CN109375948A - A kind of software pricing method of intelligent function point identification - Google Patents
A kind of software pricing method of intelligent function point identification Download PDFInfo
- Publication number
- CN109375948A CN109375948A CN201811133405.7A CN201811133405A CN109375948A CN 109375948 A CN109375948 A CN 109375948A CN 201811133405 A CN201811133405 A CN 201811133405A CN 109375948 A CN109375948 A CN 109375948A
- Authority
- CN
- China
- Prior art keywords
- user function
- software
- function
- user
- file
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/70—Software maintenance or management
- G06F8/77—Software metrics
Abstract
The present invention relates to statistics, natural language processing technique and machine learning fields.Relate to a kind of software pricing method of intelligent function point identification, comprising: parse to multiple files in target software project, obtain corresponding parsing text;According to corresponding recognition rule, the user function in the parsing text is identified using natural language processing method;The user function recognized is assessed, the functional coefficient of the corresponding user function is obtained;It is valuated according to the quantity of the same kind user function recognized and corresponding functional coefficient to the type user function;According to the valuation to multiple type of user functions as a result, valuating to the target software project.The present invention has the advantages that the estimation valuation that user can obtain software without largely checking requirement documents by the method for the invention is reported;Using the method for machine learning, learn potential recognition rule, and formulate recognition rule scheme automatically, realizes the identification of intelligent automation.
Description
Technical field
The present invention relates to a kind of identifications of intelligent function point in statistics, natural language processing technique and machine learning field
Software pricing method.
Background technique
As the scale and complexity of computer software is continuously increased, software size is estimated and is measured, by
Be known as be software project standardized operation, successful execution an important link.Effectively software Quantity customizing is in software development
In be very difficult because factor involved in software development is not only more but also complex.If underestimating the rule of software
Mould, will cause that human resources are in short supply, and cost excesses budget, and influence software project quality.And over-evaluated software size, it will cause people
The power utilization of resources is insufficient, whole development inefficiency.So the exploitation scale for being computed correctly and assessing software is extremely important
's.
Traditional software pricing mode is estimated with code line analysis or expert judgments, and all there is not for these
Certainty and inaccuracy.
Summary of the invention
In view of the above-mentioned problems, a kind of software pricing method of intelligent function point identification proposed by the present invention can be by software system
System function is gradually segmented down to lesser component, so that software systems are easier to analyzed and estimate, to accurately identify
Function point in various software projects out.It is combined by natural language processing technique and machine learning, so that software valuation is more
Add simplification and intelligence, reduces the artificial time and efforts for calculating function point, provide detailed and accurate software for user and estimate
Calculate price.Specifically,
The purpose of the present invention is a kind of intelligent function point being achieved through the following technical solutions identification software pricing method,
Include: that multiple files in target software project are parsed, obtains corresponding parsing text;According to corresponding identification rule
Then, the user function in the parsing text is identified using natural language processing method;To the user function recognized
It is assessed, obtains the functional coefficient of the corresponding user function;According to the quantity of the same kind user function recognized with
And corresponding functional coefficient valuates to the type user function;According to the valuation to multiple type of user functions as a result, to described
The valuation of target software project.
Further, it is parsed in the multiple types file in target software project, obtains corresponding parsing
Before text further include: load target software project, and the multiple types file in target software project is managed.
Further, it includes: addition or removal that the multiple types file in target software project, which is managed,
File.
Further, it parses, obtains in corresponding parsing text, institute to multiple files in target software project
Parsing text is stated to show in the form of structure tree according to its titles at different levels, chapters and sections for parsing content.
Further, the structure tree includes: for choosing the content of text needed with user function identification;For
The user function recognized is checked.
Further, according to corresponding recognition rule, using natural language processing method in the parsing text
Before user function is identified further include: according to the target software project, establish corresponding dictionary dictionary;According to the word of foundation
Allusion quotation dictionary formulates recognition rule;It is trained using identification process of the sample to user function.
Further, the dictionary dictionary is used to know the matching of the recognition rule by natural language processing method
Not.
Further, the user function includes: internal logical file, external interface file, and external input, outside are defeated
Out and external inquiry.
