CN110705283A - Deep learning method and system based on matching of text laws and regulations and judicial interpretations - Google Patents
Deep learning method and system based on matching of text laws and regulations and judicial interpretations Download PDFInfo
- Publication number
- CN110705283A CN110705283A CN201910843319.3A CN201910843319A CN110705283A CN 110705283 A CN110705283 A CN 110705283A CN 201910843319 A CN201910843319 A CN 201910843319A CN 110705283 A CN110705283 A CN 110705283A
- Authority
- CN
- China
- Prior art keywords
- matching
- text
- database
- regulations
- laws
- 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
- 238000013135 deep learning Methods 0.000 title claims abstract description 30
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000007477 logistic regression Methods 0.000 claims abstract description 23
- 238000012549 training Methods 0.000 claims description 33
- 239000013598 vector Substances 0.000 claims description 31
- 238000013136 deep learning model Methods 0.000 claims description 23
- 230000011218 segmentation Effects 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 7
- 238000012350 deep sequencing Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000013527 convolutional neural network Methods 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 2
- 230000008569 process Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 5
- 239000013604 expression vector Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
Images
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
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Databases & Information Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Machine Translation (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a deep learning method and a deep learning system based on text matching of laws and regulations and judicial interpretations, wherein a judicial interpretation is input, and related laws and regulations are inquired and recommended according to the judicial interpretation; and inputting a legal rule, and inquiring and recommending relevant judicial interpretations according to the legal rule. By using logistic regression and an Attention-based deep learning network, the input text (law and legal regulations or judicial interpretations) is matched with the text (judicial interpretations or legal regulations) in the database, and a plurality of texts (judicial interpretations or legal regulations) which are most relevant are output. And automatic updating is designed, and the model is updated by using the actual click condition in the using process, so that the model can be adjusted and optimized in real time, and the blank in the aspects of judicial interpretation based on a deep learning network and automatic matching tasks of laws and regulations is filled.
Description
Technical Field
The invention relates to a deep learning method and a deep learning system based on matching of text laws and regulations and judicial interpretation, in particular to an algorithm for matching the laws and regulations and the judicial interpretation based on a text Attention deep learning model.
Background
In the legal world, people in China set a series of laws and regulations aiming at various aspects of national economy, politics, culture and social life. Legal regulations are generally some guidelines summarized to a relatively high degree, and are not described in detail with respect to a particular scenario or case. On the basis, a series of judicial interpretation files are respectively formulated by each administrative institution, inspection institute and court in China according to respective specific scenes. The judicial interpretation file is used for making detailed explanation and specific guidance for certain laws and regulations under corresponding specific scenes.
In real-life situations, legislation is often modified accordingly for some reason. Correspondingly, judicial interpretation files for the laws and regulations should be adjusted accordingly, otherwise the judicial interpretation files may be violated with the existing laws and regulations. Therefore, determining whether a piece of judicial interpretation text matches a piece of legal or not becomes an important task. The industry would like to have some automated method to determine which laws and regulations match a particular judicial interpretation context; or which judicial interpretation matches a particular law or regulation. In summary, it is necessary to determine whether a piece of judicial interpretation text matches a piece of legal text.
At present, corresponding workers still need to make judgment manually, so that due to the fact that legal and legal documents and judicial interpretation documents are huge, workload is very large, and time and labor are consumed for manual judgment. Therefore, there is an urgent need for a system capable of automatically determining the matching relationship between the legal text and the judicial interpretation text.
The only patents judged by automated methods are mostly the matching of some laws to cases. Patent document CN107818138A discloses a case law regulation recommendation method, which extracts important information of words from legal documents by means of TF-IDF and the like, and matches them, mainly intelligently matches the input legal cases and corresponding laws and regulations, and does not use the latest deep learning knowledge. Patent document CN107423374A discloses a legal recommendation method based on classification label, which includes the following steps: 1) identifying the identity of the user; 2) Preprocessing a legal provision database; adjusting a legal recommendation strategy according to the identification result; carrying out legal recommendation according to the selected legal recommendation strategy; 3) and displaying the legal recommendation result to the user. The laws are recommended mainly based on classification labels, and intelligent matching between judicial interpretation and laws and regulations is not performed.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a deep learning method and system based on text matching law and regulation and judicial interpretation.
