CN113392977A - Method, apparatus and storage medium for locating modeling errors - Google Patents
Method, apparatus and storage medium for locating modeling errors Download PDFInfo
- Publication number
- CN113392977A CN113392977A CN202010171922.4A CN202010171922A CN113392977A CN 113392977 A CN113392977 A CN 113392977A CN 202010171922 A CN202010171922 A CN 202010171922A CN 113392977 A CN113392977 A CN 113392977A
- Authority
- CN
- China
- Prior art keywords
- information
- error
- entity
- model
- target model
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Computational Linguistics (AREA)
- Databases & Information Systems (AREA)
- Debugging And Monitoring (AREA)
Abstract
The present disclosure provides a method, an apparatus, and a storage medium for locating modeling errors, wherein the method comprises: setting a model operating environment, and adding components required for establishing a target model in the model operating environment; establishing a model flow chart corresponding to the target model based on the component, and performing parameter configuration on the component to establish the target model; running the target model, acquiring first log information of the target model, and extracting error query information from the first log information; inputting the error query information into a preset knowledge graph, determining whether the target model has errors or not by using the knowledge graph, and acquiring corresponding error positioning information. The method, the device and the storage medium disclosed by the invention have the advantages that the knowledge graph is generated based on the log information, the source of errors can be deduced, the error positioning range is narrowed, root nodes causing error problems are found, the problem finding and diagnosing efficiency is improved, and the model development efficiency can be obviously accelerated.
Description
Technical Field
The present invention relates to the field of computer technologies, and in particular, to a method, an apparatus, and a storage medium for locating modeling errors.
Background
When a machine learning platform creates a model project, an experiment needs to be created first, then the component is dragged from the component column according to the requirement of the experiment, and then the experiment is operated after the component is connected and the parameters are configured. At present, after each experiment is created, no part for quickly positioning error nodes and providing error modification suggestions is provided, and if errors occur, the errors are difficult to be checked, the problems are quickly found and solved. Therefore, a technical solution for locating modeling errors is needed.
Disclosure of Invention
In view of the above, an object of the present invention is to provide a method, an apparatus and a storage medium for locating modeling errors.
According to an aspect of the present disclosure, there is provided a method for locating modeling errors, comprising: setting a model operating environment, and adding components required for establishing a target model in the model operating environment; establishing a model flow chart corresponding to the target model based on the component, and performing parameter configuration on the component to establish the target model; running the target model, acquiring first log information of the target model, and extracting error query information from the first log information; inputting the error query information into a preset knowledge graph, determining whether the target model has errors or not by using the knowledge graph, and acquiring corresponding error positioning information.
Optionally, before extracting the error query information from the first log information, a step of constructing the knowledge-graph is further included, the step including: acquiring second log information corresponding to the target model, and extracting error information from the second log information; wherein the error information includes: abnormal information and alarm information; acquiring a logic relation and a causal relation corresponding to the error information; acquiring knowledge graph triples according to the logic relation, the causal relation and the error information; wherein the knowledge-graph triplets include: a first entity, a second entity, and relationship information between the first entity and the second entity; and constructing the knowledge graph based on the knowledge graph triples.
Optionally, the first log information and the second log information include: log files corresponding to nodes of the model flow graph.
Optionally, obtaining a knowledge graph triple according to the logical relationship, the causal relationship, and the error information includes: extracting the first entity and the relationship information from the error information; wherein the first entity comprises: first node, error entity information in the model flow chart; acquiring error source information corresponding to the error information based on the logic relationship and the causal relationship, and generating the second entity according to the error source information; wherein the second entity comprises: error location information; the error positioning information includes: a second node in the model flowchart, error cause information.
Optionally, normalizing the first entity, the relationship information, and the second entity; and generating the knowledge graph triple according to the first entity, the relationship information and the second entity after normalization processing.
