A customized software function point identification and diagnosis system based on big data technology

Abstract

The present invention discloses a customized software function point identification and diagnosis system based on big data technology, which relates to the technical field of software function point identification, including a data acquisition module, a preprocessing module, a function point identification module, a diagnosis module, a function point dynamic optimization module and a cost management module. The customized software function point identification and diagnosis system based on big data technology provided by the present invention adopts fuzzy matching function and consistency discrimination function to evaluate the matching degree of function points between different documents and implementations, ensures the alignment of function point description and actual implementation, improves the consistency checking capability of the system, and reduces the deviation between function point definition and implementation. Cost prediction is performed when performing function point maintenance to avoid the situation where the cost is higher than the value. The present invention achieves better results in terms of accuracy and cost.

Description

Technical Field

[0001] The present invention relates to the technical field of software function point identification, and in particular to a customized software function point identification and diagnosis system based on big data technology. Background Art

[0002] At present, big data technology has been widely used, but in the field of customized software function point identification and diagnosis and cost, the maturity of related technologies may not be high enough. This may lead to problems in project establishment, budget preparation, and technical document writing, such as the inconsistency of software function points, the inconsistency between the budget content and the technical description of the function points, and the inability of technical, project management and supervision personnel to effectively identify and diagnose function points and effectively judge costs. Big data technology is updated and iterated quickly, which requires the system to keep up with the trend of technological development and introduce new technical methods and service directions in a timely manner to meet actual work needs. At the same time, in actual applications, due to the diversity, complexity and hugeness of software function points, as well as the differences and dynamics of function point adjustment and calculation, it may increase the difficulty of optimizing the number and cost of function points, thereby affecting the construction and management effect of the system and causing investment to be out of control.

[0003] Therefore, it is necessary to provide a new technology, optimization algorithm, improve the function point identification and diagnosis of customized software, and strengthen the calculation, inspection and user-friendly customized software function point identification and diagnosis system of the function point number of the customized development system to overcome the above-mentioned problems. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: the existing software function point identification and detection methods have insufficient recognition accuracy, low diagnostic efficiency, and inaccurate cost control, as well as how to improve the accuracy and efficiency of function point identification, diagnosis and budget management through optimization algorithms.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a customized software function point identification and diagnosis system based on big data technology, including a data acquisition module, a preprocessing module, a function point identification module, a diagnosis module, a function point dynamic optimization module and a cost management module.

[0007] The data acquisition module is used to acquire multi-source data in real time.

[0008] The preprocessing module is used to process multi-source data and provide standardized input for the function point identification module, the diagnosis module, and the cost management module.

[0009] The function point identification module is used to identify and vectorize function points in a document.

[0010] The diagnosis module is used to detect the consistency of the function point description.

[0011] The function point dynamic optimization module is used to optimize the number of function points and configure resources.

[0012] The cost management module is used to dynamically predict project costs.

[0013] As a preferred solution of the customized software function point identification and diagnosis system based on big data technology described in the present invention, the data acquisition module is connected to the preprocessing module to transmit the collected data to the preprocessing module.

[0014] The data collection module collects structured data, unstructured text data, time series data and abnormal data.

[0015] Structured data includes tabular data in the database, function call relationships in the code, and change records in the version control system.

[0016] Unstructured text data includes requirement documents and technical specifications.

[0017] Time series data includes project progress and test results.

[0018] Abnormal data includes failure logs and abnormal feedback information during the test process.

[0019] As a preferred solution of the customized software function point identification and diagnosis system based on big data technology described in the present invention, the preprocessing module receives the data provided by the data acquisition module and constructs the data into a data set. , For the first data points, the Gaussian noise model is used to remove outliers in the data, and the exponential weighting function with dynamic weighting is used to optimize the data distribution. The data with data distribution trend is passed to the function point identification module, which is expressed as:

[0020]

[0021] in, is the data mean, is the data standard deviation, is a noise term with a mean of zero and a variance of The normal distribution of is the normalized data point, are optimized and standardized data points.

[0022] As a preferred solution of the customized software function point identification and diagnosis system based on big data technology described in the present invention, the function point identification module is connected with the preprocessing module and the diagnosis module, and based on the improved vector space model and deep text embedding, the function points are identified by semantic understanding of the requirement documents and technical specifications, and the local text feature extraction is performed using the convolution kernel, and the context understanding capability is performed in combination with the dynamic word vector embedding, which is expressed as:

[0023]

[0024] in, and are two texts to be compared. and yes and In Features The weight on is the embedding vector of the text, is the convolution kernel function, which is used to extract local information of text. is the text similarity.

