Enterprise work data redisk method and device, electronic equipment and medium

Abstract

The invention provides an enterprise work data redisk method and device, electronic equipment and a medium, and relates to the technical field of data processing, and the method comprises the steps: carrying out the problem recognition, emotion analysis and quality evaluation of a redisk report of each employee according to a configured rule; according to the sentiment analysis result, the problem identification result and the quality evaluation result, generating redisk analysis data of the target employee in the target time period; comparing the historical redisk analysis data of the target employee with the redisk analysis data to generate change trend analysis data; determining a target problem of the target enterprise in the target time period based on the problem identification results of the different target employees; and generating a redisk summary report according to the target problem, the redisk analysis data of the different target employees and the change trend analysis data of the corresponding target employees. According to the method, emotion categories and problems in the redisk report are automatically analyzed, the quality of the redisk report is evaluated, and enterprise managers can visually master the working conditions of all employees.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method, device, electronic device and medium for replaying enterprise work data. Background Art

[0002] Existing enterprise data reviews are generally completed by department managers manually organizing and compiling data after employees from various departments fill out the forms freely. For large-scale enterprises, manual analysis and statistical methods are not only inefficient and highly subjective, but also lack quantifiable multi-dimensional indicators, making it difficult to accurately measure work status and review quality. Trend analysis is also lagging, making it impossible to timely discover fluctuations or improvement points in employee performance. Employee review reports are scattered and lack systematic integration methods, making it difficult to quickly extract common problems at the enterprise level from individual problems. As a result, the review results cannot be fed back into the actual enterprise management process. Summary of the Invention

[0003] The purpose of the embodiments of the present application is to provide a method, device, electronic device and medium for reviewing enterprise work data, so as to solve the above-mentioned problems existing in the prior art and to quantitatively review the enterprise's work data from multiple dimensions.

[0004] In a first aspect, the present invention provides a method for retrieving enterprise work data, the method comprising:

[0005] Obtaining review reports generated by different target employees of the target enterprise regarding their work performance during the target time period, as well as historical review analysis data for each target employee; wherein the historical review analysis data is obtained by reviewing and analyzing the historical review reports of the target employees during the historical time period;

[0006] For any target employee's review report, identify problems based on the configured problem identification rules and obtain problem identification results;

[0007] Performing sentiment analysis on the review report to obtain a sentiment analysis result;

[0008] Perform a quality evaluation on the replay report according to the configured quality evaluation rules to obtain a quality evaluation result;

[0009] Generate review analysis data of the target employee within the target time period based on the sentiment analysis results, the problem identification results, and the quality evaluation results;

[0010] Comparing the target employee's historical review analysis data with the review analysis data to generate change trend analysis data;

[0011] Based on the problem identification results of different target employees, determine the target problems of the target enterprise within the target time period;

[0012] A review summary report for the target enterprise within the target time period is generated based on the target problem, the review analysis data of different target employees and the change trend analysis data of the corresponding target employees.

[0013] In an optional embodiment, the review report includes: the target employee's work completion status within the target time period;

[0014] The question identification rules include: different keywords, different question words, and the question level, first threshold and second threshold corresponding to each question word;

[0015] According to the configured problem identification rules, the problem identification is performed on the replay report to obtain the problem identification results, including:

[0016] For any keyword, extract the problem description containing the keyword from the review report;

[0017] Analyze the problem description to obtain target problem words;

[0018] Counting the occurrences of each target question word and the corresponding target question word in the review report;

[0019] If the number of occurrences of the target question word is greater than a first threshold corresponding to the corresponding question word, the question word is determined as the first target word;

[0020] A problem identification result is generated based on the first target word and the corresponding number of occurrences, the problem level and the problem description, as well as the completion status of the work.

[0021] In an optional embodiment, sentiment analysis is performed on the review report to obtain sentiment analysis results, including:

[0022] Segmenting the review report to obtain a plurality of text paragraphs and a position of each text paragraph in the review report;

[0023] Perform semantic understanding on multiple text paragraphs respectively to obtain semantic understanding results corresponding to each text paragraph;

[0024] Different text paragraphs and corresponding semantic understanding results and positions are input into a pre-trained sentiment analysis model to obtain the sentiment analysis results of the review report.

[0025] In an optional embodiment, comparing the target employee's historical review analysis data with the review analysis data to generate change trend analysis data includes:

[0026] Comparing historical sentiment analysis results in the historical replay analysis data with the sentiment analysis results to generate sentiment change analysis data;

[0027] Comparing the historical problem identification results in the historical replay analysis data with the problem identification results to generate problem change analysis data;

[0028] Comparing the historical quality evaluation results in the historical replay analysis data with the quality evaluation results to generate quality change analysis data;

[0029] Based on the sentiment change analysis data, the problem change analysis data and the quality change analysis data, change trend analysis data is obtained.

[0030] In an optional embodiment, based on the problem identification results of different target employees, determining the target problems of the target enterprise within the target time period includes:

[0031] Count the total number of occurrences of different first target words in the problem recognition results of all target employees;

[0032] If the total number of occurrences of any first target word is greater than the second threshold, the first target word is used as the second target word;

[0033] Perform text clustering on different second target words to obtain target question clusters;

[0034] Determining the risk level corresponding to each problem cluster based on the problem level corresponding to each second target word included in different problem clusters and the completion status of the work;

[0035] Generate a problem cluster distribution map based on the proportion of each problem cluster to all problem clusters;

[0036] The target problems of the target enterprise within the target time period are obtained according to the problem cluster distribution map, the risk level corresponding to each problem cluster and each second target word contained in each problem cluster.