Optionally, inputting the error query information into a preset knowledge graph, and determining whether an error occurs in the operation of the target model and acquiring corresponding error positioning information by using the knowledge graph includes: normalizing the error query information; generating a query statement with a triple structure according to the error query information subjected to the normalization processing; and inputting the query sentence into the knowledge graph for querying to obtain the second entity.
Optionally, the second entity is input into a preset expert system, and an error elimination scheme output by the expert system and corresponding to the error query information is obtained.
Optionally, the configuration information of the component is modified based on the error exclusion scheme, and a new target model is generated and run.
According to another aspect of the present disclosure, there is provided an apparatus for locating modeling errors, comprising: the component processing module is used for setting a model operating environment and adding components required for establishing a target model in the model operating environment; the model establishing module is used for establishing a model flow chart corresponding to the target model based on the component, and carrying out parameter configuration on the component so as to establish the target model; the information acquisition module is used for operating the target model, acquiring first log information of the target model and extracting error query information from the first log information; and the error positioning module is used for inputting the error query information into a preset knowledge map, determining whether the target model has errors or not by using the knowledge map and acquiring corresponding error positioning information.
Optionally, the map generation module is configured to construct the knowledge-map before extracting the incorrect query information from the first log information, and includes: an error processing unit, configured to acquire second log information corresponding to the target model, and extract error information from the second log information; wherein the error information includes: abnormal information and alarm information; the map processing unit is used for acquiring a logic relationship and a causal relationship corresponding to the error information and acquiring a knowledge map triple according to the logic relationship, the causal relationship and the error information; wherein the knowledge-graph triplets include: a first entity, a second entity, and relationship information between the first entity and the second entity; and constructing the knowledge graph based on the knowledge graph triples.
Optionally, the first log information and the second log information include: log files corresponding to nodes of the model flow graph.
Optionally, the map processing unit is specifically configured to extract the first entity and the relationship information from the error information; wherein the first entity comprises: first node, error entity information in the model flow chart; acquiring error source information corresponding to the error information based on the logic relationship and the causal relationship, and generating the second entity according to the error source information; wherein the second entity comprises: error location information; the error positioning information includes: a second node in the model flowchart, error cause information.
Optionally, the graph processing unit is further configured to perform normalization processing on the first entity, the relationship information, and the second entity; and generating the knowledge graph triple according to the first entity, the relationship information and the second entity after normalization processing.
Optionally, the error locating module is configured to perform normalization processing on the error query information; generating a query statement with a triple structure according to the error query information subjected to the normalization processing; and inputting the query sentence into the knowledge graph for querying to obtain the second entity.
Optionally, the elimination scheme obtaining module is configured to input the second entity into a preset expert system, and obtain an error elimination scheme output by the expert system and corresponding to the error query information.
Optionally, the model updating module is configured to modify the configuration information of the component based on the error exclusion scheme, and generate and run a new target model.
According to yet another aspect of the present disclosure, there is provided an apparatus for locating modeling errors, comprising: a memory; and a processor coupled to the memory, the processor configured to perform the method as described above based on instructions stored in the memory.
According to yet another aspect of the present disclosure, a computer-readable storage medium is provided, which stores computer instructions for execution by a processor to perform the method as described above.
According to the method, the device and the storage medium for positioning the modeling errors, the knowledge graph is generated based on the log information, the alarm information of model operation and the like, the operation log and the like are sent to the log graph in real time, the source of the errors can be deduced, the error positioning range is narrowed, root nodes causing the error problems are found, the problem finding and diagnosing efficiency is improved, and the model development efficiency can be remarkably accelerated.
Drawings
In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without inventive exercise.