[0025] After obtaining the similarity, the function point recognition module makes function point judgment according to the preset similarity threshold.

[0026] After completing the calculation of all data, the identified function points are converted into structured vector representations, and a vectorized data set containing all function points is output.

[0027] As a preferred solution of the customized software function point identification and diagnosis system based on big data technology described in the present invention, the diagnosis module is connected to the function point dynamic optimization module, receives the vectorized data set, and compares it with the original data to evaluate the description consistency of the function point between different documents and implementations, which is expressed as:

[0028] ,

[0029] in, Score consistency. and They are respectively the expected and actual function point descriptions, and They are and In the Description of the characteristics, is the consistency discriminant function, It is a fuzzy matching function to reduce the impact of errors.

[0030] As a preferred solution of the customized software function point identification and diagnosis system based on big data technology described in the present invention, the function point dynamic optimization module introduces a Gaussian regression model to predict the change trend of function points by analyzing project requirements and historical cost data, and adjusts resource allocation in real time to achieve dynamic optimization of function points, which is expressed as:

[0031]

[0032] in, To optimize the variables, is the Gaussian process regression model, is an estimate of uncertainty.

[0033] As a preferred solution of the customized software function point identification and diagnosis system based on big data technology described in the present invention, when the cost management module detects that function point optimization is being performed, it is connected to the data acquisition module to collect historical cost data, current expenditure records and project progress information, construct a time series, and use a time series model to dynamically evaluate and predict project costs, which is expressed as:

[0034] ,

[0035] in, For the original cost estimate, For the Item cost, For improved cost estimates, is the exponential decay coefficient.

[0036] Generate reports regularly to output estimated costs and actual budget usage.

[0037] Another object of the present invention is to provide a customized software function point identification and diagnosis method based on big data technology, which can identify and vectorize function points through improved vector space model and deep text embedding technology, extract function points from requirement documents and technical specifications and generate vector representation, thereby improving the accuracy and efficiency of identification. The problem of insufficient accuracy of current software function point identification and diagnosis technology is solved.

[0038] As a preferred solution of the customized software function point identification and diagnosis system based on big data technology described in the present invention, it includes: a data acquisition module that collects multi-source data in real time and transmits it to a preprocessing module, the preprocessing module cleans and standardizes the data, and provides the results to a function point identification module, a diagnosis module and a cost management module, the function point identification module extracts function points from the data and represents them quantitatively, and transmits the data to the diagnosis module, the diagnosis module receives the function point data and detects the consistency between the description and the implementation, and feeds back the inconsistent information to the function point dynamic optimization module, the function point dynamic optimization module optimizes the number of function points and resource allocation according to the diagnosis results, and updates the adjustment results to the data storage, the cost management module uses the latest configuration data to perform dynamic cost forecasting, and feeds back to the project management level for budget adjustment when there is a risk of overspending.

[0039] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement a step of a customized software function point identification and diagnosis method based on big data technology.

[0040] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a customized software function point identification and diagnosis method based on big data technology.

[0041] Beneficial effects of the present invention: The customized software function point identification and diagnosis method based on big data technology provided by the present invention uses fuzzy matching functions and consistency judgment functions to evaluate the matching degree between function points in different documents and implementations, ensuring the alignment between the function point description and the actual implementation, improving the consistency checking capability of the system, and reducing the deviation between the function point definition and implementation. Cost prediction is performed when performing function point maintenance to avoid the situation where the cost is higher than the value. The present invention achieves better results in terms of accuracy and cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0043] Figure 1 An overall flow chart of a customized software function point identification and diagnosis system based on big data technology provided for the first embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0045] Example 1, reference Figure 1 , which is an embodiment of the present invention, provides a customized software function point identification and diagnosis system based on big data technology, including: a data acquisition module 100, a preprocessing module 200, a function point identification module 300, a diagnosis module 400, a function point dynamic optimization module 500 and a cost management module 600.

[0046] The data acquisition module 100 is used to acquire multi-source data in real time.

[0047] The preprocessing module 200 is used to process multi-source data and provide standardized input for the function point identification module 300 , the diagnosis module 400 , and the cost management module 600 .

[0048] The function point identification module 300 is used to identify and vectorize function points in a document.