[0037] In an optional embodiment, the quality evaluation rules include different evaluation indicators and scoring rules and weights corresponding to each evaluation indicator;

[0038] Perform a quality evaluation on the replay report according to the configured quality evaluation rules to obtain a quality evaluation result, including:

[0039] For any evaluation indicator, score the review report based on the scoring rules corresponding to the evaluation indicator to obtain the score of the review report under the evaluation indicator;

[0040] The scores of the review report under different evaluation indicators are weighted and summed to obtain the quality evaluation result of the review report.

[0041] In an optional embodiment, a review summary report of the target enterprise within the target time period is generated based on the target question, the review analysis data of different target employees, and the change trend analysis data of the corresponding target employees, including:

[0042] Calculate the completion rate of the review of each target department based on the quality evaluation results of the review reports of each target employee in different target departments of the target enterprise;

[0043] Generate a review summary report for the target department within the target time period based on the target problem, the review completion rate, the review analysis data of each target employee in the target department, and the change trend analysis data of the corresponding target employees;

[0044] Based on the review and summary reports of each target department of the target enterprise, a review and summary report of the target enterprise is obtained.

[0045] In a second aspect, the present invention provides a device for recovering enterprise work data, the device comprising:

[0046] An acquisition unit is configured to acquire a review report generated by different target employees of the target enterprise regarding their work performance during a target time period, as well as historical review analysis data of each target employee; wherein the historical review analysis data is obtained by reviewing and analyzing the historical review reports of the target employees during the historical time period;

[0047] An analysis unit is configured to, for a review report of any target employee, identify problems in the review report according to configured problem identification rules to obtain a problem identification result; perform sentiment analysis on the review report to obtain a sentiment analysis result; perform quality evaluation on the review report according to configured quality evaluation rules to obtain a quality evaluation result; and generate review analysis data of the target employee within the target time period based on the sentiment analysis result, the problem identification result, and the quality evaluation result;

[0048] a comparison unit, configured to compare the historical review analysis data of the target employee with the review analysis data to generate change trend analysis data;

[0049] A determination unit is used to determine the target problems of the target enterprise within the target time period based on the problem identification results of different target employees;

[0050] A generating unit is used to generate a review summary report of the target enterprise within the target time period based on the target problem, the review analysis data of different target employees and the change trend analysis data of the corresponding target employees.

[0051] In a third aspect, the present invention provides an electronic device, comprising a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0052] Memory for storing computer programs;

[0053] The processor is configured to implement any of the methods described in the foregoing embodiments when executing the program stored in the memory.

[0054] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in any one of the aforementioned embodiments is implemented.

[0055] This application obtains the review reports of target employees of target enterprises, automatically identifies the emotional categories and problems in the work process in the review reports, generates action suggestions and conducts quality evaluation of the review reports, which helps enterprise managers to intuitively understand the work situation of each employee; by introducing artificial intelligence technology, it improves the efficiency and quality of review, actively identifies potential problems, provides intelligent suggestions and warnings, and assists managers and employees in effective target management.

[0056] This application compares data from different review cycles (i.e., different target time periods) (such as changes in the number of risks, evolution of major obstacles, and distribution of achievement types) to analyze the evolution trend of the organization's goal management capabilities.

[0057] This application uses preset rules to quickly locate problems in the review, reduce manual screening costs, and improve the accuracy and efficiency of problem discovery; combines sentiment analysis, quality evaluation and historical data comparison to build a comprehensive employee work performance evaluation model to avoid the limitations of single-dimensional judgment; by comparing historical data with current data, it intuitively presents the changing trends of employee work performance, helping companies identify potential problems or optimization directions; based on the problem identification results of all employees, it extracts common problems at the enterprise level (such as process loopholes, training needs, etc.) to provide data support for management decision-making; automatically generates structured review summary reports with a unified evaluation caliber to facilitate management to quickly grasp the overall situation and promote the scientific nature and consistency of management decisions; through quantitative analysis of employees' work status and corporate problems, it assists companies in formulating targeted improvement measures (such as training plans, process optimization) to improve overall operational efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0059] Figure 1 A flowchart of a method for reviewing enterprise work data provided in an embodiment of the present application;

[0060] Figure 2 A schematic diagram of the structure of a device for restoring enterprise work data provided in an embodiment of the present application;

[0061] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0062] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0063] The method for replaying enterprise work data provided in the embodiment of the present application can be applied in a server or in a terminal with strong computing power. The server can be a physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and basic cloud computing services such as big data and artificial intelligence platforms. The terminal can be a user equipment (UE) such as a mobile phone, a smart phone, a laptop, a digital broadcast receiver, a personal digital assistant (PDA), a tablet computer (PAD), a handheld device, a vehicle-mounted device, a wearable device, a computing device or other processing equipment connected to a wireless modem, a mobile station (MS), a mobile terminal, etc. The terminal and the server can be directly or indirectly connected by wired or wireless communication, and this application is not limited here.

[0064] The preferred embodiments of the present application are described below in conjunction with the drawings in the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application and are not used to limit the present application. In addition, the embodiments and features in the embodiments of the present application can be combined with each other if there is no conflict.