FIG. 1 is a flow diagram of one embodiment of a method for locating modeling errors according to the present disclosure;
FIG. 2 is a schematic flow diagram of construction of a knowledge graph in one embodiment of a method for locating modeling errors according to the present disclosure;
FIG. 3 is a schematic flow diagram of obtaining knowledge-graph triples in one embodiment of a method for locating modeling errors according to the present disclosure;
FIG. 4A is a schematic flow chart diagram illustrating obtaining error localization information in one embodiment of a method for localizing modeling errors according to the present disclosure; FIG. 4B is a schematic diagram of an example of a model operating environment;
FIG. 5 is a block diagram illustration of one embodiment of an apparatus for locating modeling errors in accordance with the present disclosure;
FIG. 6 is a block schematic diagram of another embodiment of an apparatus for locating modeling errors according to the present disclosure;
FIG. 7 is a block schematic diagram of an atlas generation module in another embodiment of an apparatus for localizing modeling errors according to the present disclosure;
FIG. 8 is a block diagram illustration of yet another embodiment of an apparatus for locating modeling errors according to the present disclosure.
Detailed Description
The present disclosure now will be described more fully hereinafter with reference to the accompanying drawings, in which exemplary embodiments of the disclosure are shown. The technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the drawings in the embodiments of the present disclosure, and it is obvious that the described embodiments are only a part of the embodiments of the present disclosure, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present disclosure.
The terms "first", "second", and the like are used hereinafter only for descriptive distinction and not for other specific meanings.
A great number of tools in the field of traditional machine learning are programming machine learning libraries, such as R, Python, deep learning libraries caffe, tensorflow, etc., and these machine learning tools have certain thresholds, have high learning curves, and are not suitable for general business and data analysts.
At present, when a common machine learning platform creates a model project, an experiment needs to be created first, then a component is dragged from a component column according to the requirement of the experiment, and then the experiment is operated after the component is connected and parameter configuration is carried out. After an experiment is created every time, a part for quickly positioning error nodes and providing error modification suggestions is not provided, so that in the process of visual AI modeling, if errors occur, the errors are difficult to be eliminated, the problems are quickly found and solved, the whole model needs to be built again, time is consumed, inconvenience is caused, and the usability of the visual machine learning platform is poor.
FIG. 1 is a schematic flow chart diagram of one embodiment of a method for locating modeling errors according to the present disclosure, as shown in FIG. 1:
And 102, establishing a model flow chart corresponding to the target model based on the component, and performing parameter configuration on the component to establish the target model.
In one embodiment, a predetermined graphical algorithm component can be selected, dragged to a design area for visual modeling, a visual model flow chart is built, and parameters of algorithm nodes are adjusted.
The model operation environment can be a machine learning model project operation environment, the target model is a model project, and the model project refers to a data workflow or a data application or an experimental model project which is built on a machine learning platform by a machine learning platform user and consists of components. Through the machine learning platform, reasonable model items can be obtained according to the existing data.
When a machine learning platform creates a model project, an experiment needs to be created first, then components are dragged from a component bar according to the requirements of the experiment, a model flow chart corresponding to a target model is established, and the experiment is operated after the parameters of the components are configured. A machine learning execution environment (model execution environment) is a virtual environment in a machine learning platform for creating and executing model items, in which the model items can be created and executed.
The model flowchart may be a visual model flowchart, and each component in the model project constitutes a node of the model project, for example, one node of the visual model flowchart is an algorithm training node, and parameters such as the number of trees, the tree depth, a training column, and a target column are configured.
A component refers to an element of operation that can be invoked for execution on a machine learning platform to represent various algorithms or data sources. Such as data import and export, data processing, data analysis, model project training or prediction, etc. Adding components into the machine learning model project operating environment, establishing input and output links of the required components according to a set sequence to create a model project (target model)
The first log information may be log information of a node in the visualization model flowchart, the first log information includes information about an exception, an alarm, and the like, and the information about the exception, the alarm, and the like is extracted from the first log information to generate error query information. The error query information can be extracted from the first log information using a variety of existing methods.
And 104, inputting the error query information into a preset knowledge graph, determining whether the target model has errors or not by using the knowledge graph, and acquiring corresponding error positioning information.