The deep learning method based on text matching laws and regulations and judicial interpretation provided by the invention comprises the following steps:
an input matching step: matching the input text with each text in a database corresponding to the input text, calculating matching scores through a logistic regression model, and outputting n input texts with the matching scores ranked in the front, wherein the n input texts are marked as preliminary matching texts;
a depth matching step: and forming m data pairs by the preliminary matching text and the input text, calculating matching scores of the data pairs through a deep learning model, and sequencing the output data pairs according to the matching scores to obtain the accurate matching text.
Preferably, the deep learning method based on text matching of laws and regulations and judicial interpretation further comprises the steps of constructing a database, constructing a database of laws and regulations according to the existing laws and regulations, constructing a database of judicial interpretation according to the existing judicial interpretation, and constructing a supervision and training sample database according to the matching relationship between laws and regulations and judicial interpretation.
Preferably, the deep learning method based on text matching laws and regulations and judicial interpretations further comprises a model updating step, if the precisely matched text is confirmed to be successfully matched, the precisely matched text is used as a positive example, otherwise, the precisely matched text is used as a negative example, the positive example and the negative example form a new training sample set together, and the new training sample set is respectively used as the input of the logistic regression model and the deep learning model so as to update the logistic regression model and the deep learning model.
Preferably, the step of constructing a database comprises:
building a regulation database: forming a database according to the entries of the existing laws and regulations, and performing word segmentation and vectorization on the texts in the entries;
constructing an interpretation database: forming a database according to the existing judicial interpretation items, and performing word segmentation and vectorization on the texts in the items;
building a training database: and matching the laws and regulations in the laws and regulations database with the judicial interpretations in the judicial interpretation database one by one, and recording matching results to respectively obtain matching cases and unmatched cases.
The invention provides a deep learning system based on text matching law and regulation and judicial interpretation, which comprises:
an input matching module: matching the input text with each text in a database corresponding to the input text, calculating matching scores through a logistic regression model, and outputting n input texts with the matching scores ranked in the front, wherein the n input texts are marked as preliminary matching texts;
a depth matching module: and forming m data pairs by the preliminary matching text and the input text, calculating matching scores of the data pairs through a deep learning model, and sequencing the output data pairs according to the matching scores to obtain the accurate matching text.
Preferably, the deep learning system based on text matching of laws and regulations and judicial interpretation further comprises a database construction module, a laws and regulations database is constructed according to the existing laws and regulations, a judicial interpretation database is constructed according to the existing judicial interpretation, and a supervision training sample database is constructed according to the matching relationship between laws and regulations and judicial interpretation.
Preferably, the deep learning system based on text matching laws and regulations and judicial interpretation further comprises a model updating module, if the precisely matched text is confirmed to be successfully matched, the model updating module is used as a positive example, otherwise, the model updating module is used as a negative example, the positive example and the negative example form a new training sample set together, and the new training sample set is respectively used as the input of the logistic regression model and the deep learning model so as to update the logistic regression model and the deep learning model.
Preferably, the build database module comprises:
constructing a regulation database module: forming a database according to the entries of the existing laws and regulations, and performing word segmentation and vectorization on the texts in the entries;
constructing an interpretation database module: forming a database according to the existing judicial interpretation items, and performing word segmentation and vectorization on the texts in the items;
constructing a training database module: and matching the laws and regulations in the laws and regulations database with the judicial interpretations in the judicial interpretation database one by one, and recording matching results to respectively obtain matching cases and unmatched cases.
