An innovation achievement evaluation system based on big data

Abstract

The present invention relates to the field of data analysis technology, specifically to an innovation achievement evaluation system based on big data, the system comprising: a data capture module that screens the original data set of an enterprise's scientific and technological innovation project, extracts time series data, distinguishes it into cost data, innovation project milestone dates, and quality indicators, calculates the cost change rate, project progress deviation, and quality change index of multiple innovation projects, and statistically calculates the average value and standard deviation of multiple indicators to establish a preliminary project deviation analysis. In the present invention, by accurately analyzing the time series data of scientific and technological projects, monitoring and analyzing the cost, progress, and quality indicators, time series decomposition and independent analysis of seasonality and trend deviations improve the accuracy of identifying cyclical changes and long-term development trends of projects, can accurately capture key data changes, promote the precise positioning and prediction of key change points in scientific and technological progress, and greatly enhance the evaluation and optimization effects of R&D results.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and in particular to an innovation achievement evaluation system based on big data. Background Art

[0002] The field of data analytics encompasses everything from basic data collection and storage to advanced data processing and analysis. The key task in this field is to transform large amounts of complex data into useful information and insights to support decision-making. Data analytics techniques, including statistical analysis, machine learning, big data technology, and artificial intelligence, are widely used in finance, healthcare, market research, operational optimization, and many other fields. Using data models and algorithms, analysts can identify trends, test hypotheses, predict market behavior, and improve business efficiency. With technological advancements, data analytics methods and tools are constantly evolving, with an increasing focus on real-time data processing and the development of automated decision support systems.

[0003] The Innovation Outcomes Evaluation System is a tool that utilizes data analysis techniques to assess and optimize the results of innovation activities. Its primary purpose is to help businesses or research institutions quantify and evaluate the effectiveness of their innovation projects, thereby guiding future R&D investment and strategic adjustments. By analyzing data collected from various data sources, the system provides detailed analysis of project performance, cost-effectiveness, and market response, enabling decision-makers to more intelligently choose which innovations to continue investing in and which ones may require reconsideration.

[0004] Existing technologies for processing large and complex data sets are often limited to basic data collection and statistical analysis, making it difficult to instantly identify and deeply analyze data changes. This results in traditional methods being unable to adjust strategies in a timely manner under rapidly changing market conditions, which in turn affects project cost-effectiveness and market responsiveness. For example, failure to accurately analyze seasonal and long-term trends in data can lead to misjudgments of market demand. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the existing technology and propose an innovation achievement evaluation system based on big data.

[0006] In order to achieve the above objectives, the present invention adopts the following technical solutions: A big data-based innovation achievement evaluation system includes:

[0007] The data capture module screens the original data sets of enterprise technology projects, extracts time series data, and separates them into cost data, project milestone dates, and quality indicators. It calculates the cost change rate, project schedule deviation, and quality change index of multiple projects, calculates the mean and standard deviation of multiple indicators, and establishes preliminary project deviation analysis.

[0008] The pattern recognition module extracts data items with standard deviations exceeding the norm from the preliminary project deviation analysis, performs time series decomposition on each data item, extracts seasonal and trend components, analyzes seasonal deviations and trend deviations separately, identifies abnormal change points corresponding to time nodes of scientific and technological progress, and extracts key morphological features;

[0009] The result optimization module analyzes the key morphological features, determines the cumulative effect of data deviation, reclassifies multiple data items by setting threshold classification, adjusts the deviation limit of each type of data, and establishes the deviation dynamic prediction analysis result;

[0010] The report generation module integrates the evaluation data based on the deviation dynamic prediction analysis results, conducts a multi-dimensional evaluation of the project including project health, risk assessment, and future development forecast, and forms a project analysis result.

[0011] As a further solution of the present invention, the steps for calculating the cost change rate are specifically as follows:

[0012] Filter cost data from the original data set of enterprise technology projects, classify them by project and time series, and generate classified cost data sets;

[0013] Calculating the total cost at each time point using the classified cost data set, comparing the cost at the current time point with the cost at the previous time point, and generating cost comparison data for each time point;

[0014] Based on the cost comparison data at each time point, the formula is used:

[0015]

[0016] Calculate and generate the cost change rate at each time point;

[0017] Where, ΔC i represents the cost change rate, C i Indicates the total cost at the current time point, C i-1 Indicates the total cost at the previous point in time.

[0018] As a further solution of the present invention, the steps for calculating the project progress deviation are specifically as follows:

[0019] Filter project milestone date information from the original data set of enterprise technology projects, classify the data according to projects and time series, and generate a project milestone classification data set;

[0020] Using the project milestone classification data set, calculating the difference between the actual completion date and the scheduled completion date of each project to generate time deviation data for each project;

[0021] From the time deviation data for each project, the formula is used:

[0022]

[0023] Calculate and generate project schedule deviations;

[0024] Among them, P dev Indicates the calculated project schedule deviation percentage, T actual Indicates the actual completion time, T planned Indicates the scheduled completion time.

