CN115033280A

CN115033280A - Knowledge graph-based automatic generation method for requirement specification document and storage medium

Info

Publication number: CN115033280A
Application number: CN202210760828.1A
Authority: CN
Inventors: 杨双远
Original assignee: Individual
Current assignee: Individual
Priority date: 2022-06-29
Filing date: 2022-06-29
Publication date: 2022-09-09

Abstract

A demand specification document automatic generation method storage medium based on a knowledge graph, a demand specification document automatic generation method based on a knowledge graph and a storage medium comprise the following steps: 101, generating a named entity identification and relation extraction result, 102, describing semantic features by software requirements, disambiguating by using an entity disambiguation model, 103, embedding the result in the entity-relation set obtained in the 102 on the basis of a requirement specification map of a tree model, 106, intelligently converting a UML model map, automatically generating the UML model map, and 107, generating a standard requirement specification document. By the scheme, the problems of missing and errors of requirements for manually compiling the requirement information can be solved, the relevant software requirement documents are compiled, and the automation level of software requirement document compiling is improved.

Description

Knowledge graph-based automatic generation method for requirement specification document and storage medium

Technical Field

The present invention relates to the field of document processing, and in particular, to a method and a storage medium for performing document processing by an automatic telephone.

Background

In the prior art, software requirement document analysis is the foundation and source of the design and implementation phase of a software system. However, in the current software project requirement analysis process, a hand weaving method is mostly adopted, which is tedious, and has the problems of service understanding deviation, process error or function omission and the like, which causes fatal errors in the design of subsequent system function modules, and greatly increases project research and development cost. Service-oriented software requirement documents are difficult to understand, and customers cannot really and accurately grasp the requirements. Whether the structural information of the functional requirements in the requirement specification can be accurately extracted and modeled will determine the quality of the requirement analysis.

At present, the generation and expression method of the functional requirement specification mainly comprises UML manual compilation, extraction rules, information extraction based on natural language processing and named entity identification extraction models.

In the prior art, the patent of CN112257404A discloses an automatic generation method and device of software requirement specification document based on QT 5. The accuracy of manual extraction, although high, is labor and time intensive and is subject to compiler subjectivity. The requirements for automatically generating and analyzing the requirement specification document cannot be met.

Disclosure of Invention

Therefore, a technical scheme that a method for automatically generating a document meeting software requirements and improving automation efficiency can be provided is needed.

In order to achieve the above object, the inventor provides a method for automatically generating a requirement specification document based on a knowledge graph, comprising the following steps:

step 101, generating a named entity identification and relationship extraction result,

102, disambiguating the recognition result by using the semantic features described by the software functional requirements and the entity disambiguation model to finally obtain a functional entity-relationship set,

103, embedding the result in the functional entity-relationship set obtained in the step 102 based on the requirement specification map of the tree model,

step 106, carrying out UML model diagram intelligent conversion, automatically generating a UML model diagram,

step 107, generating a standard requirement specification document.

In some embodiments of the present application, the method further includes step 104 of performing verification on the data stream during the execution of the software engineering.

In some embodiments of the present application, the method further includes a step 105 of performing a software requirement correctness check by using energy zeros and relations in the knowledge-graph link.

In some embodiments of the present application, a pretreatment step is further included.

In some embodiments of the present application, the step 102, where the software requirement describes semantic features, and disambiguating with the entity disambiguation model specifically includes the steps of:

step 201, processing based on the word meaning disambiguation model of the part of speech,

step 202, entity generation based on regular matching,

step 203, aligning the entity acronyms and reference words based on the context information,

step 204, learning the hidden relation based on punctuation marks,

step 205, learning of hidden relations based on numbers.

A knowledge-graph-based demand specification document automatically generates a storage medium,

stored with a computer program for performing, when executed, the steps comprising:

102, disambiguating the recognition result by using the semantic features described by the software functional requirements and using an entity disambiguation model to finally obtain a functional entity-relationship set,

step 107, generating a standard requirement specification document.

In some embodiments of the present application, the computer program is further configured to perform step 104, performing a check of the data stream during the execution of the software engineering.

In some embodiments of the application, the computer program when executed is further configured to perform a software requirement correctness check comprising step 105 using energy zeros and relations in the knowledge-graph links.

In some embodiments of the application, the computer program is further adapted to perform steps including preprocessing when executed.