The knowledge map is a semantic network formed by connecting knowledge points and is used for knowledge reasoning and automatic question answering. Inputting information such as abnormity, alarm and the like into a preset knowledge map, determining whether the target model has errors by using the knowledge map, and acquiring corresponding error positioning information; and providing a visual interface, displaying error positioning information and the like in the visual interface, and adjusting codes according to the error positioning information.
Based on the method for positioning modeling errors in the embodiment, high-quality machine learning modeling can be realized, developers can develop and train machine learning without coding, and the model development efficiency can be remarkably improved.
In one embodiment, the knowledge-graph is pre-constructed prior to extracting the erroneous-query information from the first log information. The knowledge graph can be constructed in advance, and can also be constructed in the training and running processes of the target model. FIG. 2 is a schematic flow diagram of construction of a knowledge graph in one embodiment of a method for locating modeling errors according to the present disclosure, as shown in FIG. 2:
The first log information and the second log information include: log files corresponding to the nodes of the model flow graph. For example, model training is performed, each node generates second log information in the training process, and logs of the nodes can be integrated together according to a workflow and integrated into a log graph.
The associated error information appearing in the log can be analyzed by using Apriori or FP-growth and other algorithms to obtain related logical relations and causal relations. The Apriori algorithm is an association rule mining algorithm, and the FP-Growth algorithm is an association analysis algorithm.
In one embodiment, the minimum composition unit for constructing the knowledge graph is a knowledge graph triple, the knowledge graph triple comprises two knowledge graph entities and an attribute relationship between the two knowledge graph entities, and the basic form of the knowledge graph triple is as follows: entity-relationship-entity. According to the logic relation and the causal relation, error information can be processed in real time, knowledge extraction is carried out, information such as entities and relations is obtained and sent to a knowledge map database for construction of a knowledge map.
And step 204, constructing the knowledge graph based on the knowledge graph triples.
And extracting knowledge graph triples from the error information to construct a knowledge graph. The knowledge graph constructed by the triple data has the logical structure capability of knowledge inference, and the knowledge graph constructed by the data triple can be used for positioning errors.
In one embodiment, a knowledge graph can be dynamically generated based on the second log information, and the cause problem of the error information can be located through the knowledge graph. In the training process, each node generates log information, logs of each node are integrated together according to a working process and integrated into a log map, sources of error information are deduced according to incidence relations among the nodes in a visual model flow chart, logical relations and causal relations among the nodes are mined, the range of positioning problems is narrowed, and root nodes causing the problems are found. In the visual machine learning development model, visual error troubleshooting and interactive debugging can be provided, high-quality machine learning modeling is realized, developers can develop and train machine learning without coding, and the model development efficiency is accelerated.
In one embodiment, obtaining knowledge-graph triples may use a variety of methods. FIG. 3 is a schematic flow diagram of obtaining knowledge-graph triples in an embodiment of a method for locating modeling errors according to the present disclosure, as shown in FIG. 3:
By normalizing the first entity, the relationship information and the second entity, different descriptions with the same meaning in the first entity, the relationship information and the second entity can be integrated, so that redundant descriptions with the same meaning can be eliminated, and the first entity, the relationship information and the second entity become normalized standard entities and standard relationship texts.
And 304, generating a knowledge graph triple according to the first entity, the relationship information and the second entity after normalization processing.
FIG. 4A is a schematic flow chart of obtaining error localization information in an embodiment of a method for localizing modeling errors according to the present disclosure, as shown in FIG. 4:
By normalizing the error query information, different descriptions with the same meaning in the error query information can be integrated, so that redundant descriptions with the same meaning can be eliminated and become normalized standard entities and standard relation texts.
As shown in fig. 4B, a model operating environment is set, components required for building the target model are added to the model operating environment, a visual model flowchart corresponding to the target model is built based on the components, and nodes in the visual model flowchart include: data set, data preprocessing, algorithm training, model evaluation and the like. Carrying out parameter configuration on the component to establish a target model; acquiring first log information of algorithm training nodes, and extracting error query information from the first log information, wherein the error query information comprises: the number of trees of the first node is smaller than a preset set value, and the like.