Preferably, the logistic regression model is obtained by:
step 1: averaging word vectors of each word of each text in the database to obtain a representation vector of the whole text, and storing the representation vector in the database;
step 2: taking one text in a supervised training sample database as input, and obtaining the dot product distance between the input vector representation and the vector representation of each text in the corresponding database;
and step 3: performing gradient reduction on the obtained dot product distance by using a logistic regression logarithmic loss function, and updating the vector representation of the input text;
and 4, step 4: and taking another piece of text in the supervised training sample database as input to start to execute the step 2.
Preferably, the deep learning model is obtained by:
step 1: taking two texts in a supervised training sample database as input, and respectively reconstructing the two texts;
step 2: the distance between the vectors is utilized to carry out interaction on the two text vector matrixes to form a text matching graph;
and step 3: performing feature extraction on the text matching graph by using multiple layers of CNNs;
and 4, step 4: performing matching classification on the extracted features by using a multilayer full-connection layer classifier;
and 5: and (3) utilizing a multi-classification cross entropy loss function to implement a gradient descent algorithm, updating parameters of the deep learning model and updating a word vector matrix.
Compared with the prior art, the invention has the following beneficial effects:
the invention utilizes the deep learning model to automatically match judicial interpretation with laws and regulations, has high identification accuracy, can output and update the identification result in real time, is suitable for various scenes, and can be well suitable for requirements of future expansion of law and regulations libraries and judicial interpretation libraries.
Drawings
Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:
FIG. 1 is a schematic diagram of the organization of three databases according to the present invention;
FIG. 2 is a block diagram of a logistic regression model according to the present invention;
FIG. 3 is a block diagram of a deep learning model according to the present invention;
FIG. 4 is a system framework diagram of the present invention;
FIG. 5 is a block diagram illustrating the updating of two models based on actual usage according to the present invention.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the present invention.
The deep learning system based on the text matching law and regulation and the judicial interpretation can be realized through the step flow of the deep learning method based on the text matching law and regulation and the judicial interpretation. The deep learning method based on text matching laws and regulations and judicial interpretation can be understood as a preferred example of the deep learning system based on text matching laws and regulations and judicial interpretation by those skilled in the art.
In specific implementation, as shown in fig. 1, three databases need to be constructed first, and the specific steps include:
step 1, constructing a judicial interpretation database, collecting the existing judicial interpretations, organizing the existing judicial interpretations into the database according to entries, and performing word segmentation and vectorization on the databases;
step 2, constructing a law and regulation database, collecting the existing laws and regulations, organizing the existing laws and regulations into the database according to the entries, and performing word segmentation and vectorization on the databases;
step 3, constructing supervision training sample libraries by using professional knowledge, wherein each training sample library comprises a law and a judicial explanation, and the judicial explanation is matched with the law and the law to construct a matching case in the training library;
step 4, randomly sampling in a law and regulation database and a judicial interpretation database to form a mismatching case in a training library;
and 5, completing the construction of a database, wherein the database comprises three databases, namely a law and regulation database, a judicial interpretation database and a supervised training sample library. The content in each small box in a single long box represents a data type, for example, a legal and legal regulation database is taken as an example, a legal and legal regulation text represents the legal and legal regulation text of the data, a text word segmentation result represents the result of word segmentation of the legal and legal regulation, a text word vector matrix represents that words are converted into word vectors, and text vector representation is the result of obtaining a text representation vector according to a given method. In the supervised training sample library, each piece of data expresses a legal and legal regulation text, a judicial interpretation text and whether a matching relationship exists between the legal and legal regulations text and the judicial interpretation text, and the legal interpretation text and the judicial interpretation text are expressed as 1 if the legal and legal regulations text and the judicial interpretation text are matched, and are expressed as 0 if the legal and legal interpretation text and the judicial interpretation text are not.
