CN114254617A - Method, device, computing equipment and storage medium for revising clauses - Google Patents
Method, device, computing equipment and storage medium for revising clauses Download PDFInfo
- Publication number
- CN114254617A CN114254617A CN202111565016.3A CN202111565016A CN114254617A CN 114254617 A CN114254617 A CN 114254617A CN 202111565016 A CN202111565016 A CN 202111565016A CN 114254617 A CN114254617 A CN 114254617A
- Authority
- CN
- China
- Prior art keywords
- clause
- revision
- record
- clauses
- reviewed
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/194—Calculation of difference between files
-
- 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
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Entrepreneurship & Innovation (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a method for revising clauses, which comprises the following steps: acquiring clauses to be checked and determining at least one clause revision record matched with the clauses to be checked in a clause revision library, wherein the clause revision library comprises a plurality of clause revision records, and each clause revision record comprises an original clause and revision content aiming at the original clause; and then, revising the clauses to be reviewed by using the determined at least one clause revision record. Therefore, the automatic examination and revision of the clauses can be realized for each clause needing to be examined by the user, so that the labor cost can be effectively reduced, the examination result of the clauses can be prevented from being influenced by human factors, and the benefit of the user can be protected as much as possible. In addition, the application also provides a corresponding device, a computing device and a storage medium.
Description
Technical Field
The present application relates to the field of artificial intelligence technologies, and in particular, to a method, an apparatus, a computing device, and a storage medium for revising a clause.
Background
In application scenarios such as enterprises, public institutions, financial institutions, etc., there is often a need to review terms in a large number of documents. For example, after the business contract is formulated between the enterprise a and the enterprise B, both parties usually need to review the terms recorded in the contract to avoid loss of part of the terms in the contract due to unreasonable content.
Currently, the manual review method is usually adopted to correct and revise the terms in the document one by one. However, the number of terms included in each document is usually large, and the number of documents to be manually checked in some scenes is also large, which not only causes a large workload of manual checking and a high labor cost, but also causes the effect of manually checking the terms to be easily affected by human factors, such as omission of manual checking, and the like, thereby bringing about a serious benefit loss to the user.
Disclosure of Invention
The embodiment of the application provides a method and a device for revising clauses, computing equipment and a storage medium, so that the revision clauses are automatically checked, the labor cost is effectively reduced, and the quality of checking the clauses is improved.
In a first aspect, an embodiment of the present application provides a method for revising a clause, where the method includes:
acquiring terms to be checked;
determining at least one clause revision record matched with the clause to be reviewed in a clause revision library, wherein the clause revision library comprises a plurality of clause revision records, and each clause revision record comprises an original clause and revision content aiming at the original clause;
and revising the clauses to be reviewed by utilizing the at least one clause revision record.
In one possible embodiment, the determining at least one clause revision record in the clause revision library matching the clause to be reviewed includes:
and determining at least one clause revision record in the clause revision library, wherein the similarity between the clause revision record and the clause to be reviewed meets a preset condition.
In a possible embodiment, the determining at least one clause revision record in the clause revision library whose similarity to the clause to be reviewed satisfies a preset condition includes:
respectively calculating the similarity between the clause to be checked and the original clause included in each clause revision record belonging to the target semantic category in the clause revision library, wherein the semantics of the clause to be checked belong to the target semantic category;
and determining at least one clause revision record from the plurality of clause revision records belonging to the target semantic category according to the similarity between the clause to be reviewed and the original clause included in each clause revision record, wherein the similarity between the original clause included in each clause revision record in the at least one clause revision record and the clause to be reviewed is higher than the similarity between the original clause included in other clause revision records in the clause revision library and the clause to be reviewed.
In one possible embodiment, the clause revision library includes a clause revision record for a plurality of semantic categories including the target category, the method further including:
obtaining historical revision data including a clause revision record revised for a plurality of original clauses over a historical period of time;
vectorizing the historical revision data to obtain a vectorized representation corresponding to the historical revision data;
and clustering a plurality of clause revision records included in the historical revision data according to the vectorization representation corresponding to the historical revision data, and determining semantic categories to which the clause revision records in the historical revision data belong respectively.
In one possible embodiment, the revising the to-be-audited clause by using the at least one clause revision record includes:
presenting a recommendation interface including the at least one terms revision record;
determining a target clause revision record from the at least one clause revision record in response to a user selection operation on the recommendation interface for the clause revision record;
and revising the clauses to be checked according to the revision content in the target clause revision record.
In one possible embodiment, the revising the to-be-audited clause by using the at least one clause revision record includes:
determining a target clause revision record from the at least one clause revision record, wherein the similarity between the original clause in the target clause revision record and the clause to be reviewed is maximum;
and revising the clauses to be checked according to the revision content in the target clause revision record.
In one possible embodiment, before revising the to-be-reviewed clause, the method further comprises:
matching the clauses to be checked with the trust clauses in the trust clause library;
then, the determining at least one clause revision record matching the to-be-reviewed clause in the clause revision library includes:
and when the trust clause base does not have the trust clause matched with the clause to be checked, determining at least one clause revision record matched with the clause to be checked in a clause revision base.
In one possible embodiment, the method further comprises:
acquiring a collection clause library, wherein the collection clause library comprises a plurality of candidate clauses;
clustering the candidate clauses according to the semantics of the candidate clauses to obtain a plurality of clusters, wherein the semantics of the candidate clauses in each cluster are similar, and the semantics of the candidate clauses in different clusters are different;
and determining candidate terms meeting the trust condition in each class cluster as trust terms, and adding the trust terms into the trust term library.
In a second aspect, an embodiment of the present application further provides an apparatus for revising terms, where the apparatus includes:
the first acquisition module is used for acquiring clauses to be checked;
the first matching module is used for determining at least one clause revision record matched with the clause to be reviewed in a clause revision library, wherein the clause revision library comprises a plurality of clause revision records, and each clause revision record comprises an original clause and revision content aiming at the original clause;
and the revision module is used for revising the clauses to be reviewed by utilizing the at least one clause revision record.