And 403, inputting the query sentence into the knowledge graph for query to obtain a second entity.
In an embodiment, the query statement may be SPARQL, which is to perform filtering processing according to the entity words and the relation words and extract answers from the knowledge graph. The normalized error query information is that the number of trees of the first node is smaller than a preset value, a query statement ' SELECT X WHERE (the number of trees of the first node is smaller than the preset value, X) with a triple structure is generated by using the existing multiple methods and is input into a knowledge graph for query, a second entity at the other end, which has a relation of ' the number of trees of the first node is smaller than the preset value ', with the entity is found, and the second entity comprises error positioning information; the error positioning information includes: the second node in the model flowchart, error cause information, for example, the error cause information includes: the reason information corresponding to "the number of trees of the first node is smaller than the preset value" may be the error location information, which may be the second node + the reason information corresponding to "the number of trees is smaller than the preset value".
The expert system is an intelligent computer program system, which contains a large amount of knowledge and experience of expert level in a certain field and can process the problem in the field by utilizing the knowledge of human experts and a problem solving method. The expert system performs reasoning and judgment according to knowledge and experience provided by one or more experts in a certain field, and simulates the decision process of human experts so as to solve the complex problems needing to be processed by the human experts.
For example, an expert system may be preset, the second node in the model flowchart, the reason information corresponding to the "number of trees of the first node is less than the preset value", and the like may be input to the expert system, and the corresponding error elimination scheme may be output. A visual interface can be provided, error positioning information, error elimination schemes and the like can be displayed on the visual interface, and suggestions can be provided for the user.
The model developer can be a front-line operation and maintenance engineer, can be a solution person in a specific industry, only needs to understand the service, and does not need an IDE tool to debug the model at a code level. Based on the method for positioning modeling errors, a knowledge graph is generated based on log information, alarm information of model operation and the like, operation logs and the like are sent to the log graph in real time, the logical relation and the causal relation between nodes are mined according to the incidence relation between the nodes, the source of errors can be deduced, the range of positioning problems is narrowed, root cause nodes causing the problems are found, and the efficiency of finding and diagnosing the problems is improved.
In one embodiment, as shown in FIG. 5, the present disclosure provides an apparatus 50 for locating modeling errors, comprising: a component processing module 51, a model building module 52, an information acquisition module 53 and an error localization module 54.
The component processing module 51 sets a model execution environment, and adds components necessary for creating the target model to the model execution environment. The model building module 52 builds a model flow chart corresponding to the target model based on the components, and performs parameter configuration on the components to build the target model. The information obtaining module 52 runs the target model, obtains first log information of the target model, and extracts error query information from the first log information. The error localization module 54 inputs the error query information into a preset knowledge map, determines whether the target model has errors using the knowledge map, and obtains corresponding error localization information.
In one embodiment, as shown in FIG. 6, the means 50 for locating modeling errors further comprises: a map generation module 55, an exclusion scheme acquisition module 56, and a model update module 57. The map generation module 55 constructs a knowledge map prior to extracting the incorrect query information from the first log information.
As shown in fig. 7, the map generation module 55 includes: an error processing unit 551 and an atlas processing unit 552. The error processing unit 551 acquires second log information corresponding to the target model, and extracts error information from the second log information; wherein the error information includes: abnormal information, alarm information, etc.
The map processing unit 552 obtains a logical relationship and a causal relationship corresponding to the error information, and obtains a knowledge map triple according to the logical relationship, the causal relationship and the error information; wherein, the knowledge-graph triplets include: a first entity, a second entity, and relationship information between the first entity and the second entity; the atlas handling unit 552 constructs a knowledge-atlas based on knowledge-atlas triples.
The first log information and the second log information include: log files corresponding to the nodes of the model flow graph. The graph processing unit 552 extracts the first entity and the relationship information from the error information; wherein the first entity comprises: the first node in the model flow chart, error entity information, etc.