As shown in fig. 2, the model system for performing matching judgment between an input text (legal or judicial interpretation) and a text (judicial interpretation or legal) in a database by using a logistic regression model specifically includes the following steps:
step 1, averaging word vectors of each word of each text in a database to obtain an expression vector of the whole text, and storing the expression vector in the database;
step 2, taking one text in the supervised training sample library as input, and solving the dot product distance between the input vector representation and the vector representation of each text in the corresponding database, wherein the dot product distance is specifically represented as
Wherein x and y respectively represent text vectors, n represents the number of texts, and subscript i represents the position serial number corresponding to the texts.
Step 3, performing gradient reduction on the obtained dot product distance by using a logistic regression logarithmic loss function, and updating the vector representation of the input text;
and 4, taking another text in the supervised training sample library as input to perform the process again.
As shown in fig. 3, the model system for performing matching judgment on an input text (legal or legal explanation) and a text (legal or legal explanation) in a database by using a deep learning model specifically includes the following steps:
step 1, two texts (word vector matrixes) in a supervised training sample library are used as input, and the two texts are respectively reconstructed by using a Transformer based on an attention mechanism;
step 2, utilizing the distance between the vectors to carry out interaction on the two text vector matrixes to form a text matching graph;
step 3, extracting the characteristics of the text matching graph by using multiple layers of CNNs;
step 4, performing matching classification on the extracted features by using a multilayer full-connection layer classifier;
and 5, utilizing the multi-classification cross entropy loss function to implement a gradient descent algorithm, updating the parameters of the model and updating the word vector matrix.
As shown in fig. 4, the model system comprehensively utilizes the deep learning model and the logistic regression model to perform matching judgment on the input text (legal or judicial interpretation) and the text (judicial interpretation or legal) in the database, and includes the following specific steps:
step 1, matching the input text with each text in a corresponding database (legal explanation corresponds to legal regulations ) by using a logistic regression model, and calculating a score.
Step 2, outputting n texts before scoring;
and 3, organizing the output text and the original input text into n data pairs, performing refined matching by using a deep learning model, and outputting scores.
And 4, sequencing and outputting the output texts according to the scores.
As shown in fig. 5, the algorithm and system for updating the model in real time according to the actual use condition includes the following specific steps:
step 1, when the user uses the system provided by the invention, the user can judge the recommended output text, if the user clicks to identify the matching result, the output text and the corresponding input text are considered to be successfully matched, otherwise, the matching is considered to be failed.
And 2, constructing the corresponding output text and the input text as a positive example by using the clicking condition of the user, otherwise constructing the output text and the input text as a negative example, and forming a new training sample set.
And 3, respectively updating parameters based on the logistic regression and the deep learning model by using the newly constructed training sample set and by using the steps 2 and 3.
The invention aims to design an algorithm and a system for automatically judging the matching degree of law and regulation texts and judicial interpretation texts.
Those skilled in the art will appreciate that, in addition to implementing the systems, apparatus, and various modules thereof provided by the present invention in purely computer readable program code, the same procedures can be implemented entirely by logically programming method steps such that the systems, apparatus, and various modules thereof are provided in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Therefore, the system, the device and the modules thereof provided by the present invention can be considered as a hardware component, and the modules included in the system, the device and the modules thereof for implementing various programs can also be considered as structures in the hardware component; modules for performing various functions may also be considered to be both software programs for performing the methods and structures within hardware components.
The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.
Claims (10)
1. A deep learning method based on text matching law and regulation and judicial interpretation is characterized by comprising the following steps:
an input matching step: matching the input text with each text in a database corresponding to the input text, calculating matching scores through a logistic regression model, and outputting n input texts with the matching scores ranked in the front, wherein the n input texts are marked as preliminary matching texts;
a depth matching step: and forming m data pairs by the preliminary matching text and the input text, calculating matching scores of the data pairs through a deep learning model, and sequencing the output data pairs according to the matching scores to obtain the accurate matching text.
2. The deep learning method based on text matching of laws and regulations and judicial interpretation as claimed in claim 1, further comprising the steps of constructing a database, constructing a database of laws and regulations according to existing laws and regulations, constructing a database of judicial interpretation according to existing judicial interpretation, and constructing a database of supervised training samples according to the matching relationship between laws and regulations and judicial interpretation.