Further, it includes: the vocabulary such as creation, foundation, formation or formulation that the recognition rule, which includes: before noun subject,
Matching, can recognize as internal logical file;It include: the matching of the vocabulary such as reference, utilization, basis or reference before noun subject,
It can recognize as external interface file;Predicate includes: the matching of the vocabulary such as reception, detection, acquisition, reading, can recognize to be external defeated
Enter;Predicate includes: the matching vocabulary such as calculate, export, printing, exporting, and be can recognize as outside output;Predicate include: inquiry,
The matching of the vocabulary such as display, sequence, screening, can recognize as external inquiry.
Further, the software pricing method of a kind of intelligent function point identification further include: use machine learning method
Learnt.
The present invention has the advantages that user is it is not necessary that largely software can be obtained by checking requirement documents by the method for the invention
Estimation valuate report;Pass through natural language processing technique, the grammers such as the grammer dependence and identification Subject, Predicate and Object of anolytic sentence
Structure so that by the training of a small amount of sample can be thus achieved in large scope software project to the identification work of user function;
And in the process, it is only necessary to which the function that Adds User can be identified by carrying out simple maintenance to dictionary dictionary;This
Outside, the method that the present invention utilizes machine learning, is analyzed and processed a large amount of recognition result data, learns potential identification rule
Then mode realizes the identification of intelligent automation to formulate recognition rule scheme automatically.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the present invention
Limitation.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings:
Fig. 1 shows the work flow diagram of pricing method of the present invention.
Fig. 2 shows the preparation flow charts of pricing method of the present invention.
Fig. 3 shows the training process schematic diagram of pricing method of the present invention.
Fig. 4 shows the structure composition figure of pricing system of the present invention.
Specific embodiment
The illustrative embodiments of the disclosure are more fully described below with reference to accompanying drawings.Although showing this public affairs in attached drawing
The illustrative embodiments opened, it being understood, however, that may be realized in various forms the disclosure without the reality that should be illustrated here
The mode of applying is limited.It is to be able to thoroughly understand the disclosure on the contrary, providing these embodiments, and can be by this public affairs
The range opened is fully disclosed to those skilled in the art.
Embodiment according to the present invention proposes a kind of software pricing method of intelligent function point identification, passes through building
Dictionary dictionary and formulation recognition rule, identify the user function in file using natural language recognition method;By giving
The functional coefficient of corresponding function type out, and then valuate to each type function, finally according to multiple types function in project
The valuation of energy obtains the valuation to the destination item.In addition, also introducing machine learning method in the present invention, learned for self
It practises.Particular content of the invention is as follows:
As shown in Figure 1, being the work flow diagram of pricing method of the present invention.Pricing method of the invention includes: that step 001 is right
Multiple files in target software project are parsed, and corresponding parsing text is obtained;Step 002 is according to corresponding identification rule
Then, the user function in the parsing text is identified using natural language processing method;Step 003 is to the use recognized
Family function is assessed, and the functional coefficient of the corresponding user function is obtained;Step 004 is according to the same kind user recognized
The quantity of function and corresponding functional coefficient valuate to the type user function;Step 005 is according to multiple type of user function
The valuation of energy is as a result, valuate to the target software project.