The map processing unit 552 obtains error source information corresponding to the error information based on the logical relationship and the causal relationship, and generates a second entity according to the error source information; wherein the second entity comprises: error location information; the error positioning information includes: a second node in the model flow chart, error cause information, etc.
The graph processing unit 552 normalizes the first entity, the relationship information, and the second entity; the map processing unit 552 generates a knowledge map triple from the first entity, the relationship information, and the second entity after the normalization processing.
The error positioning module 54 normalizes the error query information; the error positioning module 54 generates a query statement with a triple structure according to the error query information after normalization processing; the error localization module 54 inputs the query statement into the knowledge-graph for query to obtain the second entity.
The elimination scheme obtaining module 56 inputs the second entity into a preset expert system, and obtains an error elimination scheme corresponding to the error query information output by the expert system. The model update module 57 modifies the configuration information of the components based on the error exclusion scheme, generating and running a new target model.
FIG. 8 is a block diagram illustration of yet another embodiment of an apparatus for locating modeling errors according to the present disclosure. As shown in fig. 8, the apparatus may include a memory 81, a processor 82, a communication interface 83, and a bus 84. The memory 81 is used for storing instructions, the processor 82 is coupled to the memory 81, and the processor 82 is configured to perform the method for locating modeling errors described above based on the instructions stored by the memory 81.
The memory 81 may be a high-speed RAM memory, a non-volatile memory (non-volatile memory), or the like, and the memory 81 may be a memory array. The storage 81 may also be partitioned and the blocks may be combined into virtual volumes according to certain rules. The processor 82 may be a central processing unit CPU, or an application Specific Integrated circuit asic, or one or more Integrated circuits configured to implement the methods of the present disclosure for locating modeling errors.
In one embodiment, the present disclosure provides a computer-readable storage medium having stored thereon computer instructions for execution by a processor to perform a method as in any of the above embodiments.
The method, the device and the storage medium for positioning modeling errors provided in the above embodiments generate a knowledge graph based on log information, send alarm information of model operation and the like, an operation log and the like to the log graph in real time, mine logical relationships and causal relationships between nodes according to the association relationship between each node in the log graph, can infer the source of an error, narrow the range of error positioning, find root nodes causing error problems, improve the efficiency of problem discovery and diagnosis, realize high-quality machine learning modeling, allow developers to develop and train machine learning without coding, significantly accelerate the efficiency of model development, and improve the use experience of users.
The method and system of the present disclosure may be implemented in a number of ways. For example, the methods and systems of the present disclosure may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above-described order for the steps of the method is for illustration only, and the steps of the method of the present disclosure are not limited to the order specifically described above unless specifically stated otherwise. Further, in some embodiments, the present disclosure may also be embodied as programs recorded in a recording medium, the programs including machine-readable instructions for implementing the methods according to the present disclosure. Thus, the present disclosure also covers a recording medium storing a program for executing the method according to the present disclosure.
The description of the present disclosure has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the disclosure in the form disclosed. Many modifications and variations will be apparent to practitioners skilled in this art. The embodiment was chosen and described in order to best explain the principles of the disclosure and the practical application, and to enable others of ordinary skill in the art to understand the disclosure for various embodiments with various modifications as are suited to the particular use contemplated.
Claims (18)
1. A method for locating modeling errors, comprising:
setting a model operating environment, and adding components required for establishing a target model in the model operating environment;
establishing a model flow chart corresponding to the target model based on the component, and performing parameter configuration on the component to establish the target model;
running the target model, acquiring first log information of the target model, and extracting error query information from the first log information;
inputting the error query information into a preset knowledge graph, determining whether the target model has errors or not by using the knowledge graph, and acquiring corresponding error positioning information.
2. The method of claim 1, further comprising, prior to extracting erroneous query information from the first log information, the step of constructing the knowledge-graph, the step comprising:
acquiring second log information corresponding to the target model, and extracting error information from the second log information; wherein the error information includes: abnormal information and alarm information;
acquiring a logic relation and a causal relation corresponding to the error information;
acquiring knowledge graph triples according to the logic relation, the causal relation and the error information; wherein the knowledge-graph triplets include: a first entity, a second entity, and relationship information between the first entity and the second entity;
and constructing the knowledge graph based on the knowledge graph triples.