In a possible implementation manner, the first matching module is specifically configured to determine at least one clause revision record in the clause revision library, where a similarity between the clause revision record and the to-be-reviewed clause satisfies a preset condition.
In a possible implementation, the first matching module includes:
the calculation unit is used for respectively calculating the similarity between the clause to be checked and original clauses included in each clause revision record belonging to the target semantic category in the clause revision library, wherein the semantics of the clause to be checked belong to the target semantic category;
a first determining unit, configured to determine at least one clause revision record from the multiple clause revision records belonging to the target semantic category according to a similarity between the clause to be reviewed and an original clause included in each clause revision record, where a similarity between an original clause included in each clause revision record in the at least one clause revision record and the clause to be reviewed is higher than a similarity between an original clause included in other clause revision records in the clause revision library and the clause to be reviewed.
In one possible embodiment, the clause revision database includes a clause revision record for a plurality of semantic categories including the target category, the apparatus further including:
a second acquisition module for acquiring historical revision data including a clause revision record revised for a plurality of original clauses over a historical period of time;
the vectorization module is used for vectorizing the historical revision data to obtain a vectorization representation corresponding to the historical revision data;
the first clustering module is used for clustering a plurality of clause revision records included in the historical revision data according to the vectorization representation corresponding to the historical revision data, and determining semantic categories to which the clause revision records in the historical revision data belong respectively.
In one possible embodiment, the revision module includes:
a presentation unit for presenting a recommendation interface including the at least one clause revision record;
a second determination unit configured to determine a target clause revision record from the at least one clause revision record in response to a selection operation of the user for the clause revision record on the recommendation interface;
and the first revision unit is used for revising the clauses to be checked according to revision contents in the target clause revision record.
In one possible embodiment, the revision module includes:
a third determining unit, configured to determine a target clause revision record from the at least one clause revision record, where a similarity between an original clause in the target clause revision record and the clause to be reviewed is the largest;
and the second revision unit is used for revising the clauses to be checked according to the revision content in the target clause revision record.
In a possible embodiment, the apparatus further comprises:
the second matching module is used for matching the clauses to be checked with the trust clauses in the trust clause library before revising the clauses to be checked;
the first matching module is configured to determine at least one clause revision record matching the to-be-checked clause in a clause revision library when there is no trust clause matching the to-be-checked clause in the trust clause library.
In a possible embodiment, the apparatus further comprises:
the third acquisition module is used for acquiring a collection clause library, and the collection clause library comprises a plurality of candidate clauses;
the second clustering module is used for clustering the candidate clauses according to the semantics of the candidate clauses to obtain a plurality of clusters, wherein the semantics of the candidate clauses in each cluster are similar, and the semantics of the candidate clauses in different clusters are different;
and the determining module is used for determining the candidate clauses meeting the trust condition in each class cluster as the trust clauses, and the trust clauses are added into the trust clause library.
In a third aspect, an embodiment of the present application further provides a computing device, where the computing device may include a processor and a memory:
the memory is used for storing a computer program;
the processor is configured to perform the method according to any of the embodiments of the first aspect and the first aspect.
In a fourth aspect, this embodiment of the present application further provides a computer-readable storage medium, where the computer-readable storage medium is configured to store a computer program, where the computer program is configured to execute the method described in any one of the foregoing first aspect and the first aspect.
In the implementation manner of the embodiment of the application, to-be-checked clauses are acquired, and at least one clause revision record matched with the to-be-checked clauses in a clause revision library is determined, where the clause revision library includes a plurality of clause revision records, and each clause revision record includes an original clause and revision content for the original clause; and then, revising the clauses to be reviewed by using the determined at least one clause revision record. Therefore, the automatic examination and revision of the clauses can be realized for each clause needing to be examined by the user, so that the labor cost can be effectively reduced, the examination result of the clauses can be prevented from being influenced by human factors, and the benefit of the user can be protected as much as possible.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art according to the drawings.
FIG. 1 is a diagram illustrating an exemplary system architecture for revising terms in an embodiment of the present application;
FIG. 2 is a flowchart illustrating a method for revising terms according to an embodiment of the present application;
FIG. 3 is a schematic diagram of an exemplary recommendation interface;
FIG. 4 is a flow chart illustrating another method for revising clauses in accordance with an embodiment of the present application;
FIG. 5 is a schematic diagram of an apparatus for revising clauses in an embodiment of the present application;
fig. 6 is a schematic hardware structure diagram of a computing device in an embodiment of the present application.
Detailed Description
Referring to fig. 1, a schematic structural diagram of a system for revising terms according to an embodiment of the present application is provided. As shown in fig. 1, the system 100 includes a client 101 and a server 102. Here, the client 101 and the server 102 are connected in communication, and can perform normal data communication.
The client 101 may present an interactive interface to the user and prompt the user in the interactive interface to enter terms that need to be reviewed (hereinafter referred to as terms to be reviewed). The client 101 may then send the terms to be reviewed, entered by the user, to the server 102. After receiving the clause to be reviewed, the server 102 may invoke a clause revision library storing a plurality of clause revision records, and determine at least one clause revision record matching the clause to be reviewed in the clause revision library, where each clause revision record includes the original clause and the revision content for the original clause. In this way, the server 102 may revise the to-be-reviewed clause by using the determined at least one clause revision record, and specifically, may revise the to-be-reviewed clause by using the revision content included in one of the clause revision records, and the like. Finally, server 102 may return the revised terms to be reviewed to client 101 so that client 101 presents the revised terms to be reviewed to the user.
Thus, for each clause that needs to be checked by the user, the client 101 and the server 102 can be used to automatically check and revise the clause, so that not only can the labor cost be effectively reduced, but also the checking result of the clause can be prevented from being influenced by human factors, and the benefit of the user can be protected as much as possible.