In some embodiments of the present application, the step 102, where the software requirement describes semantic features, and the specific implementation of disambiguation by using the entity disambiguation model includes the steps of:

step 202, entity generation based on regular matching,

step 204, learning the hidden relation based on punctuation marks,

step 205, learning of hidden relations based on numbers.

Through the scheme, the problems of missing and errors of requirements for manually compiling the requirement information can be solved, the relevant software requirement documents are compiled, and the automation level of software requirement document compiling is improved.

Drawings

FIG. 1 is a flow diagram of a method for automated production of a knowledge-graph based requirements specification document, in accordance with an embodiment;

FIG. 2 is a schematic diagram illustrating the automated steps for generating a knowledge-graph based requirements specification document according to an embodiment;

FIG. 3 is a schematic diagram illustrating a data verification principle according to an embodiment;

FIG. 4 is a schematic diagram of the BilSTM-CRF training process according to an embodiment;

FIG. 5 is a schematic flow chart of an entity disambiguation model according to an embodiment;

FIG. 6 is a schematic diagram of a storage medium according to an embodiment.

Detailed Description

To explain technical contents, structural features, and objects and effects of the technical solutions in detail, the following detailed description is given with reference to the accompanying drawings in conjunction with the embodiments.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or related to other embodiments specifically defined. In principle, in the present application, the technical features mentioned in the embodiments can be combined in any manner to form a corresponding implementable solution as long as there is no technical contradiction or conflict.

Unless defined otherwise, technical terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the use of relational terms herein is intended to describe specific embodiments only and is not intended to limit the present application.

In the description of the present application, the term "and/or" is a expression for describing a logical relationship between objects, meaning that three relationships may exist, for example a and/or B, meaning: there are three cases of A, B, and both A and B. In addition, the character "/" herein generally indicates that the former and latter associated objects are in a logical relationship of "or".

In this application, terms such as "first" and "second" are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.

Without further limitation, in this application, the use of "including," "comprising," "having," or other similar expressions in phrases and expressions of "including," "comprising," or "having," is intended to cover a non-exclusive inclusion, and such expressions do not exclude the presence of additional elements in a process, method, or article that includes the recited elements, such that a process, method, or article that includes a list of elements may include not only those elements but also other elements not expressly listed or inherent to such process, method, or article.

As is understood in the examination of the guidelines, the terms "greater than", "less than", "more than" and the like in this application are to be understood as excluding the number; the expressions "above", "below", "within" and the like are understood to include the present numbers. In addition, in the description of the embodiments of the present application, "a plurality" means two or more (including two), and expressions related to "a plurality" similar thereto are also understood, for example, "a plurality of groups", "a plurality of times", and the like, unless specifically defined otherwise.

In the description of the embodiments of the present application, spatially relative expressions such as "central," "longitudinal," "lateral," "length," "width," "thickness," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "vertical," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," "circumferential," and the like are used, and the indicated orientations or positional relationships are based on the orientations or positional relationships shown in the specific embodiments or drawings and are only for convenience of describing the specific embodiments of the present application or for the convenience of the reader, and do not indicate or imply that the device or component in question must have a specific position, a specific orientation, or be constructed or operated in a specific orientation and therefore should not be construed as limiting the embodiments of the present application.

Unless specifically stated or limited otherwise, the terms "mounted," "connected," "secured," and "disposed" used in the description of the embodiments of the present application should be construed broadly. For example, the connection can be a fixed connection, a detachable connection, or an integrated arrangement; it can be a mechanical connection, an electrical connection, or a communication connection; they may be directly connected or indirectly connected through an intermediate; which may be communication within two elements or an interaction of two elements. Specific meanings of the above terms in the embodiments of the present application can be understood by those skilled in the art to which the present application pertains in accordance with specific situations.

In some embodiments, please refer to fig. 1, which is a method for automatically generating a requirement specification document based on a knowledge graph, comprising the following steps:

106, performing intelligent conversion on the UML model diagram, automatically generating the UML model diagram,

step 107, generating a standard requirement specification document.