3. The method of claim 2, wherein,
the first log information and the second log information include: log files corresponding to nodes of the model flow graph.
4. The method of claim 3, wherein obtaining knowledge-graph triples based on the logical and causal relationships and the error information comprises:
extracting the first entity and the relationship information from the error information; wherein the first entity comprises: first node, error entity information in the model flow chart;
acquiring error source information corresponding to the error information based on the logic relationship and the causal relationship, and generating the second entity according to the error source information; wherein the second entity comprises: error location information; the error positioning information includes: a second node in the model flowchart, error cause information.
5. The method of claim 4, further comprising:
normalizing the first entity, the relationship information and the second entity;
and generating the knowledge graph triple according to the first entity, the relationship information and the second entity after normalization processing.
6. The method of claim 5, wherein inputting the error query information into a preset knowledge graph, wherein using the knowledge graph to determine whether the target model is in error and obtain corresponding error localization information comprises:
normalizing the error query information;
generating a query statement with a triple structure according to the error query information subjected to the normalization processing;
and inputting the query sentence into the knowledge graph for querying to obtain the second entity.
7. The method of claim 6, further comprising:
and inputting the second entity into a preset expert system, and acquiring an error elimination scheme which is output by the expert system and corresponds to the error query information.
8. The method of claim 7, further comprising:
and modifying the configuration information of the component based on the error elimination scheme, and generating and running a new target model.
9. An apparatus for locating modeling errors, comprising:
the component processing module is used for setting a model operating environment and adding components required for establishing a target model in the model operating environment;
the model establishing module is used for establishing a model flow chart corresponding to the target model based on the component, and carrying out parameter configuration on the component so as to establish the target model;
the information acquisition module is used for operating the target model, acquiring first log information of the target model and extracting error query information from the first log information;
and the error positioning module is used for inputting the error query information into a preset knowledge map, determining whether the target model has errors or not by using the knowledge map and acquiring corresponding error positioning information.
10. The apparatus of claim 9, further comprising:
a graph generation module to construct the knowledge graph prior to extracting false query information from the first log information, comprising:
an error processing unit, configured to acquire second log information corresponding to the target model, and extract error information from the second log information; wherein the error information includes: abnormal information and alarm information;
the map processing unit is used for acquiring a logic relationship and a causal relationship corresponding to the error information and acquiring a knowledge map triple according to the logic relationship, the causal relationship and the error information; wherein the knowledge-graph triplets include: a first entity, a second entity, and relationship information between the first entity and the second entity; and constructing the knowledge graph based on the knowledge graph triples.
11. The apparatus of claim 10, wherein,
the first log information and the second log information include: log files corresponding to nodes of the model flow graph.
12. The apparatus of claim 11, wherein,
the map processing unit is specifically configured to extract the first entity and the relationship information from the error information; wherein the first entity comprises: first node, error entity information in the model flow chart; acquiring error source information corresponding to the error information based on the logic relationship and the causal relationship, and generating the second entity according to the error source information; wherein the second entity comprises: error location information; the error positioning information includes: a second node in the model flowchart, error cause information.
13. The apparatus of claim 12, wherein,
the map processing unit is further configured to perform normalization processing on the first entity, the relationship information, and the second entity; and generating the knowledge graph triple according to the first entity, the relationship information and the second entity after normalization processing.
14. The apparatus of claim 13, wherein,
the error positioning module is used for carrying out normalization processing on the error query information; generating a query statement with a triple structure according to the error query information subjected to the normalization processing; and inputting the query sentence into the knowledge graph for querying to obtain the second entity.
15. The apparatus of claim 14, further comprising:
and the elimination scheme acquisition module is used for inputting the second entity into a preset expert system and acquiring the error elimination scheme which is output by the expert system and corresponds to the error query information.