It is understood that the architecture of the system for revising clauses 100 shown in fig. 1 is only one example provided by the embodiments of the present application, and the architecture of the system for revising clauses is not limited to the example shown in fig. 1 in actual application. For example, in other possible implementations, the system for revising the clauses may provide a service for locally reviewing the clauses for the user, in which case, the system may only include a locally located user device (e.g., a local computer, etc.) and the user device performs the clause revision process performed by the server 102, wherein the user device may further present a corresponding human-machine interaction interface to receive the clauses input by the user and present the revision result for the clauses to the user. In summary, the embodiments of the present application may be applied to any applicable system and are not limited to the above examples.
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, various non-limiting embodiments accompanying the present application examples are described below with reference to the accompanying drawings. It is to be understood that the embodiments described are only a few embodiments of the present application and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Referring to fig. 2, fig. 2 is a schematic flow chart of a method for revising terms in an embodiment of the present application, and the method can be applied to the system 100 shown in fig. 1, and of course, can also be applied to other applicable systems. For convenience of explanation and understanding, the following description will be given by taking as an example the review and revision process applied to the system 100 shown in fig. 1 and executed by the server 102 for the to-be-reviewed clauses. The method specifically comprises the following steps:
s201: and acquiring the clauses to be checked.
In this embodiment, the terms to be checked may be, for example, entries in legal provisions or agreement documents, such as terms for agreement on the content of responsibility of both parties in a lease contract, terms for agreement on benefit distribution in a business contract, and the like.
In an exemplary embodiment, the user may provide the client 101 with a file recorded with one or more terms, such as a rental contract, a legal file, and the like, by way of data copy or file import. The client 101 may then forward the file to the server 102. Thus, upon receiving the file, the server 102 may review each of the terms in the file and make a corresponding revision to the problematic term. For convenience of understanding and explanation, the following description takes an example of a process of auditing and revising one clause (i.e., the above-mentioned clause to be audited) of the file record by the server 102, and for each remaining clause of the file record, reference may be made to a revision process of the clause to be audited in this embodiment.
Of course, in other possible embodiments, the user may also directly input the terms that need to be checked into the client 101, for example, the user may directly input one or more terms contents on the client 101 and send them to the server 102 by the client 101. In this embodiment, a specific implementation manner of how the user provides the to-be-checked clauses to the client 101 is not limited.
S202: at least one clause revision record matched with the clause to be reviewed in a clause revision library is determined, wherein the clause revision library comprises a plurality of clause revision records, and each clause revision record comprises an original clause and revision contents aiming at the original clause.
In this embodiment, the server 102 may determine whether the to-be-reviewed clause needs to be revised based on the historical data. In a specific implementation, the server 102 may create a clause revision library in advance based on a plurality of clause revision records, so as to check whether the to-be-checked clause needs to be revised by using the clause revision record. The clause revision record refers to a record of revising the clause by a reviewer in a process of manually reviewing the clause, and includes original clauses (i.e., the clauses before manual revision) and revision contents for the original clauses. For example, the original clause is specifically "the first party should pay the subscription XXX element to the second party within 7 days after the first party signs a contract with the second party", and when the reviewer revises the original clause, the revision content is specifically to revise the "subscription" in the original clause into "fixed amount". In practical applications, after the reviewer revises a certain original clause, the client 101 or the server 102 may generate a clause revision record including the original clause and the revised content according to the reviewer's revision operation on the original clause, so as to subsequently establish a clause revision library based on the clause revision record.
As an implementation example of constructing the clause revision library, the server 102 may first obtain historical revision data, which includes a plurality of clause revision records generated in a historical time period (e.g., the past year, etc.), wherein each of the clause revision records is a record of manual revision performed by a reviewer for an original clause in the historical time period.
The server 102 may then vectorize the historical revision data, and map each clause revision record in the historical revision data to the same vector space, thereby obtaining a vectorized representation corresponding to the historical revision data. Illustratively, the server 102 may perform vectorization processing on each clause revision record in the history revision data in units of words, such as vectorization on the clause revision record through a bag-of-words model, an n-gram algorithm, a TF-IDF algorithm, a word2vec algorithm, and the like; alternatively, the server 102 may perform vectorization processing on each clause revision record in the history revision data in Sentence units, such as vectorization of the clause revision record by a Linear Discriminant Analysis (LDA) algorithm, a Long Short-Term Memory (LSTM) algorithm, a BERT (bidirectional Encoder retrieval from transformations) model, a sequence-BERT model, and the like. When the sequence-BERT model is adopted to carry out vectorization on the entry revision record, semantic features of sentences can be better captured, and correspondingly, the server 102 can take out the hidden layer representation of the [ CLS ] position in the sequence-BERT model as the vectorized representation of the clause revision record.
Finally, the server 102 may cluster the plurality of clause revision records included in the historical revision data according to the vectorization representation corresponding to the historical revision data, and determine semantic categories to which the respective clause revision records in the historical revision data belong.
For example, the process of clustering a plurality of clause revision records according to the vectorization representation of the clause revision records may specifically be:
(1) the server 102 may treat the vectorized representation corresponding to each clause revision record as a node.
(2) The server 102 calculates the similarity between different nodes, for example, the server 102 may calculate the vector distance between the vectorized representations corresponding to different nodes, and determine the similarity between two nodes according to the vector distance between the two nodes. In general, the smaller the vector distance is, the higher the similarity between the two nodes is represented, that is, the more similar the clause revision records corresponding to the two nodes are; conversely, the larger the vector distance, the lower the similarity between the two nodes is characterized.
(3) The server 102 may connect nodes with higher similarity in the order of similarity from high to low, and not connect nodes with lower similarity, for example, the server 102 may connect only 10 nodes with highest similarity at a time. The node after the connection is established can be regarded as a new node, at this time, the new node includes a plurality of clause revision records, and the vectorization representation corresponding to the new node can be obtained by merging the vectorization representations corresponding to the plurality of clause revision records. Thus, a tree-shaped graph can be formed based on the plurality of nodes, wherein connection exists between part of nodes in the tree-shaped graph, and connection does not exist between the other part of nodes in the tree-shaped graph.