[0049] The diagnosis module 400 is used to detect the consistency of the function point description.

[0050] The function point dynamic optimization module 500 is used to optimize the number of function points and configure resources.

[0051] The cost management module 600 is used to dynamically predict project costs.

[0052] Furthermore, the data acquisition module 100 is connected to the preprocessing module 200 to transfer the acquired data to the preprocessing module 200 .

[0053] The data collection module 100 collects structured data, unstructured text data, time series data and abnormal data.

[0054] Structured data includes tabular data in the database, function call relationships in the code, and change records in the version control system.

[0055] Unstructured text data includes requirement documents and technical specifications.

[0056] Time series data includes project progress and test results.

[0057] Abnormal data includes failure logs and abnormal feedback information during the test process.

[0058] It should be noted that the preprocessing module 200 receives the data provided by the data acquisition module 100 and constructs the data into a data set , For the first data points, the Gaussian noise model is used to remove outliers in the data, and the data distribution is optimized using a dynamically weighted exponential weighting function, and the data with data distribution trends are passed to the function point identification module 300, which is expressed as:

[0059]

[0060] in, is the data mean, is the data standard deviation, is a noise term with a mean of zero and a variance of The normal distribution of is the normalized data point, is an optimized standardized data point. The Gaussian noise model is used to detect and remove outliers in the data set. This process is based on the characteristics of the normal distribution. By setting the mean and standard deviation, outliers that deviate greatly from the data center trend are identified and filtered out. Traditional data cleaning methods usually use a simple mean or median to replace outliers, but this method cannot guarantee the natural distribution characteristics of the data. The use of the Gaussian noise model in the present invention helps to capture the actual distribution state of the data, thereby optimizing the data cleaning process. Since the function point identification module 300 is highly dependent on the accuracy of the data, this cleaning through the Gaussian noise model not only improves the data quality, but also improves the accuracy and reliability of function point identification.

[0061] It should also be noted that the function point identification module 300 is connected to the preprocessing module 200 and the diagnosis module 400. Based on the improved vector space model and deep text embedding, the function points are identified by semantic understanding of the requirement documents and technical specifications, and the local text feature extraction is performed using the convolution kernel, and the context understanding capability is performed in combination with the dynamic word vector embedding, which is expressed as:

[0062]

[0063] in, and are two texts to be compared. and yes and In Features The weight on is the embedding vector of the text, is the convolution kernel function, which is used to extract local information of text. is the text similarity.

[0064] After obtaining the similarity, the function point identification module 300 performs function point determination according to a preset similarity threshold.

[0065] After completing the calculation of all data, the identified function points are converted into a structured vector representation, and a vectorized data set containing all function points is output. The combination of the vector space model and the deep text embedding technology improves the accuracy of function point identification by deeply understanding the semantic information in the function point description. Traditional function point identification is mostly based on keyword matching or shallow vector models, which makes it difficult to understand complex semantic information. The deep embedding in the present invention incorporates the contextual information of the text into the analysis through word vector technology, which can capture the implicit relationship in the function point description, thereby more accurately distinguishing the subtle differences between function points. Compared with the prior art, the present invention realizes efficient function point extraction through deep semantic analysis, providing a high-quality data foundation for subsequent diagnosis and optimization.

[0066] Introducing convolution kernel functions to extract local text features in text analysis can capture important local information and phrase structures in function point descriptions. Convolution kernels are widely used in traditional image processing, but using them to extract local text features is innovative, especially in function point recognition scenarios, where they can efficiently extract features of phrases such as "function module" and "operation parameters". This local feature extraction method further improves the accuracy of function point recognition, allowing the system to more sensitively detect subtle changes in requirement descriptions, providing more detailed support for recognition and diagnosis.

[0067] Similarity calculation is the core of function point judgment. By combining convolution kernel and dynamic word vector embedding, function point recognition module 300 can perform refined calculation of text similarity. The set similarity threshold helps the system automatically judge the similarity between two function point descriptions, avoiding human intervention. Traditional similarity judgment often leads to low recognition accuracy due to a single judgment dimension or unreasonable threshold setting. The dual-layer feature extraction and threshold control mechanism of the present invention significantly improves the accuracy and automation level of recognition, laying a solid foundation for efficient function point recognition and data output.