Specifically, being parsed in the step 001 to the multiple types file in target software project, obtain corresponding
Before parsing text, the invention also includes: load target software project, and the file in target software project is managed, e.g.,
Increase file or deletes file.In addition, the file after only being loaded can just carry out the parsing operation of next step.Pass through
Specific analytic method parses the file, and the content obtained in file parses document;The present invention is exactly based on pair
The description rule of specific user function is identified in parsing document, so that it is determined that user function.And for subsequent step,
The present invention is during obtaining parsing document, according to the titles at different levels of its parsing content, chapters and sections to solution in the form of structure tree
Analysis document is shown.By structure tree so that user can be convenient transfer its content for needing to check, in the process checked
In, user can verify the user function recognized.In addition, being adopted carrying out step 002 according to corresponding recognition rule
Before being identified with natural language processing method to the user function in the parsing text, it is also necessary to carry out following beam worker
Make:
As shown in Fig. 2, being the preparation flow chart of pricing method of the present invention.Wherein, preparation includes: step 011
According to the target software project, corresponding dictionary dictionary is established;Step 021 formulates identification rule according to the dictionary dictionary of foundation
Then;Step 031 is trained the identification process of user function using sample.More specifically, due to the target software project
It is different for field, the form of presentation of the difference of running equipment, user function is also not quite similar.Therefore, the present invention is by root
Corresponding dictionary dictionary is established according to different software projects, for passing through natural language processing method to the content in parsing text
Carry out match cognization.The identification of user function is advised during establishing dictionary dictionary, while according to the content of parsing text
Then formulated.Then the matching of recognition rule can be known by natural language recognition method using the dictionary dictionary
Not, and then specific user function is determined.In the identification process of user function, the dictionary dictionary is realized to word or word
Identification, and the recognition rule realizes the composition of subject-predicate object in the identification to sentence structure, such as a word.Formal
Before the identification for carrying out user function, the present invention also needs to be trained identification process using a certain amount of sample.The step 031
Details are provided below:
As shown in figure 3, being the training process schematic diagram of pricing method of the present invention.Wherein, training process includes: step 131
By a certain amount of sample, the identification process of user function is trained;Step 231 is according to structure tree to the user recognized
Function is tested.More specifically, the purpose of the training is mainly the training to dictionary dictionary and recognition rule, according to before
The part recognition rule and dictionary dictionary of formulation identify user function;Then logical by the structure tree according to parsing text
It crosses and the user function recognized is carried out to check verifying, and then the content in dictionary dictionary is modified or increased, i.e., to word
Allusion quotation dictionary is safeguarded;It is including in training process according in identification process of the recognition rule to user function, the present invention is also
Machine learning method is introduced, for voluntarily constructing recognition rule to the recognition result for reaching certain identification scale.
After training, that is, carry out normal user function identification process.In normal user function identification process, use
Family can also verify the user function recognized, and carry out machine learning.Wherein, user function is generally divided into five classes
It include: internal logical file, external interface file, external input, external output and external inquiry.Knowledge corresponding to user function
The not matching of rule to include: before noun subject include: vocabulary such as creation, foundation, formation or formulation, can recognize as internal logic text
Part;Include: the matching of the vocabulary such as reference, utilization, basis or reference before noun subject, can recognize as external interface file;Predicate
Include: the matching of the vocabulary such as reception, detection, acquisition, reading, can recognize and be an externally input;Predicate includes: to calculate, export, beating
The matching of the vocabulary such as print, export can recognize as outside output;Predicate includes: of the vocabulary such as inquiry, display, sequence, screening
Match, can recognize as external inquiry.Certainly, user function is not only limited in above-mentioned five seed type, additionally certainly including some users
Surely the function of setting, such as dynamic guest's relation recognition rule in this example;Subject-predicate relation recognition rule and specific area noun are known
Not rule etc..The dictionary dictionary established and safeguarded in the present invention includes verb library, thesaurus etc..Dictionary dictionary can be with additions and deletions
Change and looks into.In addition, the natural language processing technique is mainly used for, to participle, part-of-speech tagging and the interdependent pass of syntax of parsing text
System etc..
Next, evaluating the user function recognized.Same type of user function is evaluated, according to function
The field of energy type and the target software project application obtains corresponding functional coefficient, and then calculates the function of the type function
Points, wherein function points=same type user function number × functional coefficient.
Finally, carrying out comprehensive function to the software project according to the function of all types in target software project points
Points statistics, and according to the price for setting the corresponding function point, it valuates to target software project.Further, it is also possible to
Comprehensive other impersonal force expenses in addition to software project itself, such as evaluation and test takes, dedicated expense, to calculate the comprehensive of entire project
Conjunction expense.A system of application the method for the present invention will be illustrated below:
As shown in figure 4, being the structure composition figure of pricing system of the present invention.Wherein, invention software pricing system includes: text
Shelves load parsing module obtains corresponding parsing for loading software project file, parsing to the file in software project
Text;Recognition rule configuration module for configuring corresponding recognition rule according to software project, and uses natural language processing side
Method identifies the user function in the parsing text;Function Quantity customizing module, for the user being finally identified to
Function is assessed, and the functional coefficient of the corresponding user function is obtained, to calculate the function points of the type user function;
Software valuation module, for being counted according to the function, in conjunction with Dynamic gene and user function time-consuming to entire software item visual estimation
Price is calculated, and provides assessment report.Wherein, synthesis described in the method for the present invention is contained in the Dynamic gene except software item
Other impersonal force expenses other than mesh itself, the expense adjustment such as situations such as evaluation and test takes, dedicated expense.The user function time-consuming is used
In the given of the functional coefficient.