[0025] As a further solution of the present invention, the steps for calculating the quality change index are specifically as follows:

[0026] Extract indicators related to project quality from the original data set of enterprise technology projects, classify and organize the data by project and time series, and generate a classified quality indicator data set;

[0027] Using the quality indicator data set, statistically analyzing the data at each time point, calculating the average quality score at each time point, and generating average quality score data at each time point;

[0028] Based on the average quality score data at each time point, the formula is used:

[0029]

[0030] Calculate and generate quality change index;

[0031] Among them, Q index Represents the quality change index, which is used to evaluate the degree of change in the quality score in the time series. current , Q previous Represent the average quality scores of the current and previous time points, respectively.

[0032] As a further solution of the present invention, the steps for obtaining the preliminary project deviation analysis are specifically as follows:

[0033] Integrating the cost change rate, the project schedule deviation, and the quality change index, aggregating indicators from multiple data sources to form a comprehensive performance indicator data set;

[0034] Based on the comprehensive performance indicator data set, a weighted analysis is performed on the key performance indicators of each project, a standardization process is applied to ensure that the weights of each indicator match, and a weighted average score is calculated to obtain a comprehensive performance score for each project;

[0035] Using the composite performance score for each item, use the formula:

[0036] D score=α·ΔC i +β·P dev +γ·Q index

[0037] Calculate the deviation for each item and establish a preliminary item deviation analysis;

[0038] Among them, D score It represents the comprehensive deviation score of the project, which is used to comprehensively evaluate the overall performance of the project. α, β, and γ are the weights of the cost change rate, project schedule deviation, and quality change index, respectively. They are adjusted according to the focus of project management to reflect their criticality to project success. ΔC i 、P dev , Q index They represent the cost change rate, project schedule deviation and quality change index extracted from the dataset respectively.

[0039] As a further solution of the present invention, the steps of obtaining the key morphological features are specifically as follows:

[0040] Screening the data items in the preliminary project deviation analysis, calculating the standard deviation of each data item, comparing it with a preset threshold, identifying data items whose standard deviation significantly exceeds the threshold, and generating a list of data with abnormal standard deviations;

[0041] Performing time series decomposition on the abnormal data items identified in the standard deviation abnormal data list, extracting and separating the seasonality and trend components of each data item, ensuring the independence of each component, and generating decomposed seasonality and trend data;

[0042] The statistical model was applied to analyze the decomposed seasonality and trend data using the formula:

[0043] A i =β·S i +γ·T i

[0044] Identify key time nodes and abnormal change points, and generate abnormal change analysis results;

[0045] Among them, A i Represents the comprehensive anomaly analysis score, which is used to quantify the degree of anomaly. i 、T i Refers to the quantitative scores of seasonal and trend components, respectively. β and γ are coefficients used to adjust the effects of seasonal and trend components, providing a means of adjusting the sensitivity of the analysis;

[0046] From the abnormal change analysis results, key morphological features corresponding to the time nodes of scientific and technological progress are extracted.

[0047] As a further solution of the present invention, the steps for obtaining the offset dynamic prediction analysis results are specifically as follows:

[0048] Analyzing the data extracted from the key morphological features, evaluating the cumulative effect of deviations for each data point, calculating the cumulative deviations, and determining which deviations are significantly higher than a normal level, and generating a list of cumulative deviation effects;

[0049] According to the list of cumulative effects of deviations, differentiated threshold levels are set, data items are reclassified using threshold grading logic, deviation limits are set for each category, and a list of reclassified data items is generated;

[0050] Using the results from the reclassified data item list, adjust the deviation limits for each category of data using the formula:

[0051] L new =θ·L old +ξ

[0052] Redefine the deviation limit of each category to generate adjusted deviation limit data;

[0053] Among them, L new Indicates the current deviation limit, which is used to control and predict data deviation, L old is the original deviation limit, representing the previous classification standard, and θ and ξ are adjustment coefficients used to fine-tune the deviation limit to match the current analysis requirements;

[0054] Based on the adjusted deviation limit data, a deviation dynamic prediction model is established to predict future deviation behaviors, and based on the changes in the adjusted deviation limit dynamic prediction data, a deviation dynamic prediction analysis result is generated.

[0055] As a further solution of the present invention, the steps for obtaining the project analysis results are specifically as follows:

[0056] Extract key data points from the dynamic forecast analysis results of the offset, combine each project dimension including cost, time and quality, calculate comprehensive health indicators, and generate preliminary comprehensive assessment data;

[0057] Utilizing the comprehensive assessment data, applying a quantitative risk assessment model to analyze potential risks and impacts, and generating risk assessment results;

[0058] Combining the risk assessment results and the health index, the formula is adopted:

[0059] P future =β·R risk +γ·H current

[0060] Conduct forecast analysis of future development, predict the future performance of the project by adjusting the weights of risk and health, and generate future development forecast results;

[0061] Among them, P future Represents the future forecast value of the project, R risk Represents the quantitative results of risk assessment,

[0062] H current represents the current project health index, β and γ are weight coefficients for adjusting the impact of risk and health;

[0063] Summarize the preliminary comprehensive assessment data, the risk assessment results and the future development forecast results, conduct a multi-dimensional assessment of the project, analyze the overall health status, risk level and future trends of the project, and generate project analysis results.