[0068] Furthermore, the diagnosis module 400 is connected to the function point dynamic optimization module 500, receives the vectorized data set, and compares it with the original data to evaluate the consistency of the description of the function point between different documents and implementations, which is expressed as:

[0069] ,

[0070] in, Score consistency. and They are respectively the expected and actual function point descriptions, and They are and In the Description of the characteristics, is the consistency discriminant function, It is a fuzzy matching function to reduce the impact of errors.

[0071] It should be noted that the function point dynamic optimization module 500 analyzes project requirements and historical cost data, introduces a Gaussian regression model to predict the change trend of function points, and adjusts resource allocation in real time to achieve dynamic optimization of function points, which is expressed as:

[0072]

[0073] in, To optimize the variables, is the Gaussian process regression model, is an estimate of uncertainty.

[0074] It should also be noted that when the cost management module 600 detects that function point optimization is being performed, it connects with the data acquisition module 100 to collect historical cost data, current expenditure records, and project progress information, construct a time series, and use a time series model to dynamically evaluate and predict project costs, which is expressed as:

[0075] ,

[0076] in, For the original cost estimate, For the Item cost, For improved cost estimates, is the exponential decay coefficient.

[0077] Generate reports regularly to output estimated costs and actual budget usage.

[0078] Example 2, an embodiment of the present invention, provides a customized software function point identification and diagnosis system based on big data technology. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0079] First, in order to verify its performance in function point identification accuracy, consistency diagnosis and cost management, the experimental objects are the requirements documents, code base, project schedule and test logs of a software project. The experimental data covers structured data (database tables, function call relationships), unstructured text data (requirements documents, technical specifications), time series data (project progress, test results) and abnormal data (test failure logs, abnormal feedback information).

[0080] Table 1 Experimental data comparison table

[0081]

[0082] Through the analysis of the table data, it can be concluded that the content of the present invention has obvious advantages in function point recognition accuracy, consistency diagnosis and cost control. Experimental data show that the number of recognized function points is close to the actual number of function points in the requirement document, indicating that the recognition accuracy of the function point recognition module 300 is high, and the average recognition accuracy is more than 95%. This is due to the improvement of the system through deep text embedding and convolution kernel extraction technology, which enables the semantic information of the requirement document to be fully analyzed, thereby achieving accurate recognition of function points.

[0083] The diagnosis module 400 performs consistency checks based on the identified function points. The number of inconsistencies in the experimental data shows that the diagnosis module 400 can effectively detect the difference between the document description and the implementation. The diagnosis module 400 accurately identifies the difference between the requirement document and the actual implementation through fuzzy matching and consistency judgment methods. The average number of inconsistencies diagnosed in the data is less than 6, indicating that the system has a strong advantage in diagnostic accuracy.

[0084] The effectiveness of the function point dynamic optimization module 500 is reflected by the number of optimized function points and the cost saving rate. After dynamic optimization, the number of function points is appropriately adjusted (reduced by about 2-4 on average), and resource allocation is more reasonable, which significantly improves the cost saving rate of each project, reaching more than 10%, with the highest reaching 15.3%. This verifies the effect of the Gaussian regression model in the dynamic optimization of function points, making resource allocation more accurate and thus optimizing project costs.

[0085] In summary, this embodiment verifies the innovation and practicality of the system in the field of function point identification and diagnosis through detailed testing of function point identification, diagnosis and optimization. The table data shows the excellent performance of the system in various indicators. Compared with traditional function point identification and cost control methods, the present invention can more efficiently identify, diagnose and optimize project function points, while improving the accuracy and efficiency of cost control, providing a scientific solution for custom software project management.

[0086] Embodiment 3, an embodiment of the present invention provides a customized software function point identification and diagnosis method based on big data technology, including a data acquisition module 100 to collect multi-source data in real time and pass it to a preprocessing module 200, the preprocessing module 200 cleans and standardizes the data, and provides the results to a function point identification module 300, a diagnosis module 400 and a cost management module 600, the function point identification module 300 extracts function points from the data and represents them quantitatively, and passes the data to the diagnosis module 400, the diagnosis module 400 receives the function point data and detects the consistency between the description and the implementation, and feeds back the inconsistent information to the function point dynamic optimization module 500, the function point dynamic optimization module 500 optimizes the number of function points and resource allocation according to the diagnosis results, and updates the adjustment results to the data storage, the cost management module 600 uses the latest configuration data to perform dynamic cost forecasting, and feeds back to the project management level for budget adjustment when there is a risk of overspending.

