both be realized in the form of hardware, it would however also be possible to employ the form of SFU software functional unit is realized.
If the integrated unit is to realize in the form of SFU software functional unit and as independent production marketing or use
When, can store in a computer read/write memory medium.Based on such understanding, the technical scheme of the application
The part for substantially being contributed to prior art in other words or all or part of the technical scheme can be produced with software
The form of product is embodied, and the computer software product is stored in a storage medium, including some instructions are used to make
Obtain a computer equipment (can be personal computer, server or network equipment etc.) and perform each implementation of the application
The all or part of step of example methods described.And foregoing storage medium includes：USB flash disk, read-only storage (ROM,
Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk,
Magnetic disc or CD etc. are various can be with the medium of store program codes.
The above is only the preferred embodiment of the application, it is noted that for the ordinary skill people of the art
For member, on the premise of the application principle is not departed from, some improvements and modifications can also be made, these improve and moisten
Decorations also should be regarded as the protection domain of the application.
Claims (10)
1. a kind of judgement document's search method, it is characterised in that including：
Judgement document's set that acquisition is matched with term；
Every the relevance score and complexity of judgement document in judgement document's set are calculated, wherein, it is described
Relevance score is used to represent the judgement document and the matching degree of the term that the complexity to be used to represent
The complexity of the judgement document；
Every judge is literary in judgement document's set according to the relevance score and the complicated dynamic behaviour respectively
The amendment relevance score of book, obtains multiple corresponding with many judgement documents in judgement document set respectively
Amendment relevance score；
Judgement document in being gathered the judgement document according to the multiple amendment relevance score is ranked up,
Obtain ranking results；And
The judgement document in judgement document's set is shown according to the ranking results.
2. method according to claim 1, it is characterised in that judgement document's set includes that the first judge is literary
Book, the complexity for calculating every judgement document in judgement document's set includes：
The index parameter of first judgement document is obtained, wherein, the index parameter includes following parameter extremely
It is one of few：The length of first judgement document, the applicable law number and described first of first judgement document
The lawsuit amount of money of judgement document；And
The complexity of first judgement document is calculated according to the index parameter.
3. method according to claim 2, it is characterised in that first judge is calculated according to the index parameter
The complexity of document includes：
Obtain the weight of parameters in the index parameter；And
The first judgement document according to the value of parameters in the index parameter and the weight calculation of parameters
Complexity.
4. method according to claim 3, it is characterised in that methods described also includes：According to following preset rules
Weight is set：
Every relevance score of judgement document in multiple default judgement document's set is calculated, wherein, it is the multiple
Default judgement document's collection is combined into the multiple judgement documents for being matched with the default training word of multiple respectively and gathers；
Every complexity of judgement document in the multiple default judgement document's set is calculated, wherein, according to described
The initial weight of the value of parameters and parameters calculates the complexity in index parameter, and every participates in calculating
The index parameter correspondence that uses of judgement document it is consistent, and in the index parameter identical parameters initial weight
It is equal；
According to every the relevance score and corresponding complexity of judgement document in the multiple default judgement document's set
Degree calculates every amendment relevance score of judgement document in the multiple default judgement document's set, and determines each
Judge's text of the first forward predetermined number of relevance score sequence is corrected in individual default judgement document's set
Book；
The judgement document of corresponding first predetermined number is gathered according to amendment phase according to each described default judgement document
The sequence and benchmark sequence of closing property score value, correct the initial weight of the parameters；And
Respectively using the initial weight of the revised parameters as the parameters weight, the phase
The revised weight of same parameter is equal.
5. method according to claim 4, it is characterised in that according to each default judgement document's set correspondence
The judgement document of the first predetermined number sorted according to the sequence and benchmark of amendment relevance score, correct described each
The initial weight of individual parameter includes：
Calculate each described default judgement document and gather the judgement document of corresponding first predetermined number according to repairing
The sequence of the positive correlation score value rate value different from benchmark sequence, obtains multiple rate values；
Judge whether the multiple rate value is respectively less than predetermined threshold value；
When there is the rate value more than the predetermined threshold value in judging the multiple rate value, correct described each
The initial weight of individual parameter；And
When judging that the multiple rate value is respectively less than the predetermined threshold value, terminate to the first of the parameters
The amendment of beginning weight.