[0064] Compared with the prior art, the advantages and positive effects of the present invention are:

[0065] This invention monitors and analyzes cost, progress, and quality indicators through precise analysis of time series data from scientific and technological projects. Time series decomposition and independent analysis of seasonality and trend deviations improve the accuracy of identifying cyclical changes and long-term development trends in projects. This approach accurately captures key data changes, facilitates the precise location and prediction of key change points in scientific and technological progress, and significantly enhances the evaluation and optimization of R&D results. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 is a system flow chart of the present invention;

[0067] Figure 2 A flow chart of the steps for calculating the cost change rate of the present invention;

[0068] Figure 3 A flow chart of the steps for calculating the project progress deviation of the present invention;

[0069] Figure 4 is a flow chart of the steps for calculating the quality change index of the present invention;

[0070] Figure 5 A flowchart of the steps for obtaining the preliminary project deviation analysis of the present invention;

[0071] Figure 6 Flowchart of the steps for obtaining the key morphological features of the present invention;

[0072] Figure 7 Flowchart of the steps for obtaining the results of the dynamic prediction analysis of the offset according to the present invention;

[0073] Figure 8The figure is a flow chart of the steps for obtaining the project analysis results of the present invention. DETAILED DESCRIPTION

[0074] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0075] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0076] Example 1

[0077] See also Figure 1 , an innovation achievement evaluation system based on big data includes:

[0078] The data capture module screens the original data sets of enterprise technology projects, extracts time series data, and separates them into cost data, project milestone dates, and quality indicators. It calculates the cost change rate, project schedule deviation, and quality change index of multiple projects, calculates the mean and standard deviation of multiple indicators, and establishes preliminary project deviation analysis.

[0079] The pattern recognition module extracts data items with standard deviations exceeding the norm from the preliminary project deviation analysis, performs time series decomposition on each data item, extracts seasonal and trend components, analyzes seasonal deviations and trend deviations separately, identifies abnormal change points corresponding to the time nodes of scientific and technological progress, and extracts key morphological features;

[0080] The result optimization module analyzes key morphological features, determines the cumulative effect of data deviation, reclassifies multiple data items by setting threshold classification, adjusts the deviation limit of each data type, and establishes the deviation dynamic prediction analysis results;

[0081] The report generation module integrates the assessment data based on the results of the dynamic prediction analysis of the offset, and conducts a multi-dimensional assessment of the project, including project health, risk assessment, and future development forecast, to form the project analysis results.

[0082] The preliminary project deviation analysis specifically includes the cost change rate, project schedule deviation, and quality change index. The key morphological characteristics include seasonal deviation, trend deviation, and abnormal change points. The results of the dynamic prediction analysis of deviation specifically refer to the cumulative effect of data deviation, threshold classification, and deviation limit adjustment. The multi-dimensional project evaluation includes project health, risk assessment, and future development forecast.

[0083] See also Figure 2 , the calculation steps of cost change rate are as follows:

[0084] Filter cost data from the original data set of enterprise technology projects, classify them by project and time series, and generate classified cost data sets;

[0085] To screen cost data, we first extract all cost-related data from the original data set of the enterprise's technology projects, including direct costs, indirect costs, fixed costs, and variable costs. Data classification involves distinguishing according to projects and time series. This process involves batch data sorting and verification to ensure that the cost data of each project is accurately archived. In this process, the acquisition of key parameters depends on the accurate labeling and timestamp information of the cost items in the data set to ensure the timeliness and accuracy of the data. Through this series of operations, the classified cost data set is obtained.

[0086] Using the classified cost data set, calculate the total cost at each time point, compare the cost at the current time point with the cost at the previous time point, and generate cost comparison data for each time point;

[0087] Using the classified cost data, we continue to analyze the total cost of each project at multiple time points, and compare the cost at the current time point with the cost at the previous time point. This comparison relies on the cost records at each time point, including labor costs and material costs. Through comparison, we can monitor the cost change trend. Key parameters include the amount of multiple types of costs and the accurate record of time points. These are all executed through the statistical function within the data system to ensure the accuracy and real-time nature of the data, and generate cost comparison data for each time point.

[0088] Based on the cost comparison data at each time point, the formula is used:

[0089]

[0090] Calculate and generate the cost change rate at each time point;

[0091] Where, ΔC i represents the cost change rate, C i Indicates the total cost at the current time point, C i-1 Indicates the total cost at the previous point in time.

[0092] formula:

[0093]

[0094] The benefit of this formula is that it can quantify the rate of change of costs at each point in time. This not only helps management monitor the trend of cost changes, but also predicts future cost fluctuations, thereby making more effective budgeting and cost control decisions. Detailed explanation of the formula and the calculation process of the formula:

[0095] Setting C i = 105,000 yuan is the cost for this month, C i-1 =100,000 yuan is the cost of last month. According to the formula:

[0096]

[0097] The results show that costs have increased by 5% from last month to this month, indicating that cost management needs to pay attention to this growth trend and iteratively analyze the reasons for the cost increase. It is necessary to investigate and verify the reasons for the increase in material costs or labor costs.