Specific embodiment
The software pricing method that the present embodiment is identified with a kind of application intelligent function point of the invention is to GNC controller application
For the valuation of software.
Navigation and control (GNC) technology mainly study Aeronautics and Astronautics, navigation, the position of all kinds of movable bodies of land row, direction,
Track, the measurement of posture, control and decision problem, be Defense Weapon System and civilian transportation system important core technology it
One.The research and development operation of GNC controller application software is very crucial for space technology and its application.The exploitation scale of the software
A quite high level is reached with complexity, the work estimated software size and measured should not be underestimated, traditional
Software pricing mode is very complicated and inaccurate, using intelligent function point of the present invention identify software pricing system according to
The exploitation demand of GNC controller becomes very convenient reliable to carry out estimation valuation to the software.It below will be simply to using this
Inventive method carries out the part treatment process in GNC project
Firstly, importing the requirement profile document of GNC development project, load files under corresponding project directory,
Classification display file name imports time and path, and chooses GNC controller demand file and carry out text resolution;Extract solution
Chapters and sections content in analysis text is showed in the form of tree structure.
Then, by the matching of existing dictionary dictionary and recognition rule, different function point in demand chapters and sections is known in realization
Not.Wherein, optional recognition rule e.g. moves guest's relation recognition, and verb is " generating and sending ", and noun is " inertial navigation parameter ", right
This can then be respectively identified as an internal logical file (ILF) and an external output (EO).In addition, being to custom rule
When certain specific area noun, then internal logical file (ILF) is also identified as;For another example " calculate rate damping posture twice ",
Wherein " rate damping posture " is identified as an internal logical file (ILF), and " calculating twice " is identified as two external outputs
(EO).It is right by the sentence paragraph of the corresponding original text of the structure tree click recognition result according to the result data of statistics identification
The some and expected result not being inconsistent is corrected, and saves result after verification.At the same time, machine learning is utilized to result data
Method carries out assisting in identifying operation, and the phrase of automatic study to certain the class phrase for belonging to specific function type combines;In machine
During study, adjustment by user to recognition result constantly enhances the correctness to lay down a regulation by machine learning.
Subsequently, the field according to applied by specific user function type and software project, provides functional coefficient respectively,
Such as: 1 ILF=10 function point, 1 EIF=7 function point, 1 EI=4 function point, 1 EO=5 function point, 1
EQ=4 function point, and according to function points=same type user function number × functional coefficient, calculate each user function class
The function point sum of type, completes the valuation of itself to target software project.
Finally, according to the comprehensive cost of the scale Dynamic gene of setting and function point time-consuming rate software for calculation project, in addition
Impersonal forces cost and the software profits such as assessment takes, dedicated expense, travel charge, meeting expense, expert consulting take, it is total to calculate final software
Valence, and generate estimation valuation report.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
In the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of by anyone skilled in the art,
It should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with the protection model of the claim
Subject to enclosing.
Claims (10)
1. a kind of software pricing method of intelligent function point identification characterized by comprising
Multiple files in target software project are parsed, corresponding parsing text is obtained;
According to corresponding recognition rule, the user function in the parsing text is known using natural language processing method
Not;
The user function recognized is assessed, the functional coefficient of the corresponding user function is obtained;
It is valuated according to the quantity of the same kind user function recognized and corresponding functional coefficient to the type user function;
According to the valuation to multiple type of user functions as a result, valuating to the target software project.