[0087] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A customized software function point identification and diagnosis system based on big data technology, characterized in that: include: A data collection module (100), a preprocessing module (200), a function point identification module (300), a diagnosis module (400), a function point dynamic optimization module (500) and a cost management module (600); The data acquisition module (100) is used to acquire multi-source data in real time; The preprocessing module (200) is used to process multi-source data and provide standardized input for the function point identification module (300), the diagnosis module (400), and the cost management module (600); The function point identification module (300) is used to identify and vectorize function points in a document; The diagnostic module (400) is used to detect the consistency of the function point description; The function point dynamic optimization module (500) is used to optimize the number of function points and configure resources; The cost management module (600) is used to dynamically predict project costs; The function point identification module (300) is connected to the preprocessing module (200) and the diagnosis module (400), and based on the improved vector space model and deep text embedding, identifies the function points through the semantic understanding of the requirement documents and technical specifications, uses the convolution kernel to extract local text features, and combines the dynamic word vector embedding to perform context understanding, which is expressed as: ， in, and are two texts to be compared. and yes and In Features The weight on is the embedding vector of the text, is the convolution kernel function, which is used to extract local information of text. is the text similarity; After obtaining the similarity, the function point identification module (300) performs function point judgment according to a preset similarity threshold; After all data are calculated, the identified function points are converted into structured vector representations, and a vectorized data set containing all function points is output; The diagnosis module (400) is connected to the function point dynamic optimization module (500), receives the vectorized data set, compares it with the original data, and evaluates the consistency of the description of the function point between different documents and implementations, which is expressed as: ， in, Score consistency. and They are respectively the expected and actual function point descriptions, and They are and In the Description of the characteristics, is the consistency discriminant function, It is a fuzzy matching function to reduce the impact of errors; The function point dynamic optimization module (500) analyzes project requirements and historical cost data, introduces a Gaussian regression model to predict the change trend of function points, and adjusts resource allocation in real time to achieve dynamic optimization of function points, which is expressed as: ， in, To optimize the variables, is the Gaussian process regression model, is the uncertainty estimate; When the cost management module (600) detects that function point optimization is being performed, it connects with the data collection module (100) to collect historical cost data, current expenditure records and project progress information, construct a time series, and use a time series model to dynamically evaluate and predict project costs, which is expressed as: ， in, For the original cost estimate, For the Item cost, For improved cost estimates, is the exponential decay coefficient; Generate reports regularly to output estimated costs and actual budget usage.

2. The customized software function point identification and diagnosis system based on big data technology according to claim 1, characterized in that: The data acquisition module (100) is connected to the preprocessing module (200) and transmits the acquired data to the preprocessing module (200); The data collection module (100) collects structured data, unstructured text data, time series data and abnormal data; Structured data includes table data in the database, function call relationships in the code, and change records in the version control system; Unstructured text data includes requirements documents and technical specifications; Time series data includes project progress and test results; Abnormal data includes failure logs and abnormal feedback information during the test process.

3. The customized software function point identification and diagnosis system based on big data technology as claimed in claim 2, characterized in that: The preprocessing module (200) receives the data provided by the data acquisition module (100) and constructs the data into a data set , For the first data points, outliers in the data are removed by using a Gaussian noise model, and the data distribution is optimized by using a dynamically adjusted exponential weighting function, and the data with data distribution trends are passed to the function point identification module (300), which is expressed as: ， in, is the data mean, is the data standard deviation, is a noise term with a mean of zero and a variance of The normal distribution of is the normalized data point, are optimized and standardized data points.

4. A method for using the customized software function point identification and diagnosis system based on big data technology as described in any one of claims 1 to 3, characterized in that: The data acquisition module (100) collects multi-source data in real time and transmits it to the preprocessing module (200). The preprocessing module (200) cleans and standardizes the data and provides the results to the function point identification module (300), the diagnosis module (400) and the cost management module (600). The function point identification module (300) extracts function points from the data and quantifies them, and transmits the data to the diagnosis module (400). The diagnosis module (400) receives the function point data and detects the consistency between the description and the implementation, and feeds back the inconsistency information to the function point dynamic optimization module (500). The function point dynamic optimization module (500) optimizes the number of function points and resource allocation according to the diagnosis results, and updates the adjustment results to the data storage. The cost management module (600) uses the latest configuration data to perform dynamic cost prediction, and feeds back to the project management level for budget adjustment when there is a risk of overspending.

Images