[0065] Figure 1 This is a flow chart of a method for replaying enterprise work data provided in an embodiment of the present application. Figure 1 As shown, the method may include:

[0066] Step S110: Obtain the review reports generated by different target employees of the target enterprise regarding their work status within the target time period, as well as the historical review analysis data of each target employee; for the review report of any target employee, identify problems in the review report according to the configured problem identification rules to obtain problem identification results.

[0067] In an embodiment of the present application, the review report is obtained by the target employee filling out the review report template pre-configured by the target enterprise; the review report template includes: risk obstacles, review summary (or results summary), expected completion targets, follow-up actions and other matters, among which risk obstacles are the obstacles or difficulties encountered by the target employee when working within the target time period; the review summary is the work completion status of the target employee within the target time period; the follow-up actions are the measures taken by the user to solve the corresponding obstacles or difficulties, or the work plan for the next target time period; other matters are used for the target employee to fill in other data, such as resources that need to be uniformly allocated by the company; the expected completion target is the workload or progress that the target employee is expected to complete within the target time period.

[0068] In the embodiment of the present application, the historical review analysis data is obtained by reviewing and analyzing the historical review reports of the target employees in the historical time period.

[0069] In an embodiment of the present application, the problem identification rules include: different keywords, different problem words, and the problem level, first threshold and second threshold corresponding to each problem word; among them, different keywords are used to identify keywords or synonyms related to difficulties or setbacks, such as difficulty, delay and unsmoothness; the configuration of keywords is to extract key sentences related to problem descriptions from the review report; problem words are words that specifically describe the problem, such as poor communication and lack of resources. Problem words not only include keywords but also specific objects or problem types.

[0070] In the embodiment of the present application, a keyword library and a question word library are pre-configured; at the same time, a weight level (high / medium / low) is set for each keyword and / or question word.

[0071] In the embodiment of the present application, according to the configured problem identification rules, the problem identification is performed on the replay report to obtain the problem identification results, including:

[0072] For any keyword, extract the problem description containing the keyword from the review report; analyze the problem description to obtain the target problem word; count the target problem words contained in the review report and the number of occurrences of the corresponding target problem words; if the number of occurrences of the target problem word is greater than the first threshold corresponding to the corresponding problem word, the problem word is determined as the first target word; generate the problem identification result based on the first target word and the corresponding number of occurrences, problem level and problem description, as well as the work completion status.

[0073] In one embodiment of the present application, extracting a problem description containing keywords from a review report includes: dividing the review report into sentences S i , the sentence vector of the corresponding sentence is expressed as i is a positive integer; keyword K j The template vector is Calculate the similarity between each sentence and the template vector of each keyword:

[0074]

[0075] If any sim(S i ,K j )>τ, the corresponding sentence is considered to be a problem description containing keywords.

[0076] In another embodiment of the present application, in order to improve data processing efficiency, the data filled in by the target employees can also be extracted from the area corresponding to the risk obstacles in the review report, and the semantic understanding and analysis of this part of the data can be targeted to obtain the target problem words.

[0077] In an embodiment of the present application, keyword matching can adopt Aho-Corasick multi-mode matching or a matching method based on semantic enhancement. After the question description sentence containing keywords is segmented, similarity matching is performed with the configured keyword template vector or question word template vector. If the similarity matching result is greater than the configured similarity threshold, it is considered to be the corresponding target question word.

[0078] In the embodiment of the present application, a problem identification result is generated based on the first target word and its corresponding number of occurrences, the problem level and the problem description, as well as the work completion status, including:

[0079] A problem summary is generated based on the first target word and the corresponding number of occurrences, problem level and problem description, as well as the work completion status; if the first target word and the corresponding number of occurrences, problem level and problem description, as well as the work completion status meet the risk warning rules, risk warning information is generated; based on the problem summary and risk warning information, a problem identification report is obtained.

[0080] In an embodiment of the present application, if the first target word and the corresponding number of occurrences, problem level and problem description, as well as the work completion status meet the risk warning rules, including: when the risk level and number of occurrences of the target problem word or the first target word meet the configured risk warning rules, then a risk warning information is generated and sent to HR or relevant managers; the risk warning rules include: problem level, third threshold and target problem word; for example, the target problem word is a synonym for progress lag or delay, the problem level is primary, and the third threshold is 2 times; then when progress lag and other target problem words appear in the review report, it is determined whether the work completion status is consistent with the expected completion target. If it lags behind the expected completion target (indicating actual progress delay), a warning information is generated; or, when a problem word with an intermediate problem level appears more than 2 times in the review report, a warning information is generated.

[0081] Step S120: Perform sentiment analysis on the review report to obtain sentiment analysis results; perform quality evaluation on the review report according to the configured quality evaluation rules to obtain quality evaluation results; generate review analysis data for the target employee within the target time period based on the sentiment analysis results, problem identification results, and quality evaluation results.

[0082] In the embodiment of the present application, sentiment analysis is performed on the review report to obtain sentiment analysis results, including:

[0083] The review report is segmented to obtain multiple text paragraphs and the position of each text paragraph in the review report; semantic understanding is performed on multiple text paragraphs respectively to obtain the semantic understanding results corresponding to each text paragraph; different text paragraphs and the corresponding semantic understanding results and positions are input into a pre-trained sentiment analysis model to obtain the sentiment analysis results of the review report.

[0084] In an embodiment of the present application, the sentiment analysis results include: sentiment tendencies and corresponding degrees, sentiment categories, and reasons for outputting corresponding results; sentiment tendencies include: positive (or active), negative (or passive) and neutral, and the corresponding degrees are mild, moderate and severe; different sentiment tendencies correspond to different sentiment categories; sentiment categories include satisfaction, confidence, fatigue, happiness and sadness, etc.; the reason for outputting the corresponding result is the reason for judging that the user has this sentiment tendency, for example, the user contains multiple negative sentiment words, and the corresponding negative sentiment words are given.