In the above scheme, the sentence, word and BMEWO labeling are firstly carried out on the original service requirement. And then, identifying the named entity by using a BilSTM-CRF method, and extracting the named entity identification and the relation to obtain the marked service requirement. And then carrying out disambiguation processing on the related requirement documents according to the extracted semantic features described by the software requirements. The software functional requirement description refers to extracting functional requirements facing users and automatically generating requirement specifications of software. And finally, embedding according to a required specification map, and automatically generating a UML (unified modeling language) diagram: aiming at the problem that the structured information of the prior invention is not visually and visually represented, the method and the device automatically convert the acquired requirement specification map into the UML map, so that the requirement specification is more standard. Meanwhile, understanding errors of service requirements caused by the fact that the requirement specifications are not standard enough are avoided. And finally, generating a standard requirement document, so that the technical problem of automatic generation of the requirement document can be met, and the technical effect of improving the generation efficiency of the requirement document is achieved.

In some embodiments of the present application, please further refer to fig. 2, which further includes step 104 of performing a verification of the data stream during the execution of the software engineering. By checking the data stream, the fault tolerance of the system scheme can be effectively improved, and in some embodiments as shown in fig. 3, the data stream check may be performed through the UC matrix. Improvement of classical Use/Create (abbreviated as U/C) matrix: the classical Use/Create (U/C for short) matrix can only record in a single mode, and cannot realize the comparison and inspection of the local demand investigation result of multiple persons and multiple departments. Aiming at the problem, the project plans to disassemble the U/C matrix into a plurality of Send/Use/Create (S/U/C for short) matrixes. As shown in fig. 3, the main operation of splitting the U/C matrix into a plurality of S/U/C matrices is: based on each functional entity (such as functions F1, F2, F3 and F4 in fig. 3), only the corresponding whole row information and data column information with "C" are retained, and then all the "U" in the column with "C" are modified to "S". Through the improved UC matrix process, the traditional U/C matrix is improved: aiming at the problem that the traditional U/C matrix can only record in a single mode and cannot realize the comparison and inspection of the investigation result of the local requirements of multiple persons and multiple departments, the invention invents the compound U/C matrix, uses the S/U/C matrix to record the relation between functions and data tables, and is convenient for the discriminator to carry out automatic verification according to the energy transfer relation between S and U.

In some embodiments of the present application, please refer to fig. 2 further, which further includes a step 105 of performing a software requirement correctness check by using energy zeros and relations in the knowledge-graph link. By designing knowledge graph links, the correctness of the required document can be further verified, and the practicability of the design method is improved.

In some embodiments of the present application, a pre-treatment step is also included. The preprocessing step can comprise the steps of character recognition, segmentation, distribution, labeling and the like. Some embodiments of the method for identifying a named entity include: the invention firstly carries out sentence segmentation, word segmentation and BMEWO labeling on the original service requirement. The named entities are then identified using the BilSTM-CRF method, and FIG. 4 discloses the training process for the model. Through designing the pretreatment step, the material can be better suitable for the subsequent steps, and the processing efficiency of the subsequent scheme of the scheme is improved.

In order to better perform disambiguation of semantic features, in some embodiments of the present application, as shown in fig. 5, the step 102 of describing semantic features by software requirements specifically includes the steps of:

step 201, processing word sense disambiguation model based on parts of speech,

step 202, entity generation based on regular matching,

step 204, learning the hidden relation based on punctuation marks,

step 205, learning of hidden relations based on numbers.

The problem of insufficient performance of the model due to insufficient context semantic analysis in the existing demand specification automatic extraction model is solved, and meanwhile, the function recognition is fine-grained. The problem of requirement omission and human error which may occur in the process of manually compiling the requirement specification is solved, the workload of software requirement specification compilation is reduced, and meanwhile, the missing of parent-child relationship among functions is avoided by utilizing the hidden relationship learning. Through the scheme, the software requirement description can be subjected to more accurate word sense disambiguation. Thereby enabling the generation of more accurate demand documents.

The technical scheme aims at the problems that a requirement specification document needs to be manually compiled, version updating iteration is difficult, interpretability is poor, quality cannot be quickly checked and quantified and the like in the traditional software requirement modeling process. Aiming at the problem that the strategy for formulating the extraction rule is limited, the patent defines a demand specification generator model which is mainly responsible for automatically extracting key words from an original software business demand corpus, generating a demand specification map and automatically converting the demand specification map into a demand specification document in the later period. The method comprises the contents of four parts, namely, the improvement of a classic Use/Create matrix, a named entity identification and relationship extraction network, a word sense and entity disambiguation network based on context information, and the embedding of a demand specification map based on a tree structure. Aiming at the problems that the existing information extraction and named entity identification extraction model based on natural language processing is insufficient in context analysis and insufficient in semantic analysis and extraction of hidden structural relations, the word meaning and entity disambiguation network based on context information in the patent requirement specification generator model performs word meaning and entity disambiguation by using context semantics and performs extraction of hidden structural relations at the same time, and errors caused by coarse-grained identification are avoided.