3. The deep learning method based on text matching laws and regulations and judicial interpretation as claimed in claim 1, further comprising a model updating step, wherein if the exact matching text is confirmed as matching success, the positive example is taken, otherwise, the positive example and the negative example are taken as negative examples to form a new training sample set together, and the new training sample set is taken as input of the logistic regression model and the deep learning model respectively, so as to update the logistic regression model and the deep learning model.
4. The deep learning method based on text matching laws and jurisdictions according to claim 2 wherein the step of constructing a database comprises:
building a regulation database: forming a database according to the entries of the existing laws and regulations, and performing word segmentation and vectorization on the texts in the entries;
constructing an interpretation database: forming a database according to the existing judicial interpretation items, and performing word segmentation and vectorization on the texts in the items;
building a training database: and matching the laws and regulations in the laws and regulations database with the judicial interpretations in the judicial interpretation database one by one, and recording matching results to respectively obtain matching cases and unmatched cases.
5. A deep learning system based on text matching legal and legal regulations and judicial interpretation, comprising:
an input matching module: matching the input text with each text in a database corresponding to the input text, calculating matching scores through a logistic regression model, and outputting n input texts with the matching scores ranked in the front, wherein the n input texts are marked as preliminary matching texts;
a depth matching module: and forming m data pairs by the preliminary matching text and the input text, calculating matching scores of the data pairs through a deep learning model, and sequencing the output data pairs according to the matching scores to obtain the accurate matching text.
6. The deep learning system based on text matching of laws and regulations and judicial interpretation of claim 5, further comprising a database construction module for constructing a database of laws and regulations according to existing laws and regulations, a database of judicial interpretation according to existing judicial interpretation, and a database of supervised training samples according to the matching relationship between laws and regulations and judicial interpretation.
7. The deep learning system based on text matching laws and regulations and judicial interpretation of claim 5, further comprising a model updating module, wherein if the exact matching text is confirmed as matching successfully, the model updating module is used as a positive example, otherwise, the model updating module is used as a negative example, the positive example and the negative example are combined to form a new training sample set, and the new training sample set is respectively used as the input of the logistic regression model and the deep learning model to update the logistic regression model and the deep learning model.
8. The deep learning system based on text matching laws and regulations and judicial interpretations of claim 6, wherein the build database module comprises:
constructing a regulation database module: forming a database according to the entries of the existing laws and regulations, and performing word segmentation and vectorization on the texts in the entries;
constructing an interpretation database module: forming a database according to the existing judicial interpretation items, and performing word segmentation and vectorization on the texts in the items;
constructing a training database module: and matching the laws and regulations in the laws and regulations database with the judicial interpretations in the judicial interpretation database one by one, and recording matching results to respectively obtain matching cases and unmatched cases.
9. The deep learning method based on text matching laws and regulations and judicial interpretation as claimed in claim 2, wherein the logistic regression model is obtained by the following steps:
step 1: averaging word vectors of each word of each text in the database to obtain a representation vector of the whole text, and storing the representation vector in the database;
step 2: taking one text in a supervised training sample database as input, and obtaining the dot product distance between the input vector representation and the vector representation of each text in the corresponding database;
and step 3: performing gradient reduction on the obtained dot product distance by using a logistic regression logarithmic loss function, and updating the vector representation of the input text;
and 4, step 4: and taking another piece of text in the supervised training sample database as input to start to execute the step 2.