[0085] In this embodiment of the application, different emotional tendencies and emotional categories correspond to different emotional prompt words. After obtaining the emotional analysis results of the target employee, the corresponding emotional prompt words are matched according to different emotional tendencies and emotional categories and sent to the user; for example, your expression is emotional, do you need to adjust it? Or you expressed your ideas confidently, which is great.

[0086] In another embodiment of the present application, a warning category is pre-configured. When the identified emotion category is a warning category, an emotional warning message will be generated and sent to HR or managers to remind an employee that he or she has serious negative emotions and needs timely attention and intervention.

[0087] In the embodiment of the present application, the sentiment analysis model not only determines the overall sentiment tendency or sentiment category of a certain paragraph of text, but also obtains comprehensive and accurate sentiment recognition results based on tone, wording, and subjective words.

[0088] In an embodiment of the present application, semantic understanding is performed on multiple text paragraphs separately, including: understanding the language structure and the emotional intensity of the words, such as the emotional weights of words such as "failure", "frustration", and "satisfaction"; the sentiment analysis model can choose the RoBERTa or ERNIE model.

[0089] In actual applications, target employees complete the review report in a certain order, and the corresponding emotions will also change with the different parts filled in. Therefore, this application also takes the position of different paragraphs in the review report into consideration to determine the overall emotion recognition results.

[0090] In an embodiment of the present application, the sentiment analysis model encodes different text paragraphs into sentence vectors, uses a classification head to output sentiment tendencies and sentiment categories, and determines the overall sentiment tendencies and corresponding degrees and sentiment categories based on the position vectors and the corresponding semantic understanding results and positions.

[0091] In the embodiment of the present application, the output function of the sentiment analysis model is as follows:

[0092]

[0093] Where P(y=c|B) represents the predicted probability that the text paragraph B is classified as the emotion category c; h B represents the sentence vector of the text paragraph T; c represents the number of emotional tendencies and emotional categories; y represents the true emotional label; w c b represents the weight vector of emotion category c in the softmax classifier; c represents the bias term of emotion category c in the softmax classifier; w x Represents the weight vector of emotion category x in the softmax classifier; b xrepresents the bias term of the emotion category x in the softmax classifier; e represents the base of the natural logarithm; T represents the vector transpose.

[0094] In an embodiment of the present application, HR or managers can also conduct targeted screening based on emotional categories and emotional tendencies to obtain employees with corresponding emotional categories and tendencies. They can also conduct emotional analysis based on departments to determine the proportion of employees with different emotional tendencies in each department, so as to analyze whether the department needs to hold cohesive activities or adjust emotions.

[0095] In an embodiment of the present application, the quality evaluation rules include different evaluation indicators and scoring rules and weights corresponding to each evaluation indicator; the evaluation indicators are customized by the target enterprise, and the evaluation indicators involve multiple aspects such as completeness, specificity, executability, rationality and word coverage. Completeness is used to reflect whether the review report contains all key fields and whether there are fields left blank; specificity is used to characterize whether the review report mentions specific numbers (such as "conversion rate reached 85%") and behavioral actions (such as "plan to negotiate with Department A"); executability is used to characterize whether the next action has a clear action object, execution method, and time point; rationality is used to characterize whether the emotional bias of the review report is overly negative and whether there is any content without feedback; there is a word count requirement for the review report. When the review report exceeds the word count requirement, it can be recommended as an excellent review report.

[0096] In one embodiment of the present application, the quality evaluation rules adopt a heuristic scoring formula, which is weighted and superimposed by dimensions, with a completeness weight of 40%, a specificity and executability weight of 30%, a rationality weight of 15%, and a word coverage weight of 15%.

[0097] In the embodiment of the present application, the quality evaluation of the replay report is performed according to the configured quality evaluation rules to obtain the quality evaluation results, including:

[0098] For any evaluation indicator, the review report is scored based on the scoring rules corresponding to the evaluation indicator to obtain the score of the review report under the evaluation indicator; the scores of the review report under different evaluation indicators are weighted and summed to obtain the quality evaluation result of the review report.

[0099] In an embodiment of the present application, the score of the review report in the completeness evaluation index is determined by analyzing whether the corresponding three parts in the review report template all have text and whether the number of words meets the configured word count threshold; the score of the review report in the specificity and executability evaluation index is determined by analyzing whether there are specific numbers in the work summary or the next step plan, or whether there are sentences containing verbs plus objects plus time points, and the proportion of the corresponding sentences in all sentences in this part of the paragraph; the score of the review report in the rationality evaluation index is determined by analyzing the proportion of neutral emotional tendencies in the emotional tendencies of each paragraph of the overall review report and the overall emotional tendency of the review report; the score of the review report in the word coverage evaluation index is obtained according to the number of words in the review report and the configured scoring rules.

[0100] In an embodiment of the present application, the quality evaluation result of the review report is a specific score; the quality evaluation of the review report is performed according to the configured quality evaluation rules, with the aim of guiding the writing of higher quality review reports and at the same time monitoring the overall review quality level for the organization.

