Engineering document quality evaluation method and device, equipment and storage medium

Abstract

The invention discloses an engineering document quality evaluation method, device and equipment and a storage medium, and is applied to the field of financial science and technology. The method comprises the steps of obtaining a to-be-evaluated engineering test document, performing text segmentation processing on the engineering test document, and performing semantic feature extraction on segmented text units to generate semantic context features; based on a bidirectional long-short-term memory network, performing dependency feature extraction on the semantic context features, endowing a dependency feature extraction result with a position attention weight through a gating mechanism, and generating a domain entity tag space; based on a legal label transfer rule and the domain entity label space, determining an optimal label sequence, and based on the optimal label sequence, determining a structured entity tree corresponding to the engineering test document; and performing quality evaluation on the engineering test document based on the structured entity tree. According to the technical scheme, the quality evaluation precision of the engineering test document is improved.

Description

Technical Field

[0001] This application relates to the field of computer technology and can be applied to the field of financial technology. Specifically, it relates to a method, apparatus, device, and storage medium for quality assessment of engineering documents. Background Technology

[0002] Currently, in the project development and testing phases of various application fields, the quality of the application project's test engineering documentation is particularly important, as it directly affects the test quality of the application project.

[0003] Therefore, there is an urgent need for a quality assessment method for engineering documents applicable to the field of software testing, in order to improve the document quality of engineering documents in the field of software testing. Summary of the Invention

[0004] This application provides a method, apparatus, device, and storage medium for quality assessment of engineering documents, in order to improve the document quality of engineering documents in the field of software testing.

[0005] According to one aspect of this application, a method for quality assessment of engineering documents is provided, the method comprising:

[0006] The engineering test document to be evaluated is obtained, the engineering test document is processed by text segmentation, and semantic features are extracted from the segmented text units to generate semantic context features; wherein, the semantic context features are context semantic features with at least two feature dimensions.

[0007] Based on a bidirectional long short-term memory network, dependency feature extraction is performed on the semantic context features, and positional attention weights are assigned to the dependency feature extraction results through a gating mechanism to generate a domain entity label space.

[0008] Based on the legal tag transfer rules and the domain entity tag space, the optimal tag sequence is determined, and based on the optimal tag sequence, the structured entity tree corresponding to the engineering test document is determined.

[0009] The quality of the engineering test documents is assessed based on the structured entity tree.

[0010] According to another aspect of this application, an engineering document quality assessment apparatus is provided, the apparatus comprising:

[0011] The semantic feature extraction module is used to acquire the engineering test document to be evaluated, perform text segmentation on the engineering test document, and extract semantic features from the segmented text units to generate semantic context features; wherein, the semantic context features are context semantic features with at least two feature dimensions.

[0012] The label space generation module is used to extract dependent features from the semantic context features based on a bidirectional long short-term memory network, and to assign positional attention weights to the dependent feature extraction results through a gating mechanism to generate a domain entity label space.

[0013] The structure tree generation module is used to determine the optimal tag sequence based on the legal tag transfer rules and the domain entity tag space, and to determine the structured entity tree corresponding to the engineering test document based on the optimal tag sequence.

[0014] The quality assessment module is used to assess the quality of the engineering test documents based on the structured entity tree.

[0015] According to another aspect of this application, an electronic device is provided, the electronic device comprising:

[0016] One or more processors;

[0017] Memory, used to store one or more programs;

[0018] When the one or more programs are executed by the one or more processors, the one or more processors implement any of the engineering document quality assessment methods provided in the embodiments of this application.

[0019] According to another aspect of this application, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements any of the engineering document quality assessment methods provided in the embodiments of this application.

[0020] According to another aspect of this application, a computer program product is provided, including a computer program that, when executed by a processor, implements any of the engineering document quality assessment methods provided in the embodiments of this application.