10. The deep learning method based on text matching laws and regulations and judicial interpretation as claimed in claim 2, wherein the deep learning model is obtained by the following steps:
step 1: taking two texts in a supervised training sample database as input, and respectively reconstructing the two texts;
step 2: the distance between the vectors is utilized to carry out interaction on the two text vector matrixes to form a text matching graph;
and step 3: performing feature extraction on the text matching graph by using multiple layers of CNNs;
and 4, step 4: performing matching classification on the extracted features by using a multilayer full-connection layer classifier;
and 5: and (3) utilizing a multi-classification cross entropy loss function to implement a gradient descent algorithm, updating parameters of the deep learning model and updating a word vector matrix.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910843319.3A CN110705283A (en) | 2019-09-06 | 2019-09-06 | Deep learning method and system based on matching of text laws and regulations and judicial interpretations |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910843319.3A CN110705283A (en) | 2019-09-06 | 2019-09-06 | Deep learning method and system based on matching of text laws and regulations and judicial interpretations |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110705283A true CN110705283A (en) | 2020-01-17 |
Family
ID=69194583
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910843319.3A Pending CN110705283A (en) | 2019-09-06 | 2019-09-06 | Deep learning method and system based on matching of text laws and regulations and judicial interpretations |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110705283A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111858830A (en) * | 2020-03-27 | 2020-10-30 | 北京梦天门科技股份有限公司 | Health supervision law enforcement data retrieval system and method based on natural language processing |
| CN111930601A (en) * | 2020-05-14 | 2020-11-13 | 国家电网有限公司客户服务中心 | Deep learning-based database state comprehensive scoring method and system |
| CN112199466A (en) * | 2020-09-08 | 2021-01-08 | 深圳价值在线信息科技股份有限公司 | Method and device for identifying associated regulation of mail |
| CN115309988A (en) * | 2022-08-10 | 2022-11-08 | 武汉圣男品牌管理有限公司 | Method, system and storage medium for matching webpage search content |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6772149B1 (en) * | 1999-09-23 | 2004-08-03 | Lexis-Nexis Group | System and method for identifying facts and legal discussion in court case law documents |
| CN106202413A (en) * | 2016-07-11 | 2016-12-07 | 北京大学深圳研究生院 | A kind of cross-media retrieval method |
| US20180308003A1 (en) * | 2017-04-21 | 2018-10-25 | Pranjal Singh | Hybrid approach to approximate string matching using machine learning |
| CN109086423A (en) * | 2018-08-08 | 2018-12-25 | 北京神州泰岳软件股份有限公司 | A kind of text matching technique and device |
| CN109472024A (en) * | 2018-10-25 | 2019-03-15 | 安徽工业大学 | A kind of file classification method based on bidirectional circulating attention neural network |
| WO2019064598A1 (en) * | 2017-09-29 | 2019-04-04 | Nec Corporation | Regression apparatus, regression method, and computer-readable storage medium |
| CN109740126A (en) * | 2019-01-04 | 2019-05-10 | 平安科技（深圳）有限公司 | Text matching technique, device and storage medium, computer equipment |
| CN109992648A (en) * | 2019-04-10 | 2019-07-09 | 北京神州泰岳软件股份有限公司 | The word-based depth text matching technique and device for migrating study |
| CN110019685A (en) * | 2019-04-10 | 2019-07-16 | 北京神州泰岳软件股份有限公司 | Depth text matching technique and device based on sequence study |
-
2019
- 2019-09-06 CN CN201910843319.3A patent/CN110705283A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6772149B1 (en) * | 1999-09-23 | 2004-08-03 | Lexis-Nexis Group | System and method for identifying facts and legal discussion in court case law documents |
| CN106202413A (en) * | 2016-07-11 | 2016-12-07 | 北京大学深圳研究生院 | A kind of cross-media retrieval method |
| US20180308003A1 (en) * | 2017-04-21 | 2018-10-25 | Pranjal Singh | Hybrid approach to approximate string matching using machine learning |
| WO2019064598A1 (en) * | 2017-09-29 | 2019-04-04 | Nec Corporation | Regression apparatus, regression method, and computer-readable storage medium |