[0101] In one embodiment of the present application, the formula for the quality evaluation result may be:

[0102] Score total =0.4·C1+0.3·C2+0.15·C3+0.15·C4

[0103]

[0104] C2 = percentage of action sentences containing verb + object + time point;

[0105] C3 = 1-|emotional score|, that is, the more neutral the score, the higher the score, and extreme emotions will result in deduction of points.

[0106] Among them, C1 represents the score of the review report in the integrity evaluation index. C2 represents the score of the review report in the evaluation index of specificity and executability, C2 = the proportion of action sentences containing verbs + objects + time points; C3 represents the score of the review report in the evaluation index of rationality, C3 = 1-|emotion score|, that is, the more neutral the score, the higher the score, and extreme emotions will result in a deduction; C4 represents the score of the review report in terms of word coverage,

[0107] Step S130: Compare the historical review analysis data and the review analysis data of the target employee to generate change trend analysis data; based on the problem identification results of different target employees, determine the target problems of the target enterprise within the target time period.

[0108] In the embodiment of the present application, the historical review analysis data and the review analysis data of the target employee are compared to generate the change trend analysis data, including:

[0109] Compare historical sentiment analysis results with sentiment analysis results in historical review analysis data to generate sentiment change analysis data; compare historical problem identification results with problem identification results in historical review analysis data to generate problem change analysis data; compare historical quality evaluation results with quality evaluation results in historical review analysis data to generate quality change analysis data; based on sentiment change analysis data, problem change analysis data and quality change analysis data, obtain change trend analysis data.

[0110] In an embodiment of the present application, multiple periods of review data (multiple target time periods, such as Q1, Q2, and Q3) are obtained, and the target word frequency, emotional label ratio, review quality score, and action item coverage in each period are extracted; a trend line is generated with the time axis as the horizontal axis (each period) and the indicator as the vertical axis (such as "average number of risk keywords", "negative emotional ratio", and "average quality score"); the increase / decrease in two consecutive time periods is compared, the rate of change is analyzed, and a prompt is generated such as "the frequency of 'collaboration problems' decreased by 15% in this period, but the 'resource bottleneck' problem increased by 20%"; the trends of different departments, different target types, and different employee groups (such as new employees and managers) can be compared, and the output can be in the form of a line graph: a periodic trend line for each type of risk theme, a stacked graph: a comparison of the total amount of different problem types in different periods, and an emotional heat map: an emotional color change trend map for the entire organization / departments.

[0111] In the embodiment of the present application, based on the problem identification results of different target employees, the target problem of the target enterprise within the target time period is determined, including:

[0112] The total number of occurrences of different first target words in the problem identification results of all target employees is counted; if the total number of occurrences of any first target word is greater than the second threshold, the first target word is used as the second target word; text clustering is performed on different second target words to obtain target problem clusters; based on the problem level corresponding to each second target word contained in different problem clusters and the work completion status, the risk level corresponding to each problem cluster is determined; according to the proportion of each problem cluster in all problem clusters, a problem cluster distribution map is generated; based on the problem cluster distribution map, the risk level corresponding to each problem cluster and the second target words contained in each problem cluster, the target problems of the target enterprise within the target time period are obtained.

[0113] In an embodiment of the present application, a text clustering algorithm (such as LDA topic model, KMeans, etc.) is used to perform text clustering on different second target words to obtain a target problem cluster; for example, the "collaboration problem" topic includes: "docking delay", "unclear interface", and "cross-team scheduling" in order to identify common problems.

[0114] In one embodiment of the present application, the problem identification results of different target employees are input into topic modeling to determine the problem distribution, word distribution, and sampling of topics for each paragraph from the topic distribution, and then sampling of words to obtain the topic category to which each review belongs and the target problem of the target enterprise within the target time period, and the target problem of the target enterprise within the target time period is displayed in the form of an obstacle topic distribution map.

[0115] Step S140: Generate a review summary report for the target enterprise within the target time period based on the target problem, the review analysis data of different target employees, and the change trend analysis data of the corresponding target employees.

[0116] In the embodiment of the present application, a review summary report of the target enterprise within the target time period is generated based on the target question, the review analysis data of different target employees, and the change trend analysis data of the corresponding target employees, including:

[0117] Based on the quality evaluation results of the review reports of the target employees in different target departments of the target enterprise, the review completion degree of each target department is calculated; based on the target issues, review completion degree, review analysis data of each target employee in the target department and the change trend analysis data of the corresponding target employees, a review summary report of the target department within the target time period is generated; based on the review summary reports of each target department of the target enterprise, the review summary report of the target enterprise is obtained.

[0118] In one embodiment of the present application, trend analysis data such as the degree of completion of the review can be calculated using the following formula:

[0119]

[0120] Where t represents each target time period; M t It represents the average score of each target employee of the target enterprise on this indicator.

[0121] In an embodiment of the present application, the number of review reports with quality evaluation results greater than or equal to 80 points is divided by the total number of review reports to calculate the proportion of high-quality review reports; the number of review reports with quality evaluation results less than or equal to 60 points is divided by the total number of review reports to calculate the proportion of low-quality review reports; the completion rate of review reports of a department or a certain level is calculated based on the number of submitters / the number of people who should submit; based on the quality evaluation results of review reports of a department or a certain level, the quality average of the department is calculated, and abnormal employees are identified, for example, if the review score is <50 for three consecutive times; if the department completion rate is lower than the set threshold (such as 80%), the system automatically reminds its supervisor and HRBP; the review summary of the department or enterprise is displayed with a variety of visual charts; for example, radar chart: dimensions include completion rate, quality average, emotional health, and risk exposure rate; list view: review scores and status details of all employees can be exported; organization-wide comparison dashboard: sorted by department / role / level.

[0122] In one embodiment of the present application, the method further includes: displaying the on-time completion rate of the review for each department / level in the form of a chart, combining it with the quality evaluation results to monitor the overall review execution; ensuring the anonymization of employee personal information and protecting privacy when aggregating and analyzing data; designing intuitive and easy-to-understand charts to support managers in quickly obtaining key information. Establishing a clear review process to ensure that the content marked as "excellent practices" is representative and practical;

[0123] In the embodiment of the present application, after generating the problem identification result, the method further includes:

[0124] From the configured mapping table of different target problem types, different target problems, and different solutions, match the target problems and target solutions corresponding to the target problem types.

[0125] In an embodiment of the present application, by converting the target problem type and the target problem into a semantic vector, the similarity between the semantic vector and each vector in the mapping table of different target problem types, different target problems and different solutions is calculated; if the similarity is greater than the configured similarity threshold, the corresponding solution is output; if the similarity is not greater than the configured similarity threshold, no suggestion is returned (to avoid mismatching).

[0126] In the embodiment of the present application, the vector models available are: S-BERT, SimCSE, iTalent-fine-tuning and re-analysis semantic model.

[0127] In one embodiment of the present application, for review reports with a quality evaluation result exceeding 90 points and containing a text description associated with the next action with more than 30 words in the review report, or review reports marked as excellent by department leaders or managers, their "action plans" and "experience summaries" can be extracted for display: action sentences containing actions, objects and time are extracted as action plans; content similar to action plans are aggregated through semantic clustering to form thematic practice entries; HR is supported to add classification labels for each practice (such as "cross-team collaboration" and "performance coaching"); keywords can be searched or dimensions such as departments, cycles, and target types can be filtered; through the practice dashboard list page: all marked practices are displayed with labels, source people, and review cycles; practices can be previewed / referenced to other goals for employees' reference and adoption.

[0128] For example, when filling in the "follow-up actions" field, an employee has not yet filled in the content, or has entered: "There is no clear plan yet, and I hope to optimize the collaboration method in the next cycle"; when it is detected that this field is empty or lacks specific execution intentions, the content in the employee's "risk description", "summary" and other fields is read; the problem category faced (such as "communication problems" and "demand changes") is extracted; when querying historical reviews of successful practices recorded by employees in dealing with such problems, or matching relevant operational suggestions from the knowledge base; from these, several typical "action plans" that are most similar to the current problem scenario are extracted as recommended content; employees can "adopt" with one click, and the suggested content is automatically filled in the form, which can be modified and supplemented before submission; if the employee already has some draft content, the system will suggest supplementing it without overwriting the original content.

[0129] For example, Product Manager Xiao Li (responsible for personal goals) is reviewing his personal OKRs for this quarter. His goal is to "deliver a new user onboarding module." He faces challenges with inefficient collaboration and delayed launch, but isn't sure how to clearly articulate these challenges. Xiao Li opens the review form, and the system automatically populates the goal details. When filling in the "Risks and Obstacles" field, Xiao Li enters, "Due to limited design team resources, our proposal was rejected twice, and progress is one week behind schedule." The system automatically identifies the sentiment as "neutral to slightly negative" and highlights keywords like "limited resources" and "schedule delay." The system prompts, "Xiao Li, you may be facing resource coordination issues," and displays smart suggestion cards: Suggestion: Establish a weekly requirements meeting with the design team manager (Source: Last Quarter's Excellent Review by the UED Team); Suggestion: Enable component library reuse to reduce custom pages (Source: PM Handbook). Xiao Li "accepts" the suggestion and fills it into the "Next Action" field. Xiao Li clicks "Sync as Task," and the action is automatically created as a Jira To-Do, assigned to him, with a due date set to the start of the next quarter.

[0130] Product Director Wang reviews the quality of his department's subordinates' reviews and risk distribution every month to assess the organization's operating status; entering the "Department Goal Review Dashboard": it shows that 7 out of 8 people have submitted reviews; the system marks two of the reviews as a "negative emotion + high risk" combination (resource bottleneck + "cross-departmental dependency" respectively); Manager Wang clicks on one of the reviews and finds that the employee mentioned "R&D priority conflict" for two consecutive cycles; in the dashboard's "obstacle word cloud", "passive scheduling" and "dependency delay" appear frequently, and most of them are concentrated in the content growth group; Manager Wang holds an ad hoc synchronous meeting, using the risk details in the system-exported report as meeting materials; at the same time, the employee's "action plan" is generated as a Jira subtask and transferred to the R&D manager for resource scheduling coordination.

[0131] HRBP Xiao Zhao needs to generate an OKR execution report every quarter to evaluate each department's "goal completion status + review quality + emotional atmosphere + action closure rate"; enter the "Organizational Insight Dashboard" to view the following content: each department's review submission rate (target>90%, actual average 83%) → start a reminder to submit; the quality score trend chart shows: the technical group has an average score of 92, and the operation group is only 61; word cloud analysis: "Insufficient budget" and "short-term temporary goals" appear frequently in the sales department's review; Xiao Zhao marks review items with a score higher than 90 and specific actions as "excellent practices": including the article "Pilot of Cross-Departmental Goal Co-construction Mechanism", which is synchronized to the knowledge base PM strategy area; the system recommended 3 action suggestions based on these excellent practices, which were directly adopted by 6 employees in the review; Xiao Zhao packaged the "review completion rate + average score + action task achievement status" into a report for HR supervisors and business heads to report on the quarterly cultural atmosphere improvement effects.

[0132] Tony, the platform administrator, is responsible for the compliance operation of the platform. He must ensure that the access rights to the review data are reasonable and that the data output is compliant and traceable. In the permission management, it is configured that: the reviews submitted by employees are visible to "themselves + line managers" by default; HR has organizational dimension review access rights, but the "risk keyword field" must be explicitly authorized; field-level access logs are enabled: all access records containing the "negative sentiment" field are recorded and audited; export policies are set: ordinary employees can only export review summaries that do not contain risk fields; only HR supervisors have "full review export" permissions (including AI analysis and scoring fields); data access audit log reports are automatically generated every month for IT audit reference.

[0133] Corresponding to the above method, the embodiment of the present application also provides a device for replaying enterprise work data, such as Figure 2 As shown, the enterprise work data recovery device includes:

[0134] The acquisition unit 210 is configured to acquire the work review reports generated by different target employees of the target enterprise for the target period of time, as well as the historical review analysis data of each target employee; wherein the historical review analysis data is obtained by reviewing and analyzing the historical review reports of the target employees in the historical period of time;

[0135] The analysis unit 220 is configured to identify problems in the review report of any target employee according to the configured problem identification rules to obtain a problem identification result; perform sentiment analysis on the review report to obtain a sentiment analysis result; perform quality evaluation on the review report according to the configured quality evaluation rules to obtain a quality evaluation result; and generate the target employee's review analysis data within the target time period based on the sentiment analysis result, the problem identification result, and the quality evaluation result;

[0136] Comparison unit 230, used to compare the historical review analysis data and the review analysis data of the target employee to generate change trend analysis data;

[0137] A determination unit 240 is configured to determine a target problem of a target enterprise within a target time period based on the problem identification results of different target employees;

[0138] The generating unit 250 is configured to generate a review summary report for the target enterprise within a target time period based on the target problem, the review analysis data of different target employees, and the change trend analysis data of the corresponding target employees.

[0139] The functions of each functional unit of the enterprise work data reconstruction device provided in the above embodiment of the present application can be realized through the above method steps. Therefore, the specific working process and beneficial effects of each unit in the enterprise work data reconstruction device provided in the embodiment of the present application will not be repeated here.

[0140] The present application also provides an electronic device, such as Figure 3 As shown, it includes a processor 310 , a communication interface 320 , a memory 330 and a communication bus 340 , wherein the processor 310 , the communication interface 320 , and the memory 330 communicate with each other via the communication bus 340 .

[0141] Memory 330, for storing computer programs;

[0142] The processor 310 is configured to execute the program stored in the memory 330 by performing the following steps:

[0143] Obtain the work review reports generated by different target employees of the target enterprise for the target period, as well as the historical review analysis data of each target employee. The historical review analysis data is obtained by reviewing and analyzing the historical review reports of the target employees in the historical period.

[0144] For any target employee's review report, identify problems based on the configured problem identification rules and obtain the problem identification results;

[0145] Perform sentiment analysis on the review report to obtain sentiment analysis results;

[0146] According to the configured quality evaluation rules, the review report is evaluated and the quality evaluation results are obtained;

[0147] Generate review analysis data for target employees within the target time period based on sentiment analysis, problem identification, and quality evaluation results;

[0148] Compare the historical and post-analysis data of target employees to generate trend analysis data;

[0149] Based on the problem identification results of different target employees, determine the target problems of the target enterprise within the target time period;

[0150] Based on the target issues, the review and analysis data of different target employees and the change trend analysis data of the corresponding target employees, a review and summary report of the target enterprise within the target time period is generated.

[0151] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. This communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is used in the figure, but this does not mean that there is only one bus or only one type of bus.

[0152] The communication interface is used for communication between the above electronic device and other devices.

[0153] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage. Alternatively, the memory may be at least one storage device located away from the processor.

[0154] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0155] The implementation methods and beneficial effects of the various components of the electronic device in the above embodiments to solve the problems can be found in Figure 1 The various steps in the embodiment shown are implemented, therefore, the specific working process and beneficial effects of the electronic device provided by the embodiment of the present application are not repeated here.

[0156] In another embodiment provided in the present application, a computer-readable storage medium is also provided, which stores instructions. When the computer-readable storage medium is run on a computer, it enables the computer to execute the enterprise work data recovery method described in any of the above embodiments.

[0157] In another embodiment provided by the present application, a computer program product including instructions is also provided, which, when executed on a computer, enables the computer to execute the enterprise work data recovery method described in any one of the above embodiments.

[0158] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the embodiments of the present application can be implemented in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the embodiments of the present application can be implemented in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0159] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0160] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0161] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0162] Although preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they become aware of the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0163] Obviously, those skilled in the art can make various changes and modifications to the embodiments of the present application without departing from the spirit and scope of the embodiments of the present application. Thus, if these modifications and variations of the embodiments of the present application fall within the scope of the claims and their equivalents, the embodiments of the present application are also intended to include these modifications and variations.

Claims

1. A method for reviewing enterprise work data, characterized in that: The method comprises: Obtaining review reports generated by different target employees of the target enterprise regarding their work performance during the target time period, as well as historical review analysis data for each target employee; wherein the historical review analysis data is obtained by reviewing and analyzing the historical review reports of the target employees during the historical time period; For any target employee's review report, identify problems based on the configured problem identification rules and obtain problem identification results; Performing sentiment analysis on the review report to obtain a sentiment analysis result; Perform a quality evaluation on the replay report according to the configured quality evaluation rules to obtain a quality evaluation result; Generate review analysis data of the target employee within the target time period based on the sentiment analysis results, the problem identification results, and the quality evaluation results; Comparing the target employee's historical review analysis data with the review analysis data to generate change trend analysis data; Based on the problem identification results of different target employees, determine the target problems of the target enterprise within the target time period; Based on the target issues, the review analysis data of different target employees and the change trend analysis data of the corresponding target employees, a review summary report of the target enterprise within the target time period is generated.

2. The method according to claim 1, wherein The review report includes: the target employee's work completion status within the target time period; The question identification rules include: different keywords, different question words, and the question level, first threshold and second threshold corresponding to each question word; According to the configured problem identification rules, the problem identification is performed on the replay report to obtain the problem identification results, including: For any keyword, extract the problem description containing the keyword from the review report; Analyze the problem description to obtain target problem words; Counting the occurrences of each target question word and the corresponding target question word in the review report; If the number of occurrences of the target question word is greater than a first threshold corresponding to the corresponding question word, the question word is determined as the first target word; A problem identification result is generated based on the first target word and the corresponding number of occurrences, the problem level and the problem description, as well as the completion status of the work.

3. The method according to claim 1, wherein Perform sentiment analysis on the review report to obtain sentiment analysis results, including: Segmenting the review report to obtain a plurality of text paragraphs and a position of each text paragraph in the review report; Perform semantic understanding on multiple text paragraphs respectively to obtain semantic understanding results corresponding to each text paragraph; Different text paragraphs and corresponding semantic understanding results and positions are input into a pre-trained sentiment analysis model to obtain the sentiment analysis results of the review report.

4. The method according to claim 1, wherein Comparing the target employee's historical review analysis data with the review analysis data to generate change trend analysis data, including: Comparing historical sentiment analysis results in the historical replay analysis data with the sentiment analysis results to generate sentiment change analysis data; Comparing the historical problem identification results in the historical replay analysis data with the problem identification results to generate problem change analysis data; Comparing the historical quality evaluation results in the historical replay analysis data with the quality evaluation results to generate quality change analysis data; Based on the sentiment change analysis data, the problem change analysis data and the quality change analysis data, change trend analysis data is obtained.

5. The method according to claim 2, wherein Based on the problem identification results of different target employees, determine the target problems of the target enterprise within the target time period, including: Count the total number of occurrences of different first target words in the problem recognition results of all target employees; If the total number of occurrences of any first target word is greater than the second threshold, the first target word is used as the second target word; Perform text clustering on different second target words to obtain target question clusters; Determining the risk level corresponding to each problem cluster based on the problem level corresponding to each second target word included in different problem clusters and the completion status of the work; Generate a problem cluster distribution map based on the proportion of each problem cluster to all problem clusters; The target problems of the target enterprise within the target time period are obtained according to the problem cluster distribution map, the risk level corresponding to each problem cluster and each second target word contained in each problem cluster.

6. The method according to claim 1, wherein The quality evaluation rules include different evaluation indicators and scoring rules and weights corresponding to each evaluation indicator; Perform a quality evaluation on the replay report according to the configured quality evaluation rules to obtain a quality evaluation result, including: For any evaluation indicator, score the review report based on the scoring rules corresponding to the evaluation indicator to obtain the score of the review report under the evaluation indicator; The scores of the review report under different evaluation indicators are weighted and summed to obtain the quality evaluation result of the review report.

7. The method according to claim 1, wherein Based on the target issues, the review analysis data of different target employees and the change trend analysis data of the corresponding target employees, a review summary report of the target enterprise within the target time period is generated, including: Calculate the completion rate of the review of each target department based on the quality evaluation results of the review reports of each target employee in different target departments of the target enterprise; Generate a review summary report for the target department within the target time period based on the target problem, the review completion rate, the review analysis data of each target employee in the target department, and the change trend analysis data of the corresponding target employees; Based on the review and summary reports of each target department of the target enterprise, a review and summary report of the target enterprise is obtained.

8. A device for replaying enterprise work data, characterized in that: The device comprises: An acquisition unit is configured to acquire a review report generated by different target employees of the target enterprise regarding their work performance during a target time period, as well as historical review analysis data of each target employee; wherein the historical review analysis data is obtained by reviewing and analyzing the historical review reports of the target employees during the historical time period; An analysis unit is configured to, for a review report of any target employee, identify problems in the review report according to configured problem identification rules to obtain a problem identification result; perform sentiment analysis on the review report to obtain a sentiment analysis result; perform quality evaluation on the review report according to configured quality evaluation rules to obtain a quality evaluation result; and generate review analysis data of the target employee within the target time period based on the sentiment analysis result, the problem identification result, and the quality evaluation result; a comparison unit, configured to compare the historical review analysis data of the target employee with the review analysis data to generate change trend analysis data; A determination unit is used to determine the target problems of the target enterprise within the target time period based on the problem identification results of different target employees; A generating unit is used to generate a review summary report of the target enterprise within the target time period based on the target problem, the review analysis data of different target employees and the change trend analysis data of the corresponding target employees.

9. An electronic device, characterized in that: The electronic device includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory for storing computer programs; A processor, configured to implement the method according to any one of claims 1 to 7 when executing a program stored in a memory.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images