(4) The server 102 may divide the tree into a plurality of subgraphs according to the node connection condition in the tree, each subgraph includes one node or a plurality of connected nodes, and the nodes in different subgraphs have no connection. Thus, the server 102 completes a round of clustering operations.
(5) The server 102 may iterate the clustering operation, that is, recalculate the similarity between different nodes, connect the partial nodes with the highest similarity in the order from high similarity to low similarity (a plurality of connected nodes may be regarded as a new node), generate a new tree graph according to the similarity, and partition the new tree graph to obtain a plurality of new subgraphs. It is to be noted that, in each iteration clustering process, the nodes processed by the server 102 are all nodes that have not established connection with other nodes in the previous iteration clustering process and new nodes generated based on a plurality of connected nodes.
The server 102 may perform iterative clustering operations, that is, iteratively perform (2) to (4), for a plurality of times until a preset iteration termination condition is satisfied, where the iteration termination condition may be, for example, that each subgraph generated by the last iteration includes a plurality of connected nodes, and the like. In this way, the server 102 may finally obtain a plurality of subgraphs, each subgraph corresponding to a semantic category to which the semantics of the provision revision record corresponding to each node in each subgraph belong. In this way, the server 102 may implement vectorization representation corresponding to the clause revision record, complete clustering of the plurality of clause revision records, and may further generate the clause revision library according to the clustered clause revision records.
After the clause revision library is constructed, the server 102 may match the clause revision record in the clause revision library with the clause to be reviewed. When there is a match between the clause revision record and the clause to be reviewed, the server 102 may determine that the clause to be reviewed has the necessity of being revised, at which point there is a revision made by the reviewer to the same or similar clause to be reviewed within the characterization history period. When there is no revision record of terms that matches the terms to be reviewed, the server 102 may determine that no revision scheme currently exists for the terms to be reviewed. In some scenarios, the server 102 may further determine that the to-be-reviewed clause passes the review, and may not need to be revised.
Notably, each of the clause revision records in the clause revision library has previously completed semantic clustering, and the differences between the clause revision records belonging to different semantic categories are typically large (including semantics as well as clause content). Therefore, in the process of matching the clause revision record in the clause revision library with the clause to be checked, the server 102 may, for example, first identify the semantics of the clause to be checked and determine the target semantic category to which the semantics belong, so that the server 102 may match the clause to be checked with the clause revision record belonging to the target semantic category in the clause revision library one by one. In this way, the number of clause revision records matching the clauses to be reviewed can be reduced, and thus the consumption of computing resources of the server 102 can be reduced.
As an example, when determining the clause revision record matching the clause to be reviewed in the clause revision library, the server 102 may determine the clause revision record according to a similarity between the clause to be reviewed and the original clause, and specifically may determine at least one clause revision record whose similarity to the clause to be reviewed satisfies a preset condition in the clause revision library.
For example, the preset condition may specifically be that the similarity between the determined clause revision record and the clause to be reviewed is greater than the similarity between other clause revision records in the clause revision library and the clause to be reviewed. In a specific implementation, the server 102 may respectively calculate similarities between the clause to be reviewed and the original clause included in each clause revision record belonging to the target semantic category in the clause revision library, for example, determine a similarity between two clauses by calculating a vector distance between a vectorized representation corresponding to the clause to be reviewed and a vectorized representation corresponding to the original clause, or calculate a similarity between two clauses by a text matching algorithm (the similarity is 100% if the texts are completely consistent), and the like. Then, the server 102 may determine at least one clause revision record from the plurality of clause revision records belonging to the target semantic category according to similarities between the clause to be reviewed and the original clause included in each clause revision record, where the similarities between the original clause included in each of the determined at least one clause revision records and the clause to be reviewed are higher than the similarities between the original clause included in other clause revision records under the target semantic category and the clause to be reviewed, and are also higher than the similarities between the original clause included in other clause revision records in the revision clause library and the clause to be reviewed. For example, the server 102 may determine that the top K (K is a positive integer) clause revision records with the highest similarity match the clause to be reviewed, or the server 102 may determine that one or more clause revision records with a similarity greater than a threshold (e.g., 85%) match the clause to be reviewed, which is not limited in this embodiment.
It should be noted that the implementation manner of building the clause revision library and determining the clause revision record matching the clause to be reviewed is only an exemplary illustration, and in practical applications, the server 102 may also build the clause revision library and determine the clause revision record matching the clause to be reviewed by other manners. For example, in other possible examples, the server 102 may directly add the plurality of clause revision records to the clause revision repository without clustering the plurality of clause revision records; accordingly, the server 102 may match the to-be-reviewed clauses with the clause revision records in the clause revision library one by one, and the like, when determining the clause revision record matching the to-be-reviewed clause.
In practice, the server 102 may periodically update the clause revision library. For example, after the reviewer checks and revises the clauses for a period of time, the server 102 may gradually accumulate a certain number of new clause revision records, so that the server 102 may merge the new clause revision records with the existing clause revision records in the clause revision library, and may perform semantic clustering again, thereby implementing the update of the clause revision library. Of course, the server 102 may also create a new provision revision library separately based on the new provision revision record, and the present embodiment does not limit this. In addition, in other embodiments, the execution subject of the construction of the provision revision library may be another device other than the server 102, and the provision revision library is transmitted to the server 102 or the like by the other device after being constructed.
S203: and revising the to-be-audited clauses by utilizing the at least one clause revision record.
In this embodiment, the server 102 may determine one or more clause revision records from a clause revision library. When only one clause revision record is matched with the clause to be reviewed, the server 102 can revise the clause to be reviewed according to the clause revision record. When there are multiple clause revision records simultaneously matching the clause to be reviewed, the server 102 may select one clause revision record from the multiple clause revision records, and revise the clause to be reviewed according to the selected clause revision record.