[0098] See also Figure 3 , the specific steps for calculating project schedule deviation are:

[0099] Filter project milestone date information from the original data set of enterprise technology projects, classify the data according to projects and time series, and generate a project milestone classification data set;

[0100] Project milestone date information is extracted from the original data set of the enterprise technology project and classified by project and time series. This process involves systematic classification and analysis of batch project data. The purpose is to effectively record and track the key time nodes of each project, ensure the integrity and availability of the data, and thus provide a reliable basic data source for subsequent progress deviation analysis. The classification processing of the data set includes not only data collection, but also preliminary data cleaning to remove invalid or erroneous time records, such as incorrect date formats or time nodes in the future time period. This step is the initial grasp of time control in project management, ensuring that each stage of the project can be executed as planned, thereby forming a milestone timeline for each project in the project management system. The data set obtained after classification directly affects the efficiency of project progress monitoring and deviation analysis.

[0101] Using the project milestone classification dataset, the difference between the actual completion date and the scheduled completion date of each project is calculated to generate the time deviation data of each project;

[0102] Using the resulting classified project milestone dataset, the actual completion date of each project is compared with the scheduled completion date. This process requires calculating the date difference, referring to the difference between working days and non-working days, and adjusting for estimated holidays. This comparative analysis helps project managers identify potential delays or early completion stages in the project, providing data support for project risk management and resource reallocation. Through this analysis, the project team can adjust resources and plans in a timely manner to address any uncertainty in the schedule. The generated time deviation data not only reflects the time management efficiency of each project but also serves as a key data point for evaluating project management performance, thereby providing decision support for senior management.

[0103] From the time deviation data for each project, use the formula:

[0104]

[0105] Calculate and generate project schedule deviations;

[0106] Among them, P dev Indicates the calculated project schedule deviation percentage, T actual Indicates the actual completion time, T planned Indicates the scheduled completion time.

[0107] formula:

[0108]

[0109] The formula is useful because it not only quantifies the extent of project delays or early completion by calculating the percentage deviation between the project's actual completion time and the scheduled time, but also provides a standardized evaluation metric to help project teams compare the efficiency of project execution across different projects or within different time periods.

[0110] Detailed explanation of the formula and the process of formula calculation and derivation:

[0111] Assume that the scheduled completion date of a project is December 31, 2021, and the actual completion date is January 10, 2022. The schedule deviation is calculated using the formula. First, calculate the date difference:

[0112] T actual =2022-01-10

[0113] T planned =2021-12-31

[0114] The date difference is 10 days. Substituting the value into the formula:

[0115]

[0116] The results show that the actual project completion time was about 2.74% later than planned, which means that the project encountered delays during execution. This delay percentage can help the management team assess the severity of the delay and refer to whether measures need to be taken to adjust the project plan or optimize resource allocation.

[0117] See also Figure 4 , the calculation steps of the quality change index are as follows:

[0118] Extract indicators related to project quality from the original data set of enterprise technology projects, classify and organize the data by project and time series, and generate a classified quality indicator data set;

[0119] Quality-related indicators are extracted from the original data set of the project, and data classification and processing are performed based on differentiated projects and time nodes. In this way, the quality performance of each project in the differentiation stage can be determined. This classification method not only helps in subsequent quality analysis, but also provides a data basis for comparing the quality performance between differentiated projects. Through data processing and classification, the accuracy and reliability of the data can be ensured, providing support for quality management.

[0120] Using the quality indicator data set, statistical analysis is performed on the data at each time point, and the average quality score at each time point is calculated to generate the average quality score data at each time point;

[0121] Using the classified quality indicator data set, the data at each time point is analyzed and the average quality score at each time point is calculated. This process involves batch data processing and statistical calculations to ensure that the results of each calculation are accurate. This detailed data analysis helps to identify at which time points the project performs poorly and then take corresponding improvement measures.

[0122] According to the average quality score data at each time point, the formula is used:

[0123]

[0124] Calculate and generate quality change index;

[0125] Among them, Q index Represents the quality change index, which is used to evaluate the degree of change in the quality score in the time series. current , Q previous Represent the average quality scores of the current and previous time points, respectively.

[0126] formula:

[0127]

[0128] The benefit of the formula is that by combining the geometric mean of the quality scores at the current and previous time points, it improves the reflection of the sensitivity to changes in project quality and is suitable for scenarios with large changes, thereby more accurately measuring quality changes between projects.

[0129] Detailed explanation of the formula and the process of formula calculation and derivation:

[0130] Set Q current =85,Q previous =80, substituting into the formula we get:

[0131]

[0132] The results show that the quality score has improved from the previous time point to the current time point, increasing by about 6.06%, which shows a small progress in the project's quality management and helps to quantitatively evaluate the effectiveness of improvement measures.