6. a kind of judgement document retrieves device, it is characterised in that including：
Acquiring unit, for obtaining the judgement document matched with term set；
First computing unit, for calculate in judgement document set every relevance score of judgement document and
Complexity, wherein, the relevance score is used to represent the judgement document and the matching degree of the term,
The complexity is used to represent the complexity of the judgement document；
Second computing unit, for judging text according to the relevance score and the complicated dynamic behaviour respectively
Every amendment relevance score of judgement document in book fair conjunction, many in obtaining gathering with the judgement document respectively
The corresponding multiple amendment relevance scores of judgement document；
Sequencing unit, for according to the multiple judge corrected during relevance score is gathered the judgement document
Document is ranked up, and obtains ranking results；And
Display unit, for showing the judgement document in judgement document's set according to the ranking results.
7. device according to claim 6, it is characterised in that judgement document's set includes that the first judge is literary
Book, first computing unit includes：
Acquisition module, the index parameter for obtaining first judgement document, wherein, the index parameter bag
Include at least one of following parameter：The length of first judgement document, the law of application of first judgement document
Bar number and the lawsuit amount of money of first judgement document；And
First computing module, the complexity for calculating first judgement document according to the index parameter.
8. device according to claim 7, it is characterised in that first computing module includes：
Acquisition submodule, obtains the weight of parameters in the index parameter；And
Calculating sub module, for the weight calculation according to the value of parameters and parameters in the index parameter
The complexity of first judgement document.
9. device according to claim 8, it is characterised in that described device also includes：
3rd computing unit, for calculating every correlation of judgement document point during multiple default judgement documents gather
Value, wherein, the multiple default judgement document's collection is combined into the multiple judges for being matched with the default training word of multiple respectively
Document set；
4th computing unit, for calculating every complexity of judgement document during the multiple default judgement document gathers
Degree, wherein, the initial weight according to the value of parameters and parameters in the index parameter calculates described multiple
Miscellaneous degree, the every index parameter that judgement document uses correspondence for participating in calculating is consistent, and in the index parameter
The initial weight of identical parameters is equal；
5th computing unit, for according to every correlation of judgement document in the multiple default judgement document's set
Property score value and the multiple default judgement document set of corresponding complicated dynamic behaviour in every amendment phase of judgement document
Closing property score value, and determine that each described default judgement document corrects forward first of relevance score sequence in gathering
The judgement document of predetermined number；
Amending unit, the judge for gathering corresponding first predetermined number according to each described default judgement document
Document sorts according to the sequence and benchmark of amendment relevance score, corrects the initial weight of the parameters；With
And
Determining unit, for respectively using the initial weight of the revised parameters as the parameters
Weight, the revised weight of identical parameters is equal.
10. device according to claim 9, it is characterised in that the amending unit includes：
Second computing module, corresponding first present count is gathered for calculating each described default judgement document
The judgement document of amount obtains many according to the sequence of the amendment relevance score rate value different from benchmark sequence
Individual rate value；
Judge module, for judging whether the multiple rate value is respectively less than predetermined threshold value；And
Correcting module, for there is the rate value more than the predetermined threshold value in the multiple rate value is judged
When, the initial weight of the parameters is corrected, judging that the multiple rate value is respectively less than the default threshold
During value, terminate the amendment to the initial weight of the parameters.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510869926.9A CN106815266B (en) | 2015-12-01 | 2015-12-01 | Referee document retrieval method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510869926.9A CN106815266B (en) | 2015-12-01 | 2015-12-01 | Referee document retrieval method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106815266A true CN106815266A (en) | 2017-06-09 |
| CN106815266B CN106815266B (en) | 2020-06-16 |
Family
ID=59107655
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510869926.9A Active CN106815266B (en) | 2015-12-01 | 2015-12-01 | Referee document retrieval method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106815266B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108573057A (en) * | 2018-04-25 | 2018-09-25 | 王慧 | A kind of legal documents and laws and regulations correspondence search method |
| CN109299382A (en) * | 2018-11-01 | 2019-02-01 | 厦门市美亚柏科信息股份有限公司 | A kind of recommended method of lteral data, system and computer storage medium |