[0021] This application improves the accuracy of entity boundary determination and thus the accuracy of quality assessment of engineering test documents by extracting semantic context features and dependency features from the engineering test documents twice. Attached Figure Description

[0022] Figure 1 This is a flowchart of a method for quality assessment of engineering documents according to Embodiment 1 of this application;

[0023] Figure 2 This is a flowchart of a method for quality assessment of engineering documents according to Embodiment 2 of this application;

[0024] Figure 3This is a schematic diagram of the structure of an engineering document quality assessment device according to Embodiment 3 of this application;

[0025] Figure 4 This is a schematic diagram of the structure of an electronic device that implements the engineering document quality assessment method of Embodiment 4 of this application. Detailed Implementation

[0026] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0027] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0028] Example 1

[0029] Figure 1 This is a flowchart of a quality assessment method for engineering documents according to Embodiment 1 of this application. This embodiment is applicable to the quality assessment of application test documents in application development testing. It can be performed by an engineering document quality assessment device, which can be implemented in hardware and / or software and can be configured in a computer device, such as a server. Figure 1 As shown, the method includes:

[0030] S110. Obtain the engineering test document to be evaluated, perform text segmentation on the engineering test document, and extract semantic features from the segmented text units to generate semantic context features.

[0031] The semantic context features can be contextual semantic features that include at least two feature dimensions; they can be used to characterize the features of text units in multiple dimensions such as terminology, logic, and intent.

[0032] Engineering test documentation can refer to a document that includes test elements such as test steps, expected results, and prerequisites for application testing.

[0033] Optionally, in a specific embodiment of the present invention, after obtaining the unstructured engineering test document to be evaluated, the engineering test document can first be preprocessed to convert the unstructured test document into a structured standardized sequence in order to eliminate format noise and domain-specific interference.

[0034] Specifically, the engineering test document can be split into its first stage according to heading levels. Irrelevant formatting symbols (such as ==, **, and table borders) can be removed, while variable names and parameter values ​​in the code blocks can be retained. This results in the first split. It should be noted that the first split result can refer to a list of structured text fragments.

[0035] Furthermore, a tokenizer model that has learned a domain-specific dictionary can be used to perform a second segmentation on the first segmentation result, obtaining at least one text unit, thus obtaining the second segmentation result. It should be noted that a tokenizer model that has learned a domain-specific dictionary can preserve the original format of code-related terms; for example, TestCase_001 is not segmented, and @Param retains the symbol. Optionally, the domain-specific dictionary can refer to a set that defines common terms in the application testing domain. For example, if the first segmentation result is "After executing TestCase_001, the assertion return code = 200", then the second segmentation result can be ["execution", "TestCase_001", "after", ",", "assertion", "return code", "=", "200"].

[0036] Finally, the results of the second split are mapped to predefined vocabulary IDs, and attention masks are generated based on these IDs. For example, "execute" → 1001, "after" → 2002. Valid vocabulary IDs are represented by a mask of 1, and invalid vocabulary IDs are represented by a mask of 0. It should be noted that a valid vocabulary ID can refer to a predefined ID corresponding to a text unit, while an invalid vocabulary ID can refer to an ID corresponding to a text unit that is not predefined.

[0037] Optionally, in this embodiment of the invention, the extraction of semantic context features can be performed using a bidirectional encoder model with a self-attention mechanism.

[0038] S120. Based on a bidirectional long short-term memory network, dependent features are extracted from semantic context features, and positional attention weights are assigned to the dependent feature extraction results through a gating mechanism to generate a domain entity label space.

[0039] Among them, dependency feature extraction is used to capture forward and backward dependencies.

[0040] Positional attention weights can be used to amplify the feature weights of key step features (such as “step,” “execution,” “verification”) while suppressing interference from irrelevant words (such as “then,” “and”).

[0041] Specifically, a forward long short-term memory (LSTM) network can be used to learn the pre-conditional dependencies of semantic context features, such as the sequential relationship between "Step 1: Start the service" and "Step 2: Connect to the database". A backward long short-term memory (LSTM) network can be used to learn the back-end dependencies of semantic context features, such as the inverse association between "assertion failure" and "step 3 parameter error detected". The forward and backward features output by the forward LSTM network are then fused to obtain the dependency feature extraction results for semantic context features. A gating mechanism is then used to add positional attention weights to the dependency feature extraction results to enhance step sequence recognition. Finally, the dependency feature extraction results with positional attention weights are mapped to the domain entity label space to determine the label score for each entity label corresponding to each text unit. Optionally, the label score for each entity label corresponding to a text unit can be represented by a label feature vector.

[0042] S130. Based on the legal label transfer rules and the domain entity label space, determine the optimal label sequence, and based on the optimal label sequence, determine the structured entity tree corresponding to the engineering test document.

[0043] The structured entity tree is used to represent the text entities identified from the engineering test document.