[0133] See also Figure 5 ,The steps for obtaining the preliminary project deviation analysis are as follows:

[0134] Integrate cost change rate, project schedule deviation and quality change index, aggregate indicators from multiple data sources to form a comprehensive performance indicator data set;

[0135] To integrate key performance indicators, we first extracted and cleaned the cost change rate, project schedule deviation, and quality change index in the dataset using the target script. We then used SQL queries to merge the three indicators into a single table. This approach ensured data consistency and integrity. Next, we applied data normalization to unify the scales of multiple indicators, creating conditions for subsequent weighted calculations. This process involved calculating the mean and standard deviation of each indicator and adjusting each data point so that it was distributed on the same scale. Finally, we generated a comprehensive performance score for each project, which was used to evaluate the overall performance of the project.

[0136] Based on the comprehensive performance indicator data set, we conduct a weighted analysis of the key performance indicators of each project, apply standardization to ensure that the weights of each indicator match, and calculate the weighted average score to obtain the comprehensive performance score of each project.

[0137] Based on the comprehensive performance indicator data set, correlation analysis and regression testing are conducted to determine the degree of correlation between multiple indicators and project success. This analysis helps to analyze which factors have a significant impact on project success. Then, based on the analysis results, the weights of multiple indicators are preliminarily set. The simulation model is used to test the changes in results under differentiated weight settings, optimize the weight configuration, and establish the weight coefficient of each indicator to ensure that these weights can accurately reflect the contribution of multiple indicators to project success.

[0138] Using the comprehensive performance score for each project, the formula is:

[0139] D score =α·ΔC i +β·P dev +γ·Q index

[0140] Calculate the deviation for each item and establish a preliminary item deviation analysis;

[0141] Among them, D score It represents the comprehensive deviation score of the project, which is used to comprehensively evaluate the overall performance of the project. α, β, and γ are the weights of the cost change rate, project schedule deviation, and quality change index, respectively. They are adjusted according to the focus of project management to reflect their criticality to project success. ΔC i 、P dev , Q index They represent the cost change rate, project schedule deviation and quality change index extracted from the dataset respectively.

[0142] formula:

[0143] D score =α·ΔC i +β·P dev +γ·Q index

[0144] The benefit of the formula is that it can flexibly reflect the needs and characteristics of differentiated projects by dynamically adjusting the weight coefficients, thereby improving the accuracy and practicality of project deviation analysis.

[0145] Detailed explanation of the formula and the process of formula calculation and derivation:

[0146] Set the cost change rate ΔC for a project i is 0.05, indicating a 5% increase in cost and a project schedule deviation of P dev -0.1 means the project is 10% ahead of schedule. index If the weight coefficients α = 0.3, β = 0.5, and γ = 0.2 are set, the calculation process is as follows:

[0147] D score =0.3·0.05+0.5·(-0.1)+0.2·0.02=0.015-0.05+0.004=-0.031

[0148] The results show that although the project has improved in quality and slightly increased in cost, the overall deviation score is negative due to the early completion of the schedule, indicating that the overall performance of the project is good, but attention should be paid to other potential problems caused by schedule management.

[0149] See also Figure 6 , the steps for obtaining key morphological features are as follows:

[0150] Screen the data items in the preliminary project deviation analysis, calculate the standard deviation of each data item, compare it with the preset threshold, identify the data items whose standard deviation significantly exceeds the threshold, and generate a list of data with abnormal standard deviation;

[0151] Screen the data items in the preliminary project deviation analysis, calculate the standard deviation of each data item, compare it with the preset threshold, and determine which data items have standard deviations significantly exceeding the threshold. This step calculates the standard deviation of each data item and then compares it with the pre-set threshold to identify those abnormal data items. The standard deviation of the data items is significantly higher than the average level. These data items are identified as the focus of subsequent analysis, and then a list of all standard deviation abnormal data items is generated. This list will be used for time series decomposition analysis to understand the changing trends and cyclical changes of the data.

[0152] Perform time series decomposition on the abnormal data items identified in the standard deviation abnormal data list, extract and separate the seasonality and trend components of each data item, ensure the independence of each component, and generate decomposed seasonality and trend data;

[0153] Perform time series decomposition on data items identified as abnormal to separate the seasonal and trend components of each data item. The key to this step is to apply time series analysis technology to analyze the data and separate the main components of the data through mathematical modeling, including seasonal factors and long-term trends. This analysis helps to better analyze the cyclical fluctuations and trend changes of the data, providing a basis for outlier analysis. The decomposition results will be recorded and used for the next step of seasonal and trend deviation analysis.

[0154] Apply statistical models to analyze the decomposed seasonal and trend data using the formula:

[0155] A i =β·S i +γ·T i

[0156] Identify key time nodes and abnormal change points, and generate abnormal change analysis results;

[0157] Among them, A i Represents the comprehensive anomaly analysis score, which is used to quantify the degree of anomaly. i 、T i Refers to the quantitative scores of seasonal and trend components, respectively. β and γ are coefficients used to adjust the effects of seasonal and trend components, providing a means of adjusting the sensitivity of the analysis;

[0158] formula:

[0159] A i =β·S i +γ·T i

[0160] The benefit of the formula is that by combining the weighted sum of seasonal and trend components, the weights can be flexibly adjusted to highlight the target behavior of the data, thereby more accurately identifying key change points. The β and γ in the formula provide this flexibility, and analysts can adjust the parameters according to the situation to reflect the actual influence of seasonality and trend.