16. The apparatus of claim 15, further comprising:
and the model updating module is used for modifying the configuration information of the component based on the error elimination scheme, and generating and operating a new target model.
17. An apparatus for remapping a network slice, comprising:
a memory; and a processor coupled to the memory, the processor configured to perform the method of any of claims 1-8 based on instructions stored in the memory.
18. A computer-readable storage medium having stored thereon computer instructions for execution by a processor of the method of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010171922.4A CN113392977A (en) | 2020-03-12 | 2020-03-12 | Method, apparatus and storage medium for locating modeling errors |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010171922.4A CN113392977A (en) | 2020-03-12 | 2020-03-12 | Method, apparatus and storage medium for locating modeling errors |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113392977A true CN113392977A (en) | 2021-09-14 |
Family
ID=77615781
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010171922.4A Pending CN113392977A (en) | 2020-03-12 | 2020-03-12 | Method, apparatus and storage medium for locating modeling errors |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113392977A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115510782A (en) * | 2022-08-31 | 2022-12-23 | 芯华章科技股份有限公司 | Method for locating verification error, electronic device and storage medium |
-
2020
- 2020-03-12 CN CN202010171922.4A patent/CN113392977A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115510782A (en) * | 2022-08-31 | 2022-12-23 | 芯华章科技股份有限公司 | Method for locating verification error, electronic device and storage medium |
| CN115510782B (en) * | 2022-08-31 | 2024-04-26 | 芯华章科技股份有限公司 | Method for locating verification errors, electronic device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jannach et al. | Avoiding, finding and fixing spreadsheet errors–A survey of automated approaches for spreadsheet QA | |
| US6965887B2 (en) | Rule processing methods for automating a decision and assessing satisfiability of rule-based decision diagrams | |
| US7809669B2 (en) | Rule processing system for determining a result response | |
| US7587379B2 (en) | Method and system for capturing business rules for automated decision procession | |
| CN113742205A (en) | Code vulnerability intelligent detection method based on man-machine cooperation | |
| CN113392977A (en) | Method, apparatus and storage medium for locating modeling errors | |
| CN111694570A (en) | JavaScript function parameter mismatching detection method based on static program analysis | |
| US11436514B2 (en) | Designing plans using requirements knowledge graph | |
| Maita et al. | Mining unstructured processes: An exploratory study on a distance learning domain | |
| Koivisto et al. | Evaluating CodeClusters for effectively providing feedback on code submissions | |
| CN115454702A (en) | Log fault analysis method and device, storage medium and electronic equipment | |
| Esgin et al. | Sequence alignment adaptation for process diagnostics and delta analysis | |
| US8019716B2 (en) | Reflective processing of TMK hierarchies | |
| CN111913872A (en) | Software static inspection warning sequencing optimization method based on defect prediction | |
| CN113721977B (en) | Programming data processing method and device | |
| Orvalho et al. | Graph Neural Networks for Mapping Variables Between Programs | |
| CA2479342A1 (en) | Rule processing system and method | |
| Orvalho et al. | Graph Neural Networks For Mapping Variables Between Programs--Extended Version | |
| CN113127341B (en) | Incremental code defect detection method and system based on graph network model | |
| Schekotihin et al. | AI-based Spreadsheet Debugging | |
| CN117251477B (en) | Standardized inspection method and system for data development script based on SQL (structured query language) analysis | |
| Xu et al. | A log-linear probabilistic model for prioritizing extract method refactorings | |
| Fouladfard | Software security flaw prediction using rich contextualized language use vectors: A case study on the linux kernel | |
| WO2024057046A1 (en) | A method for computing and merging staticand dynamic metricsof a computer program, a method for training a machine learning model based on hybrid metrics, a machine learning model, and a method for using the machine learning model | |
| CN118012767A (en) | Multi-coverage-rate-use-case-oriented generation and multi-method-fusion-oriented software bug positioning method |
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 |