As an implementation example, the server 102 may implement a user to decide which clause revision record to select to revise the terms to be reviewed. Specifically, the server 102 may present a recommendation interface to the user via the client 101, the recommendation interface including one or more of the clause revision records determined by the server 102, such as the clause revision record 1 through the clause revision record 3 shown in fig. 3. In this way, the user can select which clause revision record to revise the to-be-reviewed clause on the recommendation interface. In actual application, specific content of the clause to be reviewed can be presented in the recommendation interface, and a detailed control is further arranged in the recommendation interface, so that after a user clicks the detailed control corresponding to one of the clause revision records, the recommendation interface can further present specific information corresponding to the clause revision record, such as specific text content of the original clause and revision content of the original clause, and the like, so that the user can determine which clause revision record is selected to revise the clause to be reviewed by comparing text content differences between the clause to be reviewed and the original clause on the recommendation interface, and use experience of the user is improved. Accordingly, the server 102 may determine a target clause revision record for revising the to-be-reviewed clause in response to a selection operation of the user on the recommendation interface for the clause revision record, and may further revise the to-be-reviewed clause according to revision contents of the target clause revision record for the original clause.
For example, assuming that the original clause is "the first party should pay the subscription XXX to the second party within 7 days after signing a lease contract with the second party", the revision content is to modify the "subscription" in the original clause into "fixed fee", and when the clause to be reviewed is "the first party should pay the subscription XXX to the second party within 7 days after signing a business contract with the second party", the server 102 may modify the "subscription" in the clause to be reviewed into "fixed fee" according to the revision content, so that the revised clause to be reviewed is: "Party A shall pay Party B a subscription XXX dollar 7 days after contracting with Party B.
Further, for the clause revision record selected by the user, the server 102 may increase the similarity between the original clauses in the clause revision record and the clauses to be reviewed. Thus, when the server 102 subsequently performs review and revision on the same to-be-reviewed clauses input for other users, the clause revision record can be preferentially recommended to other users, such as at the primary recommendation position, or the recommended position of the clause revision record is more advanced relative to the last recommended position, etc.
In an actual application scenario, when a part of the clauses is manually revised by the auditing staff, information related to a specific application scenario, such as a date, an amount, an address and the like, which are related to the clauses, may be revised, and when another part of the clauses is manually revised, the revision content is irrelevant to the specific application scenario, for example, "a subscription" in the clauses is revised to "a deposit" and the like, which is applicable to various application scenarios. Thus, in one example, the clause revision library constructed by the server 102 may be divided into two categories, a specific revision and a general revision. The item revision record belonging to the specific revision characterizes that the revision of the original item by the auditor is related to a specific scene, such as the revision of the date, the money and the address in the original item; the clause revision record belongs to the general revision, and represents that the revision of the original clause by the auditor is independent of a specific scene, so that the method can be applied to various scenes. In this way, before recommending the clause revision record to the user, the server 102 may determine whether the revision content included in the clause revision record belongs to a general revision, and if so, present the recommended clause revision record to the user, and if not, recommend no clause revision record to the user, such as revising the to-be-checked clause directly according to the clause revision record, and the like. Of course, in other possible examples, the server 102 may also filter the clause revision records when constructing the clause revision library, and only retain the clause revision records belonging to the general revision, and thus establish the clause revision library, which is not limited in this embodiment.
In addition to the above-described implementation of selecting a target clause revision record by a user, in other possible implementation examples, the server 102 may not need user involvement in the process of reviewing and revising the clauses to be reviewed. Specifically, the server 102 may automatically determine a target clause revision record from the determined at least one clause revision record, for example, randomly select a clause revision record as the target clause revision record through a random algorithm, or use the clause revision record with the greatest similarity between the original clause and the clause to be reviewed as the target clause revision record, etc. Then, the server 102 revises the to-be-checked clause according to the revised content of the original clause in the target clause revision record.
Thus, for each clause that needs to be checked by the user, the server 102 can automatically check and revise the clause, so that not only can the labor cost be effectively reduced, but also the checking result of the clause can be prevented from being influenced by human factors, and the benefit of the user can be protected as much as possible.
In the above embodiment, the server 102 uses the clause revision library to revise the clause to be reviewed as an example, and in an actual application scenario, there are usually a large number of clauses in the document that do not need to be revised, so the server 102 may not need to perform the revision process on the part of the clauses. Specifically, referring to fig. 4, a flowchart of another method for revising terms is provided in an embodiment of the present application, in this embodiment, the server 102 may first determine whether the to-be-reviewed term is a trust term that does not need to be revised, and if not, the server 102 may perform a revision process on the to-be-reviewed term. As shown in fig. 4, the method may specifically include:
s401: and acquiring the clauses to be checked.
S402: and matching the clauses to be checked with the trust clauses in the trust clause library.
In this embodiment, the server 102 may pre-construct a trust term library, where the trust term library includes a plurality of trust terms. The trust clauses are high-quality clauses which have correct clause contents and can be trusted, and when the clauses to be checked are matched with the trust clauses, the server 102 can determine that the clauses to be checked pass the check without revision.
Illustratively, the server 102 may determine the terms that have been reviewed by the reviewers as the trust terms by adaptive learning. Specifically, the server 102 may first obtain a collection clause library including a plurality of candidate clauses, where the candidate clauses may be, for example, terms recognized by one or more reviewers as correct and reasonable terms when manually reviewing the clauses, and the terms are added to the collection clause library. Due to the fact that different auditors have different professional levels and have large influence on human factors identified by the terms, some terms with low quality may be added to the collection term library when the auditors audit the terms. Thus, the server 102 may filter out a portion of the candidate terms from the collection term library as trust terms. For example, the server 102 may cluster the semantics of the candidate terms according to the semantics of the candidate terms to obtain a plurality of clusters, where the semantics of the candidate terms in each cluster are similar, and the semantics of the candidate terms in different clusters are different. The specific implementation manner of the server 102 performing semantic clustering on the multiple candidate terms may refer to the foregoing text clustering process, or may adopt other available implementation manners, which is not limited and described in this embodiment. Finally, server 102 may determine candidate terms in each class cluster that satisfy the trust criteria as trust terms and add the trust terms to the library of trust terms.