[0161] Detailed explanation of the formula and the process of formula calculation and derivation:

[0162] Set at a certain point in time, the seasonal score S i is 0.3, trend score T i is 0.7, and weights β = 0.5 and γ = 0.5 are set, then the calculation is:

[0163] A i =0.5·0.3+0.5·0.7=0.15+0.35=0.5

[0164] This means that the overall score of the current data point is 0.5, and a higher score indicates that there is a significant anomaly or change at this point.

[0165] The results show that the analysis successfully identifies key time nodes associated with scientific and technological progress, and the scores directly affect the overall evaluation and decision-making of the project.

[0166] From the results of abnormal change analysis, key morphological features corresponding to the time nodes of scientific and technological progress are extracted.

[0167] Extract key morphological features from the results of abnormal change analysis. This process includes quantitative analysis of identified abnormal change points, revealing key behavioral patterns of the data through graphical and numerical analysis methods. These patterns represent potential technological advances or risk points that require attention. By quantifying these patterns, key morphological features of the data can be extracted. These features help guide subsequent project decisions and strategy adjustments, ensuring the synchronization of scientific and technological development and project management, thereby generating a result that includes all key morphological features and providing empirical support for project management.

[0168] See also Figure 7 ,The specific steps for obtaining the results of the offset dynamic prediction analysis are as follows:

[0169] Analyze the data extracted from key morphological features, evaluate the cumulative effect of deviations at each data point, calculate the cumulative deviations, and determine which deviations are significantly higher than the normal level, and generate a list of cumulative deviation effects;

[0170] After analyzing the data of key morphological features, we determined those data items that significantly exceeded normal levels by calculating a cumulative effect list. This process involved evaluating the deviation of each data point and identifying anomalies through statistical methods and threshold judgment. This strategy not only simplified the subsequent data processing steps but also ensured the accuracy and efficiency of the analysis, resulting in a classified cumulative effect list of deviations.

[0171] According to the list of cumulative effects of deviations, differentiated threshold levels are set, data items are reclassified through threshold grading logic, deviation limits are set for each category, and a list of reclassified data items is generated;

[0172] Based on the list of cumulative effects of deviations that have been obtained, differentiated threshold levels are reset and data items are reclassified. This process adjusts the deviation limits of each category by comprehensively referring to the deviation size and the actual impact of the data. Statistical methods are used to dynamically adjust the deviation thresholds. With reference to the seasonal and cyclical characteristics of the data, data classification is made more precise and personalized, and a list of data items classified according to the new standards is obtained, providing accurate data support for subsequent predictive analysis.

[0173] Using the results from the reclassified data item list, adjust the deviation limits for each category of data using the formula:

[0174] L new =θ·L old +ξ

[0175] Redefine the deviation limit of each category to generate adjusted deviation limit data;

[0176] Among them, L new Indicates the current deviation limit, which is used to control and predict data deviation, L old is the original deviation limit, representing the previous classification standard, and θ and ξ are adjustment coefficients used to fine-tune the deviation limit to match the current analysis requirements;

[0177] formula:

[0178] L new =θ·L old +ξ

[0179] The benefit of the formula is that by introducing the adjustment parameters θ and ξ, the deviation bounds can be flexibly adjusted to match differentiated data analysis requirements, thereby more accurately predicting data deviations in future time periods.

[0180] Detailed explanation of the formula and the process of formula calculation and derivation:

[0181] Set the raw deviation limit L old=10, the adjustment coefficient θ is 1.2 to amplify the limit, the fixed value ξ is 2 to handle the system error, insert the numerical value for calculation, L new =1.2×10+2=14, current deviation limit L new It is 14, indicating that the adjusted limit is more relaxed, matching the current data trend;

[0182] The results show that by adjusting the deviation bound, the sensitivity and responsiveness of data classification can be more effectively controlled, enabling the model to match a wider range of data fluctuations and improving the flexibility and accuracy of predictions.

[0183] Based on the adjusted deviation limit data, an offset dynamic prediction model is established to predict future deviation behaviors. Based on the adjusted deviation limit dynamic prediction data changes, the offset dynamic prediction analysis results are generated.

[0184] Through the adjusted deviation limits, a dynamic prediction model is established to predict data deviations in future time periods. This process uses time series analysis technology and machine learning models to evaluate the development trend of deviations, and refers to the currently adjusted deviation limits. The established prediction model is not only highly adaptable, but also able to dynamically adjust the prediction strategy according to actual data changes. The generated deviation dynamic prediction analysis results provide a scientific basis for future data management and risk control, and have high practical value and strategic guidance significance.