[0044] Optionally, the optimal label sequence is determined based on the legal label transfer rules and the domain entity label space, including: determining a label transfer probability matrix according to the legal label transfer rules; wherein the label transfer probability matrix is ​​used to characterize the connection probability between different entity labels; and according to the label transfer probability matrix, performing weighted summation on the label feature vectors corresponding to the text units in the domain entity label space, and taking the entity label sequence with the highest path score as the optimal label sequence.

[0045] In the domain entity label space, each text unit has its own corresponding label feature vector. The label feature vector can be used to represent the maximum label probability (label score) of each text unit under at least one entity label. The maximum label probability can be used to represent the likelihood of a text unit belonging to a certain entity label. It should be noted that entity labels can be used to represent the semantic role of a text unit in the engineering test document, such as test cases, test steps, test parameters, assertion conditions, preconditions, expected results, and actual results.

[0046] Specifically, at least one text unit is arranged and combined according to entity labels to generate at least one candidate label sequence. The label scores corresponding to entity labels in each candidate label sequence are weighted and summed using the label transition probability matrix, and the candidate label sequence with the highest score is taken as the optimal label sequence.

[0047] Optionally, the rules for legitimate label transfer can be adapted to the needs of those skilled in the art.

[0048] The optimal label sequence is determined by the legal label transfer rules, which solves the problem of maximizing local probability but being unreasonable overall, and makes the optimal label sequence more in line with language rules.

[0049] Optionally, based on the optimal label sequence, the structured entity tree corresponding to the engineering test document is determined, including: traversing the optimal label sequence; if the currently traversed object is the first target label, then entity extraction is triggered to start; during the entity extraction process, if the currently traversed object is the second target label, then entity extraction is triggered to end; the text units corresponding to the entity label sequence traversed during the entity extraction process are merged, and the structured entity tree corresponding to the engineering test document is generated based on the merged text object and the text attributes of the text units corresponding to the label sequence.

[0050] The first target label can be used to represent the start mark of merging text units, and the second target label can be used to represent the end mark of merging text units.

[0051] It should be noted that a structured entity tree can refer to standardized structured data that can be recognized by machines, and the entity labels of the text units used to merge and generate text objects are the same. The merged text objects can be used as tree nodes in a structured entity tree.

[0052] S140. Based on the structured entity tree, perform quality assessment on engineering test documents.

[0053] This application embodiment extracts semantic context features and dependency features from the engineering test document twice to determine the domain entity label space corresponding to the engineering test document, thereby improving the accuracy of entity boundary determination and thus improving the quality assessment accuracy of the engineering test document.

[0054] Example 2

[0055] Figure 2This is a flowchart of a quality assessment method for engineering documents according to Embodiment 2 of this application. This embodiment further refines the "quality assessment of engineering test documents based on a structured entity tree" based on the technical solutions of the above embodiments. It should be noted that for parts not detailed in this embodiment, please refer to the relevant descriptions in other embodiments. Figure 2 As shown, the method includes:

[0056] S210. Obtain the engineering test document to be evaluated, perform text segmentation on the engineering test document, and extract semantic features from the segmented text units to generate semantic context features.

[0057] S220, based on a bidirectional long short-term memory network, performs dependency feature extraction on semantic context features, and assigns positional attention weights to the dependency feature extraction results through a gating mechanism to generate a domain entity label space.

[0058] S230. Based on the legal label transfer rules and the domain entity label space, determine the optimal label sequence, and based on the optimal label sequence, determine the structured entity tree corresponding to the engineering test document.

[0059] S240. Traverse the tree nodes in the structured entity tree, and cluster the tree nodes according to the label attributes of the nodes to generate at least one tree node group.

[0060] S250. For any two tree nodes in the tree node group, perform term consistency detection based on semantic similarity calculation and character similarity calculation.

[0061] Specifically, by traversing the tree nodes in the structured entity tree, tree nodes with the same entity labels are grouped together to generate at least one tree node group. For any two tree nodes within each group, a terminology consistency check is performed on the corresponding text objects. This check can involve calculating the semantic similarity and character similarity between the two text objects. If the semantic similarity meets a preset semantic similarity threshold, but the character similarity does not meet a preset character similarity threshold, the terminology consistency check result indicates a terminology inconsistency. Optionally, the preset semantic similarity threshold and preset character similarity threshold can be adaptively set according to those skilled in the art.