For example, when determining whether the candidate clause satisfies the trust condition, the server 102 may specifically present the candidate clause to one or more other reviewers for review, and when the candidate clause is also deemed by the other reviewers to pass the review, the server 102 may determine that the candidate clause satisfies the trust condition, otherwise, it does not satisfy the trust condition. Alternatively, when determining whether the candidate clause satisfies the trust condition, the server 102 may also calculate an average similarity between each candidate clause in the class cluster and other candidate clauses for each class cluster, and rank each candidate clause according to the average similarity, so that the server 102 may determine the candidate clause whose average similarity is greater than the similarity threshold as the trust clause satisfying the trust condition, and determine that the candidate clause whose average similarity is less than the similarity threshold does not satisfy the trust condition. Or, the server 102 may select the trust terms by combining the two implementation manners, for example, the server 102 may select candidate terms in each class cluster whose average similarity is greater than the similarity threshold, and present the selected candidate conditions to other one or more reviewers for review, so that the server 102 may determine the candidate terms that pass the review as the trust terms, and the like. In this embodiment, the specific implementation manner of the server 102 for determining whether the candidate clause satisfies the trust condition is not limited.
In addition, if there are multiple reviewers who all add the same terms to the collection term library in actual application, the server 102 may also directly determine the terms as the trust terms.
In other embodiments, the execution subject for building the trust term library may be another device besides the server 102, and the other device sends the trust term library to the server 102 after the trust term library is built.
After obtaining the trust clause base, the server 102 may determine the trust clauses matching the to-be-checked clauses in the trust clause base by means of exact matching or fuzzy matching. When the exact match is adopted, the server 102 may traverse each trust clause in the trust clause library, and if the trust clause is completely consistent with the clause to be reviewed, the server 102 may determine that the trust clause matches the clause to be reviewed, and may further determine that the clause to be reviewed passes the review; otherwise, the server 102 determines that the trust terms do not match the terms to be reviewed. When fuzzy matching is adopted, if the trust clauses in the trust clause library are not completely consistent with the clauses to be checked, but the difference between the trust clauses and the clauses to be checked is only specific contents such as date, amount, address and the like, the server 102 can also determine that the trust clauses are matched with the clauses to be checked, and further determine that the clauses to be checked pass the checking; if the difference between the trust terms and the terms to be reviewed includes other content, the server 102 determines that the trust terms and the terms to be reviewed do not match. In an actual application scenario, the server 102 may also present an examination result for terms to be examined, and if the terms to be examined pass the examination, the server 102 may mark the terms through a specific color (such as black on green, etc.) or in another way, and if the terms to be examined do not pass the examination, the server 102 may mark the terms through another color or in another way, so as to distinguish the terms that pass the examination from the terms that do not pass the examination. In practical applications, if the number of the trust terms in the trust term library is large and the expression of the trust terms is standardized, the term hit rate of the server 102 based on the trust term library is usually high, so that the workload of manual review can be effectively reduced.
In this embodiment, when the to-be-checked clause is not successfully matched with the trust clause in the trust clause library, the server 102 may revise the to-be-checked clause through the clause revising library to improve the content quality of the to-be-checked clause. Or, on the basis of matching by using the trust clause library, the server 102 may further check the clause based on a preset clause checking rule and a checking element, and when the check is passed, the server 102 may determine that the clause to be checked passes the check, and does not need to be revised, etc. And when the review fails, the server 102 may revise the to-be-reviewed clauses through the clause revision library. At this time, the server 102 may also make a risk prompt, such as a prompt of "clause is absent" (e.g., the clause content is incomplete, the entire clause content is absent, etc.), "clause quality is poor and needs to be replaced" (e.g., the clause content is old, the sentence is not smooth, etc.).
The term of the term to be checked is the content of the term which is important to be checked, and may be, for example, "date", "contact", "identification number", "address", "payment method", "confidential term", "responsibility after violation of confidential obligation", "amount of money to be breached", or the like. The audit rule refers to a rule that the audit element needs to satisfy, and can be generally formulated by an auditor. For example, for an audit element "identity card number", an audit rule defined by an auditor may be: the number of the ID card number is correct, and if the check code passes the check, the verification is output; and if the audit rule is not met, alarming and prompting.
S403: when the trust clause matched with the clause to be checked does not exist in the trust clause base, at least one clause revision record matched with the clause to be checked in the clause revision base is determined, wherein the clause revision base comprises a plurality of clause revision records, and each clause revision record comprises the original clause and revision contents aiming at the original clause.
S404: and revising the to-be-checked clauses by using the determined at least one clause revision record.
In this embodiment, specific implementation manners of step S401 and steps S403 and S404 may refer to the descriptions of the relevant parts of the foregoing embodiments, and are not described herein again.
Thus, through the trust clause library and the clause revision library, the server 102 can perform double review and automatic revision on the clause to be reviewed, so that the accuracy of the clause review can be improved, and the review quality can be improved.
In addition, the embodiment of the application also provides a device for revising the clauses. Referring to fig. 5, fig. 5 is a schematic structural diagram illustrating an apparatus 500 for revising a clause in an embodiment of the present application, including:
a first obtaining module 501, configured to obtain terms to be checked;
a first matching module 502, configured to determine at least one clause revision record matching the clause to be reviewed in a clause revision library, where the clause revision library includes a plurality of clause revision records, and each clause revision record includes an original clause and revision content for the original clause;
a revising module 503, configured to revise the to-be-reviewed clause by using the at least one clause revision record.