[0185] See also Figure 8 , the specific steps for obtaining project analysis results are:

[0186] Extract key data points from the results of the dynamic forecast analysis of the offset, combine each project dimension including cost, time and quality, calculate comprehensive health indicators, and generate preliminary comprehensive assessment data;

[0187] Key data points are extracted from the results of the dynamic offset prediction analysis. Comprehensive health indicators are calculated based on each project dimension, including cost, time, and quality. Through data standardization and multi-dimensional weighted analysis, the combined impact of multiple dimensions is calculated, and preliminary comprehensive assessment data is generated based on this. This data helps project managers understand the overall health of the project, effectively monitor the project, and make timely adjustments.

[0188] Utilize comprehensive assessment data and apply quantitative risk assessment models to analyze potential risks and impacts and generate risk assessment results;

[0189] A quantitative risk assessment model is applied to analyze potential risks and their impacts. The model combines on-site monitoring data with statistical analysis methods and machine learning techniques to assess the various risks faced by the project, including budget overruns and schedule delays. The model also generates risk assessment results by calculating the severity and likelihood of potential impacts. This result is crucial for project risk management and assists the project team in formulating corresponding risk response strategies.

[0190] Combining the risk assessment results and health indicators, the formula is used:

[0191] P future =β·R risk +γ·H current

[0192] Conduct forecast analysis of future development, predict the future performance of the project by adjusting the weights of risk and health, and generate future development forecast results;

[0193] Among them, P future Represents the future forecast value of the project, R risk Represents the quantitative results of risk assessment,

[0194] H current represents the current project health index, β and γ are weight coefficients for adjusting the impact of risk and health;

[0195] formula:

[0196] P future =β·R risk +γ·H current

[0197] The benefit of the formula is that by adjusting the weight parameters of risk and health, it can flexibly reflect the prediction of future project performance in different project management environments and project characteristics. This approach increases the accuracy of the prediction and the breadth of its application.

[0198] Detailed explanation of the formula and the process of formula calculation and derivation:

[0199] Set in the project, R risk is the current risk assessment value, which is determined to be 0.65 through analysis of similar situations in past projects. current is the current health indicator, which is 0.85 through real-time data monitoring. The weight parameters β and γ are set to 0.6 and 0.4 respectively based on the data and expert solution. The calculation process is as follows:

[0200] P future =0.6·0.65+0.4·0.85=0.39+0.34=0.73

[0201] The results show that the project's future performance is assessed at 0.73, based on its current risk and health status. This indicates that the project is in a relatively stable state, but potential risks still need to be paid attention to.

[0202] Summarize preliminary comprehensive assessment data, risk assessment results and future development forecast results, conduct multi-dimensional project assessment, analyze the overall health status, risk level and future trends of the project, and generate project analysis results.

[0203] A multi-dimensional project assessment is conducted based on preliminary comprehensive assessment data, risk assessment results, and future development forecast results. This process includes analysis of the overall health status, risk level, and future trends of the project. By using comprehensive data models and multivariate statistical methods, project analysis results are formed to provide a scientific basis for project decision-making and ensure that the project can proceed smoothly on the expected track.

[0204] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. An innovation achievement evaluation system based on big data, characterized by: The system comprises: The data capture module screens the original data sets of enterprise scientific and technological innovation projects, extracts time series data, and divides them into cost data, innovation project milestone dates, and quality indicators. It calculates the cost change rate, project schedule deviation, and quality change index of multiple innovation projects, calculates the mean and standard deviation of multiple indicators, and establishes a preliminary project deviation analysis. The pattern recognition module extracts data items with standard deviations exceeding the norm from the preliminary project deviation analysis, performs time series decomposition on each data item, extracts seasonal and trend components, analyzes seasonal deviations and trend deviations separately, identifies abnormal change points corresponding to time nodes of scientific and technological progress, and extracts key morphological features; The steps for obtaining the key morphological features are specifically as follows: Screening the data items in the preliminary project deviation analysis, calculating the standard deviation of each data item, comparing it with a preset threshold, identifying data items whose standard deviation significantly exceeds the threshold, and generating a list of data with abnormal standard deviations; Performing time series decomposition on the abnormal data items identified in the standard deviation abnormal data list, extracting and separating the seasonality and trend components of each data item, ensuring the independence of each component, and generating decomposed seasonality and trend data; The statistical model was applied to analyze the decomposed seasonality and trend data using the formula: A i =β·S i +γ·T i Identify key time nodes and abnormal change points, and generate abnormal change analysis results; Among them, A i Represents the comprehensive anomaly analysis score, which is used to quantify the degree of anomaly. i 、T i Refers to the quantitative scores of seasonal and trend components, respectively. β and γ are coefficients used to adjust the effects of seasonal and trend components, providing a means of adjusting the sensitivity of the analysis; Extracting key morphological features corresponding to time nodes of scientific and technological progress from the abnormal change analysis results; The result optimization module analyzes the key morphological features, determines the cumulative effect of data deviation, reclassifies multiple data items by setting threshold classification, adjusts the deviation limit of each type of data, and establishes the deviation dynamic prediction analysis result; The report generation module integrates the evaluation data based on the deviation dynamic prediction analysis results, conducts a multi-dimensional evaluation of the innovation project including project health, risk assessment, and future development forecast, and forms a project analysis result.