[0062] S260. Traverse the tree nodes in the structured entity tree, extract at least one target tree node whose entity label attribute is a step label, and construct a target directed acyclic graph for at least one target tree node based on the node attributes of the target tree node.

[0063] The node attributes can include the node value (text object to text content) of the tree node and step attributes, etc. Optionally, the step attributes can include startup steps, configuration steps, execution steps, and verification steps, etc.

[0064] S270. According to the preset test step execution rules, traverse the target directed acyclic graph and perform step timing detection.

[0065] Specifically, a target directed acyclic graph can be constructed for at least one target tree node based on the step sequence indicated by the node value of the target tree node, and the step sequence detection of the target directed acyclic graph can be performed according to the preset test step execution rules and the step attributes of the target tree node.

[0066] Among them, the test step timing detection is used to determine whether there are edge dependency errors and loop risks in the target directed acyclic graph. Optionally, the preset test step execution rules are used to characterize the standard step execution logic. The preset test step execution rules can be adaptively set according to those skilled in the art. For example, the startup step must be the first step type executed in the test step flow and is not allowed to have a preceding step. The preceding step of the configuration step must be a "startup step" or "configuration step". The preceding step of the execution step must be a "configuration step" or "execution step". The preceding step of the verification step must be an "execution step" or "verification step".

[0067] S280. Based on the results of terminology consistency detection and step sequence detection, conduct a quality assessment of the engineering test documents.

[0068] Optionally, based on the results of terminology consistency detection and step sequence detection, the quality assessment of the engineering test document is performed, including: weighting different document defects according to the number of defects with inconsistent terminology and defects with incorrect step sequence, and determining the quality assessment score of the engineering test document.

[0069] Optionally, the quality assessment score can be determined using the following formula:

[0070] ;

[0071] Where n represents the number of document defect types (different terminology may result in different document defect types). The weight corresponding to the i-th document defect type. This represents the number of times the defect of the i-th document defect type occurs.

[0072] This application embodiment performs terminology consistency checks on engineering test documents based on a structured entity tree, and performs timing checks on the test steps in the engineering test documents. Based on the consistency check results and timing check results, the quality of the engineering test documents is evaluated, thereby improving the testing accuracy of the engineering test documents and thus improving the document quality.

[0073] Example 3

[0074] Figure 3 This is a schematic diagram of a quality assessment device for engineering documents according to Embodiment 3 of this application. It is applicable to the quality assessment of application test documents during application development and testing. This quality assessment device can be implemented in hardware and / or software and can be configured in a computer device, such as a server. Figure 3 As shown, the device includes:

[0075] The semantic feature extraction module 310 is used to acquire the engineering test document to be evaluated, perform text segmentation on the engineering test document, and extract semantic features from the segmented text units to generate semantic context features; wherein, the semantic context features are context semantic features with at least two feature dimensions.

[0076] The label space generation module 320 is used to extract dependent features from the semantic context features based on a bidirectional long short-term memory network, and to assign positional attention weights to the dependent feature extraction results through a gating mechanism to generate a domain entity label space.

[0077] The structure tree generation module 330 is used to determine the optimal tag sequence based on the legal tag transfer rules and the domain entity tag space, and to determine the structured entity tree corresponding to the engineering test document based on the optimal tag sequence.

[0078] The quality assessment module 340 is used to perform quality assessment on the engineering test documents based on the structured entity tree.

[0079] This application embodiment extracts semantic context features and dependency features from the engineering test document twice to determine the domain entity label space corresponding to the engineering test document, thereby improving the accuracy of entity boundary determination and thus improving the quality assessment accuracy of the engineering test document.

[0080] Optionally, the quality assessment module 340 includes:

[0081] A clustering unit is used to traverse the tree nodes in the structured entity tree, cluster the tree nodes according to the label attributes of the nodes, and generate at least one tree node group.

[0082] The consistency detection unit is used to perform term consistency detection based on semantic similarity calculation and character similarity calculation for the node values ​​of any two tree nodes in the tree node group;

[0083] The graph construction unit is used to traverse the tree nodes in the structured entity tree, extract at least one target tree node whose entity label attribute is a step label, and construct a target directed acyclic graph for at least one target tree node based on the node attributes of the target tree node.