| CN109086423A (en) * | 2018-08-08 | 2018-12-25 | 北京神州泰岳软件股份有限公司 | A kind of text matching technique and device |
| CN109472024A (en) * | 2018-10-25 | 2019-03-15 | 安徽工业大学 | A kind of file classification method based on bidirectional circulating attention neural network |
| CN109740126A (en) * | 2019-01-04 | 2019-05-10 | 平安科技（深圳）有限公司 | Text matching technique, device and storage medium, computer equipment |
| CN109992648A (en) * | 2019-04-10 | 2019-07-09 | 北京神州泰岳软件股份有限公司 | The word-based depth text matching technique and device for migrating study |
| CN110019685A (en) * | 2019-04-10 | 2019-07-16 | 北京神州泰岳软件股份有限公司 | Depth text matching technique and device based on sequence study |
Non-Patent Citations (1)
| Title |
|---|
| 赵文娟等: "基于汉语框架网的语义角色标注算法", 《图书情报工作》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111858830A (en) * | 2020-03-27 | 2020-10-30 | 北京梦天门科技股份有限公司 | Health supervision law enforcement data retrieval system and method based on natural language processing |
| CN111858830B (en) * | 2020-03-27 | 2023-11-14 | 北京梦天门科技股份有限公司 | Health supervision law enforcement data retrieval system and method based on natural language processing |
| CN111930601A (en) * | 2020-05-14 | 2020-11-13 | 国家电网有限公司客户服务中心 | Deep learning-based database state comprehensive scoring method and system |
| CN112199466A (en) * | 2020-09-08 | 2021-01-08 | 深圳价值在线信息科技股份有限公司 | Method and device for identifying associated regulation of mail |
| CN112199466B (en) * | 2020-09-08 | 2024-04-12 | 深圳价值在线信息科技股份有限公司 | Method and device for identifying associated rule of mail |
| CN115309988A (en) * | 2022-08-10 | 2022-11-08 | 武汉圣男品牌管理有限公司 | Method, system and storage medium for matching webpage search content |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110705283A (en) | Deep learning method and system based on matching of text laws and regulations and judicial interpretations | |
| CN107729468B (en) | answer extraction method and system based on deep learning | |
| CN111767716B (en) | Method and device for determining enterprise multi-level industry information and computer equipment | |
| CN106776538A (en) | The information extracting method of enterprise's noncanonical format document | |
| JP2021504789A (en) | ESG-based corporate evaluation execution device and its operation method | |
| CN112287037B (en) | Multi-entity mixed knowledge graph construction method, device and storage medium | |
| CN113434688B (en) | Data processing method and device for public opinion classification model training | |
| CN111415131A (en) | Big data talent resume analysis method based on natural language processing technology | |
| CN110310012B (en) | Data analysis method, device, equipment and computer readable storage medium | |
| CN111930937A (en) | BERT-based intelligent government affair text multi-classification method and system | |
| CN114330318A (en) | Method and device for recognizing Chinese fine-grained entities in financial field | |
| CN109344400A (en) | A kind of judgment method and device of document storage | |
| CN109299007A (en) | A kind of defect repair person's auto recommending method | |
| CN110674263B (en) | Method and device for automatically classifying model component files | |
| CN110362828B (en) | Network information risk identification method and system | |
| CN111191029B (en) | AC construction method based on supervised learning and text classification | |
| CN110618980A (en) | System and method based on legal text accurate matching and contradiction detection | |
| CN115952282A (en) | Intelligent bank customer complaint diversion handling method and system based on NLP technology | |
| CN108763565B (en) | Deep learning-based data automatic association matching construction method | |
| CN113837554A (en) | Food safety risk identification method and system based on multi-mode key information matching | |
| CN110414819B (en) | Work order scoring method | |
| CN111209375B (en) | Universal clause and document matching method | |
| CN112487214A (en) | Knowledge graph relation extraction method and system based on entity co-occurrence matrix | |
| CN117541044B (en) | Project classification method, system, medium and equipment based on project risk analysis | |
| US20190311265A1 (en) | Method and device for obtaining a system for labelling images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200117 |