2. The innovation achievement evaluation system based on big data according to claim 1 is characterized in that: The specific steps for calculating the cost change rate are: Filter cost data from the original data set of enterprise technology projects, classify them by project and time series, and generate classified cost data sets; Calculating the total cost at each time point using the classified cost data set, comparing the cost at the current time point with the cost at the previous time point, and generating cost comparison data for each time point; Based on the cost comparison data at each time point, the formula is used: Calculate and generate the cost change rate at each time point; Where, ΔC i represents the cost change rate, C i Indicates the total cost at the current time point, C i-1 Indicates the total cost at the previous point in time.

3. The innovation achievement evaluation system based on big data according to claim 2 is characterized in that: The specific steps for calculating the project progress deviation are as follows: Filter project milestone date information from the original data set of enterprise technology projects, classify the data according to projects and time series, and generate a project milestone classification data set; Using the project milestone classification data set, calculating the difference between the actual completion date and the scheduled completion date of each project to generate time deviation data for each project; From the time deviation data for each project, the formula is used: Calculate and generate project schedule deviations; Among them, P dev Indicates the calculated project schedule deviation percentage, T actual Indicates the actual completion time, T planned Indicates the scheduled completion time.

4. The innovation achievement evaluation system based on big data according to claim 3 is characterized in that: The calculation steps of the quality change index are specifically as follows: Extract indicators related to project quality from the original data set of enterprise technology projects, classify and organize the data by project and time series, and generate a classified quality indicator data set; Using the quality indicator data set, statistically analyzing the data at each time point, calculating the average quality score at each time point, and generating average quality score data at each time point; Based on the average quality score data at each time point, the formula is used: Calculate and generate quality change index; Among them, Q index Represents the quality change index, which is used to evaluate the degree of change in the quality score in the time series. current , Q previous Represent the average quality scores of the current and previous time points, respectively.

5. The innovation achievement evaluation system based on big data according to claim 4 is characterized in that: The steps for obtaining the preliminary project deviation analysis are specifically as follows: Integrating the cost change rate, the project schedule deviation, and the quality change index, aggregating indicators from multiple data sources to form a comprehensive performance indicator data set; Based on the comprehensive performance indicator data set, a weighted analysis is performed on the key performance indicators of each project, a standardization process is applied to ensure that the weights of each indicator match, and a weighted average score is calculated to obtain a comprehensive performance score for each project; Using the composite performance score for each item, use the formula: D score =α·ΔC i +β·P dev +γ·Q index Calculate the deviation for each item and establish a preliminary item deviation analysis; Among them, D score It represents the comprehensive deviation score of the project, which is used to comprehensively evaluate the overall performance of the project. α, β, and γ are the weights of the cost change rate, project schedule deviation, and quality change index, respectively. They are adjusted according to the focus of project management to reflect their criticality to project success. ΔC i 、P dev , Q index They represent the cost change rate, project schedule deviation and quality change index extracted from the dataset respectively.

6. The innovation achievement evaluation system based on big data according to claim 1 is characterized in that: The steps for obtaining the offset dynamic prediction analysis results are specifically as follows: Analyzing the data extracted from the key morphological features, evaluating the cumulative effect of deviations for each data point, calculating the cumulative deviations, and determining which deviations are significantly higher than a normal level, and generating a list of cumulative deviation effects; According to the list of cumulative effects of deviations, differentiated threshold levels are set, data items are reclassified using threshold grading logic, deviation limits are set for each category, and a list of reclassified data items is generated; Using the results from the reclassified data item list, adjust the deviation limits for each category of data using the formula: L new =θ·L old +ξ Redefine the deviation limit of each category to generate adjusted deviation limit data; Among them, L new Indicates the current deviation limit, which is used to control and predict data deviation, L old is the original deviation limit, representing the previous classification standard, and θ and ξ are adjustment coefficients used to fine-tune the deviation limit to match the current analysis requirements; Based on the adjusted deviation limit data, a deviation dynamic prediction model is established to predict future deviation behaviors, and based on the changes in the adjusted deviation limit dynamic prediction data, a deviation dynamic prediction analysis result is generated.

7. The innovation achievement evaluation system based on big data according to claim 6 is characterized in that: The steps for obtaining the project analysis results are specifically as follows: Extract key data points from the dynamic forecast analysis results of the offset, combine each project dimension including cost, time and quality, calculate comprehensive health indicators, and generate preliminary comprehensive assessment data; Utilizing the comprehensive assessment data, applying a quantitative risk assessment model to analyze potential risks and impacts, and generating risk assessment results; Combining the risk assessment results and the health index, the formula is adopted: P future =β·R risk +γ·H current Conduct forecast analysis of future development, predict the future performance of the project by adjusting the weights of risk and health, and generate future development forecast results; Among them, P future Represents the future forecast value of the project, R risk Represents the quantitative result of risk assessment, H current represents the current project health index, β and γ are weight coefficients for adjusting the impact of risk and health; Summarize the preliminary comprehensive assessment data, the risk assessment results and the future development forecast results, conduct a multi-dimensional assessment of the project, analyze the overall health status, risk level and future trends of the project, and generate project analysis results.

Images