In a possible implementation manner, the first matching module 502 is specifically configured to determine at least one clause revision record in the clause revision library, where a similarity between the clause revision record and the to-be-reviewed clause satisfies a preset condition.
In a possible implementation, the first matching module 502 includes:
the calculation unit is used for respectively calculating the similarity between the clause to be checked and original clauses included in each clause revision record belonging to the target semantic category in the clause revision library, wherein the semantics of the clause to be checked belong to the target semantic category;
a first determining unit, configured to determine at least one clause revision record from the multiple clause revision records belonging to the target semantic category according to a similarity between the clause to be reviewed and an original clause included in each clause revision record, where a similarity between an original clause included in each clause revision record in the at least one clause revision record and the clause to be reviewed is higher than a similarity between an original clause included in other clause revision records in the clause revision library and the clause to be reviewed.
In one possible embodiment, the clause revision database includes a clause revision record of a plurality of semantic categories including the target category, and the apparatus 500 further includes:
a second acquisition module for acquiring historical revision data including a clause revision record revised for a plurality of original clauses over a historical period of time;
the vectorization module is used for vectorizing the historical revision data to obtain a vectorization representation corresponding to the historical revision data;
the first clustering module is used for clustering a plurality of clause revision records included in the historical revision data according to the vectorization representation corresponding to the historical revision data, and determining semantic categories to which the clause revision records in the historical revision data belong respectively.
In one possible implementation, the revision module 503 includes:
a presentation unit for presenting a recommendation interface including the at least one clause revision record;
a second determination unit configured to determine a target clause revision record from the at least one clause revision record in response to a selection operation of the user for the clause revision record on the recommendation interface;
and the first revision unit is used for revising the clauses to be checked according to revision contents in the target clause revision record.
In one possible implementation, the revision module 503 includes:
a third determining unit, configured to determine a target clause revision record from the at least one clause revision record, where a similarity between an original clause in the target clause revision record and the clause to be reviewed is the largest;
and the second revision unit is used for revising the clauses to be checked according to the revision content in the target clause revision record.
In a possible implementation, the apparatus 500 further includes:
the second matching module is used for matching the clauses to be checked with the trust clauses in the trust clause library before revising the clauses to be checked;
then, the first matching module 502 is configured to determine at least one clause revision record matching the clause to be reviewed in a clause revision library when there is no trust clause matching the clause to be reviewed in the trust clause library.
In a possible implementation, the apparatus 500 further includes:
the third acquisition module is used for acquiring a collection clause library, and the collection clause library comprises a plurality of candidate clauses;
the second clustering module is used for clustering the candidate clauses according to the semantics of the candidate clauses to obtain a plurality of clusters, wherein the semantics of the candidate clauses in each cluster are similar, and the semantics of the candidate clauses in different clusters are different;
and the determining module is used for determining the candidate clauses meeting the trust condition in each class cluster as the trust clauses, and the trust clauses are added into the trust clause library.
It should be noted that, for the contents of information interaction, execution process, and the like between the modules and units of the apparatus, since the same concept is based on the method embodiment in the embodiment of the present application, the technical effect brought by the contents is the same as that of the method embodiment in the embodiment of the present application, and specific contents may refer to the description in the foregoing method embodiment in the embodiment of the present application, and are not described herein again.
In addition, the embodiment of the application also provides the computing equipment. Referring to fig. 6, fig. 6 is a schematic diagram illustrating a hardware structure of a computing device in an embodiment of the present application, where the device 600 may include a processor 601 and a memory 602.
Wherein the memory 602 is used for storing computer programs;
the processor 601 is configured to execute the method for revising the clauses in the above method embodiments according to the computer program.
In addition, the embodiment of the present application further provides a computer-readable storage medium for storing a computer program for executing the method for revising the clauses described in the above method embodiment.
In the names of the "first obtaining module", "first matching module", and the like, the "first" mentioned in the embodiments of the present application is only used for name identification, and does not represent the first in sequence. The same applies to "second", "third", etc.
As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that all or part of the steps in the above embodiment methods can be implemented by software plus a general hardware platform. Based on such understanding, the technical solution of the present application may be embodied in the form of a software product, which may be stored in a storage medium, such as a read-only memory (ROM)/RAM, a magnetic disk, an optical disk, or the like, and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network communication device such as a router) to execute the method according to the embodiments or some parts of the embodiments of the present application.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the apparatus embodiment, since it is substantially similar to the method embodiment, it is relatively simple to describe, and reference may be made to some descriptions of the method embodiment for relevant points. The above-described embodiments of the apparatus are merely illustrative, and the modules described as separate parts may or may not be physically separate, and the parts displayed as modules may or may not be physical modules, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
The above description is only an exemplary embodiment of the present application, and is not intended to limit the scope of the present application.
Claims (11)
1. A method of revising a clause, the method comprising:
acquiring terms to be checked;
determining at least one clause revision record matched with the clause to be reviewed in a clause revision library, wherein the clause revision library comprises a plurality of clause revision records, and each clause revision record comprises an original clause and revision content aiming at the original clause;
and revising the clauses to be reviewed by utilizing the at least one clause revision record.
2. The method of claim 1, wherein determining at least one clause revision record in a clause revision library that matches the clause to be reviewed comprises:
and determining at least one clause revision record in the clause revision library, wherein the similarity between the clause revision record and the clause to be reviewed meets a preset condition.
3. The method of claim 2, wherein the determining at least one clause revision record in the clause revision library whose similarity to the clause to be reviewed satisfies a preset condition comprises:
respectively calculating the similarity between the clause to be checked and the original clause included in each clause revision record belonging to the target semantic category in the clause revision library, wherein the semantics of the clause to be checked belong to the target semantic category;
and determining at least one clause revision record from the plurality of clause revision records belonging to the target semantic category according to the similarity between the clause to be reviewed and the original clause included in each clause revision record, wherein the similarity between the original clause included in each clause revision record in the at least one clause revision record and the clause to be reviewed is higher than the similarity between the original clause included in other clause revision records in the clause revision library and the clause to be reviewed.