2. software pricing method according to claim 1, which is characterized in that described to a variety of in target software project
Type file is parsed, before obtaining corresponding parsing text further include:
Target software project is loaded, and the multiple types file in target software project is managed.
3. software pricing method according to claim 2, which is characterized in that the multiple types in target software project
Type file, which is managed, includes:
Addition removes file.
4. software pricing method according to claim 1, which is characterized in that multiple files in target software project
It is parsed, is obtained in corresponding parsing text, titles at different levels that the parsing text parses content according to it, chapters and sections are with structure
The form of tree is shown.
5. software pricing method according to claim 4, which is characterized in that the structure tree includes:
For choosing the content of text needed with user function identification;
For being checked to the user function recognized.
6. software pricing method according to claim 1, which is characterized in that according to corresponding recognition rule, using from
Before right language processing method identifies the user function in the parsing text further include:
According to the target software project, corresponding dictionary dictionary is established;
According to the dictionary dictionary of foundation, recognition rule is formulated;
It is trained using identification process of the sample to user function.
7. software pricing method according to claim 6, which is characterized in that the dictionary dictionary is for passing through natural language
Match cognization of the processing method to the recognition rule.
8. according to claim 1, software pricing method described in 5 or 6 any one, which is characterized in that the user function packet
It includes:
Internal logical file, external interface file, external input, external output and external inquiry.
9. according to claim 1, software pricing method described in 6 or 7 any one, which is characterized in that the recognition rule packet
It includes:
Include: the matching of the vocabulary such as creation, foundation, formation or formulation before noun subject, can recognize as internal logical file;
Include: the matching of the vocabulary such as reference, utilization, basis or reference before noun subject, can recognize as external interface file;
Predicate includes: the matching of the vocabulary such as reception, detection, acquisition, reading, can recognize and is an externally input;
Predicate includes: the matching vocabulary such as calculate, export, printing, exporting, and be can recognize as outside output;
Predicate includes: the matching of the vocabulary such as inquiry, display, sequence, screening, be can recognize as external inquiry.
10. software pricing method according to claim 1, which is characterized in that further include:
Learnt using machine learning method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811133405.7A CN109375948A (en) | 2018-09-27 | 2018-09-27 | A kind of software pricing method of intelligent function point identification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811133405.7A CN109375948A (en) | 2018-09-27 | 2018-09-27 | A kind of software pricing method of intelligent function point identification |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109375948A true CN109375948A (en) | 2019-02-22 |
Family
ID=65402711
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811133405.7A Pending CN109375948A (en) | 2018-09-27 | 2018-09-27 | A kind of software pricing method of intelligent function point identification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109375948A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111274817A (en) * | 2020-01-16 | 2020-06-12 | 北京航空航天大学 | Intelligent software cost measurement method based on natural language processing technology |
CN112734468A (en) * | 2020-12-31 | 2021-04-30 | 内蒙古安盾信息安全评测有限公司 | Service processing method and device |
CN113011689A (en) * | 2019-12-19 | 2021-06-22 | 中国移动通信集团辽宁有限公司 | Software development workload assessment method and device and computing equipment |
CN113705239A (en) * | 2021-07-06 | 2021-11-26 | 北京明德微科科技有限公司 | Electronic generation method, device, system and medium for software cost |
CN116523544A (en) * | 2023-06-25 | 2023-08-01 | 江西省机电设备招标有限公司 | Software price measuring and calculating method, system, storage medium and equipment |
CN117493568A (en) * | 2023-11-09 | 2024-02-02 | 中安启成科技有限公司 | End-to-end software function point extraction and identification method |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001117763A (en) * | 1999-10-19 | 2001-04-27 | Nippon Steel Corp | Arithmetic unit for software scale, method for operating software scale and computer readable recording medium |