[0084] The timing detection unit is used to traverse the target directed acyclic graph according to the preset test step execution rules and perform step timing detection; wherein, the test step timing detection is used to determine whether there are edge dependency errors and loop risks in the target directed acyclic graph;

[0085] The quality assessment unit is used to assess the quality of the engineering test documents based on the results of terminology consistency detection and step sequence detection.

[0086] Optionally, the quality assessment unit can be specifically used for:

[0087] Based on the number of defects occurring due to inconsistent terminology and errors in the timing of steps, different document defects are weighted to determine the quality assessment score of the engineering test document.

[0088] Optionally, the structure tree generation module 330 includes:

[0089] The transition matrix determination unit is used to determine the label transition probability matrix according to the legal label transition rules; wherein the label transition probability matrix is ​​used to characterize the connection probability between different entity labels;

[0090] The optimal sequence unit is used to perform weighted summation on the label feature vectors corresponding to the text units in the domain entity label space according to the label transition probability matrix, and take the entity label sequence with the highest path score as the optimal label sequence.

[0091] Optionally, in the domain entity label space, each text unit has its own corresponding label feature vector; the label feature vector is used to characterize the maximum label probability of each text unit corresponding to at least one entity label, and the maximum label probability is used to characterize the probability that the text unit belongs to one entity label.

[0092] Optionally, the structure tree generation module 330 also includes:

[0093] Extraction start unit, used to traverse the optimal label sequence; if the current traversal object is the first target label, then entity extraction start is triggered.

[0094] The extraction end unit is used to trigger the end of entity extraction if the currently traversed object is the second target tag during the entity extraction process.

[0095] The structure tree construction unit is used to merge the text units corresponding to the entity label sequences traversed during entity extraction, and generate a structured entity tree corresponding to the engineering test document based on the merged text object and the text attributes of the text units corresponding to the label sequences.

[0096] The engineering document quality assessment device provided in this application embodiment can execute the engineering document quality assessment method provided in any embodiment of this application, and has the corresponding functional modules and beneficial effects for executing the quality assessment method of each engineering document.

[0097] According to embodiments of this application, this application also provides an electronic device, a readable storage medium, and a computer program product.

[0098] Example 4

[0099] Figure 4 This is a schematic diagram of the structure of an electronic device 410 implementing the engineering document quality assessment method of the embodiments of this application. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or claimed herein.

[0100] like Figure 4 As shown, the electronic device 410 includes at least one processor 411 and a memory, such as a read-only memory 412 or a random access memory 413, communicatively connected to the at least one processor 411. The memory stores computer programs executable by the at least one processor. The processor 411 can perform various appropriate actions and processes based on the computer program stored in the read-only memory 412 or loaded from storage unit 418 into the random access memory 413. The random access memory 413 can also store various programs and data required for the operation of the electronic device 410. The processor 411, read-only memory 412, and random access memory 413 are interconnected via a bus 414. An input / output interface 415 is also connected to the bus 414.

[0101] Multiple components in electronic device 410 are connected to input / output interface 415, including: input unit 416, such as keyboard, mouse, etc.; output unit 417, such as various types of monitors, speakers, etc.; storage unit 418, such as disk, optical disk, etc.; and communication unit 419, such as network card, modem, wireless transceiver, etc. Communication unit 419 allows electronic device 410 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0102] Processor 411 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 411 include, but are not limited to, central processing units, graphics processing units, various special-purpose artificial intelligence computing chips, various processors running machine learning model algorithms, digital signal processors, and any suitable processor, controller, microcontroller, etc. Processor 411 performs the various methods and processes described above, such as quality assessment methods for engineering documents.

[0103] In some embodiments, the engineering document quality assessment method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 418. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 410 via read-only memory 412 and / or communication unit 419. When the computer program is loaded into random access memory 413 and executed by processor 411, one or more steps of the engineering document quality assessment method described above may be performed. Alternatively, in other embodiments, processor 411 may be configured as the engineering document quality assessment method by any other suitable means (e.g., by means of firmware).

[0104] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays, application-specific integrated circuits (ASICs), application-specific standard products (ASICs), systems-on-a-chip (SoCs), payload programmable logic devices, computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0105] Computer programs used to implement the methods of this application may be written in any combination of one or more programming languages. These computer programs may be provided to the processor of a general-purpose computer, a special-purpose computer, or other programmable engineering document quality assessment device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0106] In the context of this application, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, optical fibers, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0107] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a cathode ray tube or liquid crystal display monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0108] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0109] A computing system can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product within the cloud computing service system to address the shortcomings of traditional physical hosts and virtual private servers, such as high management difficulty and weak business scalability.