4. The method of claim 3, wherein the clause revision library comprises a clause revision record for a plurality of semantic categories including the target category, the method further comprising:
obtaining historical revision data including a clause revision record revised for a plurality of original clauses over a historical period of time;
vectorizing the historical revision data to obtain a vectorized representation corresponding to the historical revision data;
and clustering a plurality of clause revision records included in the historical revision data according to the vectorization representation corresponding to the historical revision data, and determining semantic categories to which the clause revision records in the historical revision data belong respectively.
5. The method of claim 1, wherein the revising the to-be-reviewed clause using the at least one clause revision record comprises:
presenting a recommendation interface including the at least one terms revision record;
determining a target clause revision record from the at least one clause revision record in response to a user selection operation on the recommendation interface for the clause revision record;
and revising the clauses to be checked according to the revision content in the target clause revision record.
6. The method of claim 1, wherein the revising the to-be-reviewed clause using the at least one clause revision record comprises:
determining a target clause revision record from the at least one clause revision record, wherein the similarity between the original clause in the target clause revision record and the clause to be reviewed is maximum;
and revising the clauses to be checked according to the revision content in the target clause revision record.
7. The method of claim 1, wherein prior to revising the to-be-reviewed clauses, the method further comprises:
matching the clauses to be checked with the trust clauses in the trust clause library;
then, the determining at least one clause revision record matching the to-be-reviewed clause in the clause revision library includes:
and when the trust clause base does not have the trust clause matched with the clause to be checked, determining at least one clause revision record matched with the clause to be checked in a clause revision base.
8. The method of claim 7, further comprising:
acquiring a collection clause library, wherein the collection clause library comprises a plurality of candidate clauses;
clustering the candidate clauses according to the semantics of the candidate clauses to obtain a plurality of clusters, wherein the semantics of the candidate clauses in each cluster are similar, and the semantics of the candidate clauses in different clusters are different;
and determining candidate terms meeting the trust condition in each class cluster as trust terms, and adding the trust terms into the trust term library.
9. An apparatus for revising clauses, the apparatus comprising:
the first acquisition module is used for acquiring clauses to be checked;
the first matching module is used for determining at least one clause revision record matched with the clause to be reviewed in a clause revision library, wherein the clause revision library comprises a plurality of clause revision records, and each clause revision record comprises an original clause and revision content aiming at the original clause;
and the revision module is used for revising the clauses to be reviewed by utilizing the at least one clause revision record.
10. A computing device, the device comprising a processor and a memory:
the memory is used for storing a computer program;
the processor is configured to perform the method of any one of claims 1-8 according to the computer program.
11. A computer-readable storage medium, characterized in that the computer-readable storage medium is used to store a computer program for performing the method of any of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111565016.3A CN114254617A (en) | 2021-12-20 | 2021-12-20 | Method, device, computing equipment and storage medium for revising clauses |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111565016.3A CN114254617A (en) | 2021-12-20 | 2021-12-20 | Method, device, computing equipment and storage medium for revising clauses |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114254617A true CN114254617A (en) | 2022-03-29 |
Family
ID=80796118
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111565016.3A Pending CN114254617A (en) | 2021-12-20 | 2021-12-20 | Method, device, computing equipment and storage medium for revising clauses |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114254617A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115618844A (en) * | 2022-12-20 | 2023-01-17 | 江苏龙虎网信息科技股份有限公司 | Content auditing system and method based on artificial intelligence |
-
2021
- 2021-12-20 CN CN202111565016.3A patent/CN114254617A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115618844A (en) * | 2022-12-20 | 2023-01-17 | 江苏龙虎网信息科技股份有限公司 | Content auditing system and method based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11176154B1 (en) | Collaborative dataset management system for machine learning data | |
| US20200202071A1 (en) | Content scoring | |
| CN110543374B (en) | Centralized data coordination using artificial intelligence mechanism | |
| US20190188326A1 (en) | Domain specific natural language understanding of customer intent in self-help | |
| CN109190098A (en) | A kind of document automatic creation method and system based on natural language processing | |
| CN111767716B (en) | Method and device for determining enterprise multi-level industry information and computer equipment | |
| CN105590055A (en) | Method and apparatus for identifying trustworthy user behavior in network interaction system | |
| CN109857846B (en) | Method and device for matching user question and knowledge point | |
| CN113256383B (en) | Recommendation method and device for insurance products, electronic equipment and storage medium | |
| CN114265967A (en) | Sensitive data security level marking method and device | |
| CN111144087A (en) | Enterprise legal flow assistant decision-making system and method based on artificial intelligence | |
| CN113538154A (en) | Risk object identification method and device, storage medium and electronic equipment | |
| US20190295110A1 (en) | Performance analytics system for scripted media | |
| CN109408643B (en) | Fund similarity calculation method, system, computer equipment and storage medium | |
| CN114254617A (en) | Method, device, computing equipment and storage medium for revising clauses | |
| CN114330476A (en) | Model training method for media content recognition and media content recognition method | |
| Li et al. | Reverse engineering variability from requirement documents based on probabilistic relevance and word embedding | |
| CN116228402A (en) | Financial credit investigation feature warehouse technical support system | |
| CN115526315A (en) | Generation method and device of rating card model | |
| CN109242690A (en) | Finance product recommended method, device, computer equipment and readable storage medium storing program for executing | |
| CN112685389B (en) | Data management method, data management device, electronic device, and storage medium | |
| CN115587192A (en) | Relationship information extraction method, device and computer readable storage medium | |
| US20210073247A1 (en) | System and method for machine learning architecture for interdependence detection | |
| US11544327B2 (en) | Method and system for streamlined auditing | |
| US20230127714A1 (en) | Systems and methods for consensus in a blockchain network |
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 |