CN101344889A (en) * | 2008-07-31 | 2009-01-14 | 中国农业大学 | Method and system for network information extraction |
CN103577195A (en) * | 2013-11-14 | 2014-02-12 | 中国联合网络通信集团有限公司 | Software requirement analysis quantifying method and system |
CN106776538A (en) * | 2016-11-23 | 2017-05-31 | 国网福建省电力有限公司 | The information extracting method of enterprise's noncanonical format document |
-
2018
- 2018-09-27 CN CN201811133405.7A patent/CN109375948A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001117763A (en) * | 1999-10-19 | 2001-04-27 | Nippon Steel Corp | Arithmetic unit for software scale, method for operating software scale and computer readable recording medium |
CN101344889A (en) * | 2008-07-31 | 2009-01-14 | 中国农业大学 | Method and system for network information extraction |
CN103577195A (en) * | 2013-11-14 | 2014-02-12 | 中国联合网络通信集团有限公司 | Software requirement analysis quantifying method and system |
CN106776538A (en) * | 2016-11-23 | 2017-05-31 | 国网福建省电力有限公司 | The information extracting method of enterprise's noncanonical format document |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113011689A (en) * | 2019-12-19 | 2021-06-22 | 中国移动通信集团辽宁有限公司 | Software development workload assessment method and device and computing equipment |
CN111274817A (en) * | 2020-01-16 | 2020-06-12 | 北京航空航天大学 | Intelligent software cost measurement method based on natural language processing technology |
CN112734468A (en) * | 2020-12-31 | 2021-04-30 | 内蒙古安盾信息安全评测有限公司 | Service processing method and device |
CN113705239A (en) * | 2021-07-06 | 2021-11-26 | 北京明德微科科技有限公司 | Electronic generation method, device, system and medium for software cost |
CN116523544A (en) * | 2023-06-25 | 2023-08-01 | 江西省机电设备招标有限公司 | Software price measuring and calculating method, system, storage medium and equipment |
CN116523544B (en) * | 2023-06-25 | 2023-11-14 | 江西省机电设备招标有限公司 | Software price measuring and calculating method, system, storage medium and equipment |
CN117493568A (en) * | 2023-11-09 | 2024-02-02 | 中安启成科技有限公司 | End-to-end software function point extraction and identification method |
CN117493568B (en) * | 2023-11-09 | 2024-04-19 | 中安启成科技有限公司 | End-to-end software function point extraction and identification method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109375948A (en) | A kind of software pricing method of intelligent function point identification | |
CN109271201A (en) | A kind of software pricing system of intelligent function point identification | |
US7685082B1 (en) | System and method for identifying, prioritizing and encapsulating errors in accounting data | |
Parra et al. | A methodology for the classification of quality of requirements using machine learning techniques | |
CN103207855A (en) | Fine-grained sentiment analysis system and method specific to product comment information | |
CN110517130A (en) | A kind of intelligence bookkeeping methods and its system | |
CN107844558A (en) | The determination method and relevant apparatus of a kind of classification information | |
CN115547466B (en) | Medical institution registration and review system and method based on big data | |
CN115186780B (en) | Discipline knowledge point classification model training method, system, storage medium and equipment | |
CN113268561B (en) | Problem generation method based on multi-task joint training | |
CN114722820A (en) | Chinese entity relation extraction method based on gating mechanism and graph attention network | |
CN110472040A (en) | Extracting method and device, storage medium, the computer equipment of evaluation information | |
CN107577738A (en) | A kind of FMECA method by SVM text mining processing datas | |
CN112579777B (en) | Semi-supervised classification method for unlabeled text | |
CN107798137B (en) | A kind of multi-source heterogeneous data fusion architecture system based on additive models | |
Osman et al. | Generate use case from the requirements written in a natural language using machine learning | |
CN103970654B (en) | Software reliability virtual test method | |
CN103019924A (en) | Input method intelligence evaluation system and input method intelligence evaluation method | |
Hussain et al. | Towards approximating COSMIC functional size from user requirements in agile development processes using text mining | |
CN110502745B (en) | Text information evaluation method and device, computer equipment and storage medium | |
Zhou et al. | Automatic acquisition of linguistic patterns for conceptual modeling | |
CN110414819B (en) | Work order scoring method | |
Cholissodin et al. | Audit system development for government institution documents using stream deep learning to support smart governance | |
Li et al. | Self‐admitted technical debt detection by learning its comprehensive semantics via graph neural networks | |
Zhu et al. | Automated qualitative rule extraction based on bidirectional long shortterm memory model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190222 |