[0110] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this application can be achieved, and this is not limited herein.

[0111] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for quality assessment of engineering documents, characterized in that, include: The engineering test document to be evaluated is obtained, the engineering test document is processed by text segmentation, and semantic features are extracted from the segmented text units to generate semantic context features; wherein, the semantic context features are context semantic features with at least two feature dimensions. Based on a bidirectional long short-term memory network, dependency feature extraction is performed on the semantic context features, and positional attention weights are assigned to the dependency feature extraction results through a gating mechanism to generate a domain entity label space. Based on the legal tag transfer rules and the domain entity tag space, the optimal tag sequence is determined, and based on the optimal tag sequence, the structured entity tree corresponding to the engineering test document is determined. The quality of the engineering test documents is assessed based on the structured entity tree.

2. The method according to claim 1, characterized in that, The quality assessment of the engineering test documents based on the structured entity tree includes: The tree nodes in the structured entity tree are traversed, and the tree nodes are clustered according to the label attributes of the nodes to generate at least one tree node group. For any two tree nodes in the aforementioned tree node group, terminology consistency detection is performed based on semantic similarity calculation and character similarity calculation; The tree nodes in the structured entity tree are traversed to extract at least one target tree node whose entity label attribute is a step label, and a target directed acyclic graph is constructed for at least one target tree node based on the node attributes of the target tree node. According to the preset test step execution rules, the target directed acyclic graph is traversed and the step timing detection is performed; wherein, the test step timing detection is used to determine whether there are edge dependency errors and loop risks in the target directed acyclic graph; The quality of the engineering test documents is assessed based on the results of terminology consistency detection and step sequence detection.

3. The method according to claim 1, characterized in that, Based on the results of terminology consistency checks and step sequence checks, the quality of the engineering test documents is assessed, including: Based on the number of defects occurring due to inconsistent terminology and errors in the timing of steps, different document defects are weighted to determine the quality assessment score of the engineering test document.

4. The method according to claim 1, characterized in that, The process of determining the optimal label sequence based on the legal label transfer rules and the domain entity label space includes: Based on the legal label transfer rules, a label transfer probability matrix is ​​determined; wherein, the label transfer probability matrix is ​​used to characterize the connection probability between different entity labels; Based on the label transition probability matrix, the label feature vectors corresponding to the text units in the domain entity label space are weighted and summed respectively, and the entity label sequence with the highest path score is taken as the optimal label sequence.

5. The method according to claim 4, characterized in that, In the domain entity label space, each text unit has its own corresponding label feature vector; the label feature vector is used to characterize the maximum label probability of each text unit under at least one entity label, and the maximum label probability is used to characterize the probability that the text unit belongs to one entity label.

6. The method according to claim 1, characterized in that, The step of determining the structured entity tree corresponding to the engineering test document based on the optimal tag sequence includes: The optimal label sequence is traversed. If the current traversed object is the first target label, entity extraction is triggered. During entity extraction, if the currently traversed object is the second target label, the entity extraction will end. The text units corresponding to the entity label sequences traversed during entity extraction are merged, and a structured entity tree corresponding to the engineering test document is generated based on the merged text object and the text attributes of the text units corresponding to the label sequences.

7. A quality assessment device for engineering documents, characterized in that, include: The semantic feature extraction module is used to acquire the engineering test document to be evaluated, perform text segmentation on the engineering test document, and extract semantic features from the segmented text units to generate semantic context features; wherein, the semantic context features are context semantic features with at least two feature dimensions. The label space generation module is used to extract dependent features from the semantic context features based on a bidirectional long short-term memory network, and to assign positional attention weights to the dependent feature extraction results through a gating mechanism to generate a domain entity label space. The structure tree generation module is used to determine the optimal tag sequence based on the legal tag transfer rules and the domain entity tag space, and to determine the structured entity tree corresponding to the engineering test document based on the optimal tag sequence. The quality assessment module is used to assess the quality of the engineering test documents based on the structured entity tree.

8. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the quality assessment method for engineering documents as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the quality assessment method for engineering documents as described in any one of claims 1-6.

10. A computer program product comprising a computer program that, when executed by a processor, implements the method for quality assessment of engineering documents according to any one of claims 1-6.

Images