In certain embodiments, as shown in FIG. 6, further comprising a knowledge-graph based requirements specification document automated production storage medium 600 storing a computer program which when executed is operable to perform steps comprising:

step 102, describing semantic features by software requirements, carrying out disambiguation by utilizing an entity disambiguation model,

103, embedding the result in the entity-relationship set obtained in the step 102 based on the requirement specification map of the tree model,

step 107, generating a standard requirement specification document.

The storage medium of the scheme firstly carries out sentence segmentation, word segmentation and BMEWO labeling on the original service requirement. And then, identifying the named entity by using a BilSTM-CRF method, and extracting the named entity identification and the relation to obtain the marked service requirement. And then carrying out disambiguation processing on the related requirement documents according to the extracted semantic features described by the software requirements. And finally, embedding according to a required specification map, and automatically generating a UML (unified modeling language) diagram: aiming at the problem that the structured information of the prior invention is not visually represented, the method and the device automatically convert the acquired requirement specification map into the UML map, so that the requirement specification is more standard. Meanwhile, understanding errors of service requirements caused by the fact that the requirement specifications are not standard enough are avoided. And finally, generating a standard requirement document, so that the technical problem of automatic generation of the requirement document can be met, and the technical effect of improving the generation efficiency of the requirement document is achieved.

In some embodiments of the present application, the computer program when executed is further configured to perform a software requirement correctness check comprising step 105 using energy zeros and relations in the knowledge-graph links.

step 202, entity generation based on regular matching,

step 204, learning the hidden relation based on punctuation marks,

step 205, learning of hidden relations based on numbers.

It should be noted that, although the above embodiments have been described herein, the invention is not limited thereto. Therefore, based on the innovative concepts of the present invention, the technical solutions of the present invention can be directly or indirectly applied to other related technical fields by making changes and modifications to the embodiments described herein, or by using equivalent structures or equivalent processes performed in the content of the present specification and the attached drawings, which are included in the scope of the present invention.

Claims

1. An automatic generation method of a requirement specification document based on a knowledge graph is characterized in that,

the method comprises the following steps:

step 107, generating a standard requirement specification document.

2. The method for automatically generating a knowledge-graph-based requirements specification document according to claim 1, further comprising step 104 of performing a check on data flow during execution of software engineering.

3. The method of claim 1, further comprising a step 105 of performing a software requirement correctness check using energy zeros and relations in the knowledge-graph links.

4. The method for automated knowledge-graph-based requirement specification document generation of claim 1, further comprising a preprocessing step.

5. The method for automatically generating a knowledge-graph-based requirement specification document according to claim 1, wherein the step 102 of describing semantic features by software requirements, and the disambiguation by using an entity disambiguation model specifically comprises the steps of:

step 202, entity generation based on regular matching,

step 204, learning the hidden relation based on punctuation marks,

step 205, learning of hidden relations based on numbers.

6. A knowledge-graph based demand specification document automated production storage medium, having stored thereon a computer program for executing steps comprising:

step 107, generating a standard requirement specification document.

7. The knowledgegraph-based requirements specification document of claim 6, wherein the computer program when executed is further configured to perform step 104, performing a check of data flow during execution of a software project.

8. The automated knowledgegraph-based requirements specification document according to claim 6, wherein the computer program when executed is further configured to perform a software requirements correctness check including step 105 using energy zeros and relations in the knowledgegraph links.

9. The knowledgegraph-based requirements specification document of claim 6, wherein the computer program when executed is further configured to perform steps including preprocessing.

10. The automated knowledgegraph-based requirements specification document generation storage medium of claim 6, wherein the step 102, the software requirements describe semantic features, and the disambiguation using the entity disambiguation model specifically performs the steps of:

step 202, entity generation based on regular matching,

step 204, learning the hidden relation based on punctuation marks,

step 205, learning of hidden relations based on numbers.