| CN109388796A (en) * | 2017-08-11 | 2019-02-26 | 北京国双科技有限公司 | The method for pushing and device of judgement document |
| CN109446511A (en) * | 2018-09-10 | 2019-03-08 | 平安科技（深圳）有限公司 | Judgement document's processing method, device, computer equipment and storage medium |
| CN110019659A (en) * | 2017-07-31 | 2019-07-16 | 北京国双科技有限公司 | The search method and device of judgement document |
| CN113297438A (en) * | 2021-05-21 | 2021-08-24 | 深圳市智尊宝数据开发有限公司 | Information retrieval method, electronic equipment and related products |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1172994A (en) * | 1996-05-29 | 1998-02-11 | 松下电器产业株式会社 | Document retrieval system |
| WO2001080087A1 (en) * | 2000-04-14 | 2001-10-25 | Rightnow Technologies, Inc. | Temporal updates of relevancy rating of retrieved information in an information search system |
| CN1517914A (en) * | 2003-01-06 | 2004-08-04 | Searching of structural file | |
| CN103793418A (en) * | 2012-10-31 | 2014-05-14 | 珠海富讯网络科技有限公司 | Search method of real-time vertical search engine for security industry |
-
2015
- 2015-12-01 CN CN201510869926.9A patent/CN106815266B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1172994A (en) * | 1996-05-29 | 1998-02-11 | 松下电器产业株式会社 | Document retrieval system |
| WO2001080087A1 (en) * | 2000-04-14 | 2001-10-25 | Rightnow Technologies, Inc. | Temporal updates of relevancy rating of retrieved information in an information search system |
| CN1517914A (en) * | 2003-01-06 | 2004-08-04 | Searching of structural file | |
| CN103793418A (en) * | 2012-10-31 | 2014-05-14 | 珠海富讯网络科技有限公司 | Search method of real-time vertical search engine for security industry |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110019659A (en) * | 2017-07-31 | 2019-07-16 | 北京国双科技有限公司 | The search method and device of judgement document |
| CN109388796A (en) * | 2017-08-11 | 2019-02-26 | 北京国双科技有限公司 | The method for pushing and device of judgement document |
| CN108573057A (en) * | 2018-04-25 | 2018-09-25 | 王慧 | A kind of legal documents and laws and regulations correspondence search method |
| CN109446511A (en) * | 2018-09-10 | 2019-03-08 | 平安科技（深圳）有限公司 | Judgement document's processing method, device, computer equipment and storage medium |
| CN109446511B (en) * | 2018-09-10 | 2022-07-08 | 平安科技（深圳）有限公司 | Referee document processing method, referee document processing device, computer equipment and storage medium |
| CN109299382A (en) * | 2018-11-01 | 2019-02-01 | 厦门市美亚柏科信息股份有限公司 | A kind of recommended method of lteral data, system and computer storage medium |
| CN109299382B (en) * | 2018-11-01 | 2021-08-10 | 厦门市美亚柏科信息股份有限公司 | Recommendation method and system for character data and computer storage medium |
| CN113297438A (en) * | 2021-05-21 | 2021-08-24 | 深圳市智尊宝数据开发有限公司 | Information retrieval method, electronic equipment and related products |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106815266B (en) | 2020-06-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106815266A (en) | Judgement document's search method and device | |
| CN102841946B (en) | Commodity data retrieval ordering and Method of Commodity Recommendation and system | |
| CN104111933B (en) | Obtain business object label, set up the method and device of training pattern | |
| CN110019658B (en) | Method and related device for generating search term | |
| US11514242B2 (en) | Method for automatically summarizing internet web page and text information | |
| WO2017092622A1 (en) | Legal provision search method and device | |
| CN110737859B (en) | UP master matching method and device | |
| CN104199833B (en) | The clustering method and clustering apparatus of a kind of network search words | |
| CN108304437A (en) | A kind of automatic question-answering method, device and storage medium | |
| CN106446071B (en) | Information processing apparatus and method | |
| CN106445963B (en) | Advertisement index keyword automatic generation method and device of APP platform | |
| CN105260362A (en) | New word extraction method and device | |
| CN107798600A (en) | The credit risk recognition methods of the small micro- loan of internet finance and device | |
| CN105260359A (en) | Semantic keyword extraction method and apparatus | |
| CN104298715B (en) | A kind of more indexed results ordering by merging methods based on TF IDF | |
| CN105205090A (en) | Web page text classification algorithm research based on web page link analysis and support vector machine | |
| CN107247743A (en) | A kind of judicial class case search method and system | |
| CN108920488A (en) | The natural language processing method and device that multisystem combines | |
| CN108520038B (en) | Biomedical literature retrieval method based on sequencing learning algorithm | |
| CN104636407B (en) | Parameter value training and searching request treating method and apparatus | |
| CN113626700B (en) | Lawyer recommendation method, system and equipment | |
| CN103123624A (en) | Method of confirming head word, device of confirming head word, searching method and device | |
| CN105069561A (en) | Bidding evaluation expert extraction method and device | |
| CN106815265B (en) | Method and device for searching referee document | |
| CN109522275B (en) | Label mining method based on user production content, electronic device and storage medium |
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: 100083 No. 401, 4th Floor, Haitai Building, 229 North Fourth Ring Road, Haidian District, Beijing Applicant after: Beijing Guoshuang Technology Co.,Ltd. Address before: 100086 Cuigong Hotel, 76 Zhichun Road, Shuangyushu District, Haidian District, Beijing Applicant before: Beijing Guoshuang Technology Co.,Ltd. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |