Object evaluation method and device, electronic equipment and storage medium

Abstract

The invention provides an object evaluation method and device, electronic equipment and a storage medium, and relates to the technical field of information processing. The method comprises the steps of obtaining to-be-evaluated data of a to-be-evaluated object, and avoiding the problem of single evaluation index to a certain extent due to the fact that the data type of the to-be-evaluated data comprises at least one of texts, audios and images; furthermore, for any to-be-evaluated data, the analysis model corresponding to the data type of the to-be-evaluated data is selected, and the to-be-evaluated data is analyzed to obtain an analysis result, so that the problem that the evaluation result is influenced by subjective factors due to manual evaluation is avoided; and finally, obtained analysis results are fused, and a fusion result is used as an object evaluation result of the to-be-evaluated object, so that the evaluation accuracy is improved.

Description

Technical Field

[0001] The present disclosure belongs to the field of information processing technology, and specifically relates to an object evaluation method, device, electronic device and storage medium. Background Art

[0002] Performance management involves the reflection and recording of employees' workload and work efficiency, and serves as a guide and reference for the arrangement and adjustment of employees' work tasks.

[0003] At present, employee performance management usually sets quantitative indicators for a single dimension of work content, such as sales, etc. However, when evaluating a single dimension of work content, the evaluation results are easily affected by the subjective factors of the evaluator, and on the other hand, the dimensions of the employee's ability, contribution, and growth are not taken into account, resulting in inaccurate and incomplete evaluation of the object. Summary of the invention

[0004] The present disclosure provides an object evaluation method, device, electronic device and storage medium.

[0005] The first aspect of the present disclosure provides an object evaluation method, the method comprising:

[0006] Acquire data to be evaluated of the object to be evaluated, wherein the data type of the data to be evaluated includes at least one of text, audio, and image;

[0007] If the data type of the data to be evaluated is text, calling a preset large language model to perform a first operation on the data to be evaluated to obtain a positive representation, wherein the first operation is to extract semantic features of the data to be evaluated and determine the positive representation based on the semantic features;

[0008] If the data type of the data to be evaluated is an image, calling the image recognition model to perform a second operation on the data to be evaluated to obtain a target working state representation, wherein the second operation is to identify facial features of the object to be evaluated in the data to be evaluated, and determine the target working state representation based on a mapping relationship between the facial features and the preset working state;

[0009] If the data type of the data to be evaluated is audio, calling the speech recognition model to perform a third operation on the data to be evaluated to obtain a target emotion representation, wherein the third operation is to extract audio features of the data to be evaluated, and based on a mapping relationship between the audio features and a preset emotion representation, determine a target emotion representation corresponding to the audio features;

[0010] A fusion operation is performed on the positivity representation, the target work state representation, and the target emotion representation, and a fusion result is used as an object evaluation result of the object to be evaluated.

[0011] In the embodiment of the present disclosure, the performing of a fusion operation on the positivity representation, the target work state representation, and the target emotion representation, and using the fusion result as the object evaluation result of the object to be evaluated, includes:

[0012] Determining the work type of the object to be evaluated; and determining a fusion rule matching the work type;

[0013] Performing a fusion operation on the positivity representation, the target work state representation, and the target emotion representation according to the fusion rule to obtain overall evaluation information;

[0014] An object evaluation result is generated according to the overall evaluation information and the task completion amount of the object to be evaluated.

[0015] In the embodiment of the present disclosure, the fusion operation is performed on the positivity representation, the target work state representation, and the target emotion representation according to the fusion rule to obtain overall evaluation information, including:

[0016] According to the evaluation sub-rule of the positivity representation dimension in the fusion rule, generating first evaluation information based on the positivity representation;

[0017] According to the evaluation sub-rule of the target working state representation dimension in the fusion rule, second evaluation information is generated based on the target working state representation;

[0018] According to the evaluation sub-rule of the target emotion representation dimension in the fusion rule, generating third evaluation information based on the target emotion representation;

[0019] The first evaluation information, the second evaluation information and the third evaluation information are combined to obtain the overall evaluation information.

[0020] In the embodiment of the present disclosure, generating the object evaluation result according to the overall evaluation information and the task completion amount of the object to be evaluated includes:

[0021] Generate work completion parameters of the object to be evaluated based on the task completion amount of the object to be evaluated and the preset workload standard of the work type of the object to be evaluated;

[0022] An object evaluation result is generated based on the overall evaluation information and the work completion parameter.

[0023] In the embodiment of the present disclosure, the work completion parameters include the work completion type and the work completion degree, the overall evaluation information includes the evaluation type, and the object evaluation result is generated based on the overall evaluation information and the work completion parameters, including:

[0024] If the work completion type is failure to meet the preset performance standard and the evaluation type is negative attitude, generating a first object evaluation result based on the preset performance standard and the work completion degree;

[0025] If the work completion type is not meeting the preset performance standard and the evaluation type is positive attitude, generating a second object evaluation result based on the preset performance standard and the work completion degree;

[0026] If the work completion type is satisfying the preset performance standard and the evaluation type is positive attitude, a third object evaluation result is generated based on the preset performance standard and the work completion degree.

[0027] An embodiment of a second aspect of the present disclosure provides an object evaluation device, the device comprising:

[0028] An acquisition module is used to acquire the data to be evaluated of the object to be evaluated, wherein the data type of the data to be evaluated includes at least one of text, audio and image;

[0029] A text analysis module, for, if the data type of the data to be evaluated is text, calling a preset large language model to perform a first operation on the data to be evaluated to obtain a positive representation, wherein the first operation is to extract semantic features of the data to be evaluated and determine the positive representation based on the semantic features;

[0030] an image analysis module, for, if the data type of the data to be evaluated is an image, calling an image recognition model to perform a second operation on the data to be evaluated to obtain a target working state representation, wherein the second operation is to identify facial features of the object to be evaluated in the data to be evaluated, and determine the target working state representation based on a mapping relationship between the facial features and a preset working state;

[0031] An audio analysis module, for, if the data type of the data to be evaluated is audio, calling a speech recognition model to perform a third operation on the data to be evaluated to obtain a target emotion representation, wherein the third operation is to extract audio features of the data to be evaluated, and determine the target emotion representation corresponding to the audio features based on a mapping relationship between the audio features and a preset emotion representation;

[0032] The determination module is used to perform a fusion operation on the positivity representation, the target working state representation and the target emotion representation, and use the fusion result as the object evaluation result of the object to be evaluated.

[0033] In the embodiment of the present disclosure, the determining module includes:

[0034] A type rule determination unit, used to determine the work type of the object to be evaluated; and determine a fusion rule matching the work type;

[0035] A fusion unit, configured to perform a fusion operation on the positivity representation, the target work state representation, and the target emotion representation according to the fusion rule to obtain overall evaluation information;

[0036] A result generating unit is used to generate an object evaluation result according to the overall evaluation information and the task completion amount of the object to be evaluated.

[0037] In the embodiment of the present disclosure, the fusion unit includes:

[0038] A first evaluation subunit, configured to generate first evaluation information based on the positivity representation according to the evaluation subrule of the positivity representation dimension in the fusion rule;

[0039] A second evaluation subunit, configured to generate second evaluation information based on the target working state representation according to the evaluation subrule of the target working state representation dimension in the fusion rule;

[0040] A third evaluation subunit, configured to generate third evaluation information based on the target emotion representation according to the evaluation subrule of the target emotion representation dimension in the fusion rule;

[0041] The merging subunit is used to merge the first evaluation information, the second evaluation information and the third evaluation information to obtain the overall evaluation information.

[0042] An embodiment of the third aspect of the present disclosure provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described in the first aspect or any optional implementation manner of the first aspect.

[0043] An embodiment of the fourth aspect of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, wherein the program is executed by a processor to implement the method described in the first aspect and any optional implementation manner of the first aspect.

[0044] The technical solution provided in the embodiments of the present disclosure has at least the following technical effects or advantages:

[0045] The disclosed embodiment obtains data to be evaluated of an object to be evaluated. Since the data type of the data to be evaluated includes at least one of text, audio and image, the problem of single evaluation index is avoided to a certain extent. Furthermore, for any data to be evaluated, an analysis model corresponding to the data type of the data to be evaluated is selected, and the data to be evaluated is analyzed to obtain an analysis result, thereby avoiding the problem that the evaluation result is affected by subjective factors due to manual evaluation. Finally, the obtained analysis results are fused, and the fused result is used as the object evaluation result of the object to be evaluated, thereby improving the accuracy of the evaluation.

[0046] Additional aspects and advantages of the present disclosure will be given in part in the following description and in part will become apparent from the following description or be learned through practice of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] By reading the detailed description of the preferred embodiment below, various other advantages and benefits will become clear to those of ordinary skill in the art. The accompanying drawings are only used for the purpose of illustrating the preferred embodiment and are not considered to be limitations of the present disclosure. In addition, the same reference symbols are used to represent the same components throughout the accompanying drawings.

[0048] In the attached picture:

[0049] Figure 1 A schematic diagram of the structure of an object evaluation system provided by an embodiment of the present disclosure is shown;

[0050] Figure 2 A schematic diagram of an object evaluation method provided by an embodiment of the present disclosure is shown;

[0051] Figure 3 A schematic diagram of a method flow of an object evaluation method provided by an embodiment of the present disclosure is shown;

[0052] Figure 4 A schematic diagram of the structure of an object evaluation device provided by another embodiment of the present disclosure is shown;

[0053] Figure 5 A schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure is shown;

[0054] Figure 6 A schematic diagram of a storage medium provided by an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0055] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0056] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in the present disclosure should have the common meanings understood by those skilled in the art to which the present disclosure belongs.

[0057] The following introduces the implementation scenarios and related technologies involved in the embodiments of the present disclosure.

[0058] Performance management is an important part of organizational management. It measures and evaluates employees' workload, work efficiency and work results through systematic and scientific methods. In current practice, although the method of setting quantitative indicators based on a single dimension (such as sales volume, production quantity, etc.) can intuitively reflect the performance of employees on certain specific tasks to a certain extent, this evaluation model has many limitations, as follows: When relying on quantitative indicators of a single dimension for evaluation, the evaluator's personal bias, emotional state, experience level and other factors may be incorporated into the evaluation process, resulting in the evaluation results deviating from objective reality. For example, different sales managers' different understandings of "difficulty in developing potential customers" may lead to deviations in the performance evaluation of the same salesperson. For another example, focusing only on quantitative indicators of work results ignores non-quantitative factors such as employee ability development, teamwork, innovative thinking, and work attitude. These factors are equally important for the long-term development of the organization and the personal growth of employees. Lack of consideration of these dimensions may lead to the loss of excellent employees due to insufficient recognition, and may also encourage "short-sighted" behavior, such as ignoring long-term customer relationship maintenance or product quality in pursuit of short-term performance.

[0059] A single-dimensional evaluation system is difficult to stimulate employees' intrinsic motivation and innovation potential. Employees may adopt conservative strategies to achieve set indicators and avoid trying new methods or projects that may bring higher returns but also come with higher risks. In addition, for non-sales positions or creative work, such as R&D, design, customer service, etc., single-dimensional quantitative indicators are often difficult to fully reflect the value of their work. Effective performance management should include guidance and feedback on employees' personal growth and career development. Single-dimensional evaluation models often lack in-depth discussions on employee skill improvement, career path planning, etc., which limits the discovery of employees' personal potential and the organization's ability to continuously innovate.

[0060] Furthermore, relying on a single or limited data source makes it difficult to present a full picture of employee performance. Soft factors such as employee communication details, work attitude, and team interaction are ignored. The evaluation process is long, which is not conducive to employees improving their performance in a timely manner. When employees learn the evaluation results, they often have missed the best time to improve. Whether it is the subjective evaluation of the supervisor or the feedback from colleagues, it may be affected by personal bias and position bias, affecting the fairness of the evaluation. Soft skills and non-quantitative indicators are difficult to measure accurately, and traditional systems usually only focus on hard indicators.

[0061] In view of this, the embodiments of the present disclosure propose an object evaluation method, device, electronic device and storage medium. The technical solution of the present disclosure is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present disclosure will be described below in conjunction with the accompanying drawings.

[0062] See also Figure 1 , Figure 1 An exemplary system structure diagram of an object evaluation system provided by the present disclosure is shown. The object evaluation system provided by the present disclosure includes a server 10 and a client 20. The server 10 and the client 20 communicate with each other through a network.

[0063] It is to be understood that the structure shown in the present disclosure does not constitute a specific limitation on the object evaluation system. In other embodiments of the present disclosure, the object evaluation system may include more or less devices than those shown in the figure, which is not limited here.

[0064] The server 10 can be an independent physical server, or a server cluster, a distributed system or a cloud platform composed of multiple physical servers. The server 10 can deploy resources and computing power for object evaluation. For example, the server 10 can be used to store analysis models. The server 10 performs the process of analyzing the data to be evaluated in the object evaluation method of the present disclosure by calling the analysis corresponding to the data type to be evaluated. Computing power resources may include, for example, deep semantic models, image recognition models, and audio analysis models for data analysis in the process of implementing the technical solution of the present disclosure.

[0065] The server 10 can be used to receive the data to be evaluated collected by the client 20, and then call the preset analysis model to analyze the analysis results of the data to be evaluated according to the type of the data to be evaluated. The server 10 can also be used to fuse the obtained analysis results according to the fusion rules of the analysis results and the task matching of the corresponding object to be evaluated, and use the fusion results as the object evaluation results of the object to be evaluated.

[0066] The client 20 can be an electronic device such as a computer or a work phone used by an employee during work. The client 20 obtains data corresponding to the employee's work, and the data is used as the data to be evaluated for the object evaluation of the employee. Exemplarily, the client 20 can be installed to run data collection tools, provide application program interfaces, AI technologies, etc. to obtain data to be evaluated, obtain information on the client 20 and store it in a structured manner. This information can be a variety of data types such as text, audio, and images. Specifically, the text can be various types of text such as work emails corresponding to the object to be evaluated, communication texts with colleagues, work documents, etc.; the audio can be the audio of the object to be evaluated during a phone call or meeting; the image can be a facial image of the object to be evaluated captured during a video conference or during a break in work.

[0067] Understandably, Figure 1The server 10 and the client 20 are both schematic diagrams at the logical function level. In actual implementation, the client 20 may include at least one electronic device entity, and the server 10 may include at least one server device entity. This is not limited here.

[0068] like Figure 2 As shown, it is a structural block diagram corresponding to the object evaluation method corresponding to the embodiment of the present disclosure. After obtaining the data to be evaluated, the data type in the data to be evaluated is analyzed, and the analysis models used for the data to be evaluated of different data types are different. Exemplarily, if the data to be evaluated is of text type, the text analysis model is called for analysis, and specifically, the text analysis model can be natural language processing (Natural Language Processing, NLP), deep semantic understanding, large language model (Large Language Models, LLMs), etc.; if the data to be evaluated is of image type, the image analysis model is called for image analysis, and specifically, the image analysis model can be image recognition technology, image segmentation technology, target recognition and other technologies; if the object to be evaluated is audio data, the audio analysis model is called for audio analysis, and specifically, the audio analysis can use spectrum analysis, waveform analysis, etc., and machine learning algorithms, deep learning algorithms, etc. can also be used to further identify the audio for sentiment analysis, etc.

[0069] After obtaining the analysis result according to at least one data to be evaluated, multimodal fusion is performed on each analysis result. The analysis results obtained from different types of evaluation data represent different information. It is necessary to fuse each analysis result to obtain evaluation information, and then generate the corresponding object evaluation result with the task completion amount of the object to be evaluated. The object evaluation result can be fed back to the client 20. In addition, on this basis, the embodiment of the present disclosure can also obtain feedback from the user of the object to be evaluated on the object evaluation result to further adjust each analysis process or multimodal fusion process to achieve the effect of adaptive learning of the above-mentioned object evaluation result.

[0070] It should be understood that the algorithms and networks included in the above-mentioned analysis modules are illustrative examples of the present disclosure and do not constitute a limitation on the modules in the embodiments of the present disclosure. The algorithms included in the modules in the embodiments of the present disclosure can be flexibly deployed by those skilled in the art according to actual conditions. No limitation is given here.

[0071] Combine the following Figure 1 and Figure 2 An illustrative embodiment is used to describe an object evaluation method proposed according to an embodiment of the present disclosure.

[0072] like Figure 3 As shown, the embodiment of the present disclosure provides an object evaluation method, which can be applied to a server, and the server can be Figure 1 The method comprises:

[0073] In step S11, the data to be evaluated of the object to be evaluated is obtained.

[0074] The data to be evaluated includes at least one of text, audio and image of the object to be evaluated.

[0075] For example, obtaining the data to be evaluated may receive data from Figure 1 The data to be evaluated sent by the client is shown. Or if the data in the client is stored in the cloud to the server, the corresponding data to be evaluated can be directly obtained from the server.

[0076] For example, the acquisition of text data in the client can automatically capture text data from specific sources such as documents, emails, databases, etc., and can also obtain text data such as social media posts and comments by calling external application programming interfaces (Application Programming Interface, API). Acquisition of audio data: Use built-in or external microphone devices to record audio; the subject to be evaluated uploads existing audio files; capture audio data from real-time streaming media (such as video conferences, telephone conferences). Acquisition of image data: You can use camera devices to take photos or video frames, specifically images or video frames captured by the client's local camera in a video conference; allow users to upload existing image files. Capture image data from sources such as web pages and social media.

[0077] It should be noted that after obtaining the data to be evaluated, it is necessary to preprocess the data to be evaluated. Specifically, text preprocessing requires removing irrelevant characters, HyperText Markup Language (HTML) tags, etc. in the text, further segmenting the text into words or phrases, and converting the text to a unified format, such as lowercase processing. Audio preprocessing first uses a filter to remove background noise and converts the audio to a unified format and sampling rate. Image preprocessing uses image processing technology to remove noise in the image and adjust the image to a uniform size and ratio.

[0078] When analyzing the data to be evaluated, different analysis models are called for analysis according to different data types. The types of the data to be evaluated may include text and audio, audio and image, text and image, or only one of text, audio or image. The type of the data to be evaluated is determined according to actual conditions.

[0079] On this basis, after obtaining the data to be evaluated, Figure 2The identification method corresponding to the data type to be evaluated in the analysis shown can be specifically implemented by file extension, data format or content analysis. For example: Text data: usually stored with file extensions such as .txt, .docx, .pdf, etc., and the content is mainly human-readable characters. Audio data: usually stored with file extensions such as .mp3, .wav, .flac, etc., and the content contains audio waveform data. Image data: usually stored with file extensions such as .jpg, .png, .bmp, etc., and the content contains pixel data.

[0080] Next, the corresponding analysis model is called according to the data type to analyze the data to be evaluated.

[0081] In step S12, if the data type of the data to be evaluated is text, a preset large language model is called to perform a first operation on the data to be evaluated to obtain a positive representation, wherein the first operation is to extract semantic features of the data to be evaluated and determine the positive representation based on the semantic features.

[0082] For example, text analysis models can include sentiment analysis, topic extraction, keyword extraction, named entity recognition, and other models. The text comes from emails, reports, meeting minutes, etc. Select a suitable pre-trained large language model according to the task requirements. Load the model into the computing environment and prepare for reasoning.

[0083] Input the preprocessed text into the model. The model outputs an embedded representation (i.e., semantic features) of the text, which captures the contextual information and semantic relationships in the text. Select features related to positivity evaluation from the embedded representation as needed. This can include embedding vectors of specific words, average embedding vectors of sentences, etc. Based on the extracted semantic features, train a simple classifier (such as logistic regression, support vector machine, etc.), or use the model's built-in classification head (if the model supports multi-task learning). Use a labeled dataset (containing positive and negative texts and their corresponding labels) to train the classifier and tune it. Input the text to be evaluated into the trained classifier to obtain its positivity representation.

[0084] In step S13, if the data type of the data to be evaluated is an image, the image recognition model is called to perform a second operation on the data to be evaluated to obtain a target working state representation, the second operation is to identify facial features of the object to be evaluated in the data to be evaluated, and determine the target working state representation based on a mapping relationship between the facial features and the preset working state;

[0085] Exemplarily, the image analysis model may include models such as image classification, object detection, face recognition, and image style conversion. Use facial detection algorithms (such as Haar cascade classifiers, HOG+SVM, MTCNN, etc.) to locate faces in images. Extract facial features. This can be achieved through deep learning models (such as dlib's 68-point model, FaceNet, etc.), which can identify and extract key points and features of the face. Analyze the extracted facial features, such as the degree of opening and closing of the eyes, the position of the eyebrows, the degree of upward movement of the corners of the mouth, etc., which are related to the individual's emotions and concentration. Establish a mapping relationship to associate facial features with preset working states. This usually requires a training process, by collecting facial feature data at different concentrations, and using machine learning algorithms (such as support vector machines, random forests, neural networks, etc.) to learn the relationship between these features and concentration; based on the mapping relationship, match the extracted facial features with the preset working state to determine the target working state. This involves inputting the facial features extracted in real time into the trained model, and the model will output the corresponding target working state representation.

[0086] In step S14, if the data type of the data to be evaluated is audio, the speech recognition model is called to perform a third operation on the data to be evaluated to obtain a target emotion representation, where the third operation is to extract audio features of the data to be evaluated, and determine the target emotion representation corresponding to the audio features based on a mapping relationship between the audio features and the preset emotion representation;

[0087] Exemplarily, the audio analysis model may include models such as speech recognition, audio classification, audio segmentation, and audio enhancement. Use a speech recognition model (such as LSTM, GRU, and other neural network models) or a specialized audio feature extraction tool (such as Librosa) to extract the acoustic features of the audio. Common acoustic features include Mel-frequency cepstral coefficients (MFCC), linear predictive coding (LPC), spectral heart quality, etc. According to the task requirements, select acoustic features related to the target emotion recognition. For example, MFCC features perform well in speech recognition and emotion recognition because they can capture the perceptual characteristics of the human auditory system to sound. Create an audio dataset containing multiple emotion tags and label each audio file with emotions. This can be done by manual labeling or using an existing emotion recognition dataset. Use the labeled dataset to train an emotion recognition model. This model can be a model based on machine learning (such as SVM, random forest, etc.) or deep learning (such as neural network). During the training process, the model will learn the mapping relationship between audio features and emotion tags. Input the acoustic features of the audio to be evaluated into the trained emotion recognition model. The model will output one or more emotion tags based on the input audio features, indicating the target emotion expressed in the audio. The emotion tags output by the model are post-processed, such as deduplication and sorting, to obtain more accurate emotion recognition results. The performance of the emotion recognition model is evaluated using the test set, such as calculating accuracy, recall, F1 score and other indicators. Based on the evaluation results, the model can be tuned or improved.

[0088] In step S15, a fusion operation is performed on the positivity representation, the target work state representation, and the target emotion representation, and the fusion result is used as the object evaluation result of the object to be evaluated.

[0089] For example, the representation of positivity, the representation of target work status, and the representation of target emotion can be integrated to form a comprehensive object evaluation data set. The comprehensive performance score, the scores of various dimensions and indicators, the analysis of advantages and disadvantages, etc. are presented in the form of a report as the object evaluation result.

[0090] In an optional implementation, the work type of the object to be evaluated is determined; and a fusion rule for work type matching is determined; a fusion operation is performed on the positivity representation, the target work state representation, and the target emotion representation according to the fusion rule to obtain overall evaluation information; and an object evaluation result is generated according to the overall evaluation information and the task completion amount of the object to be evaluated.

[0091] Exemplarily, the fusion rule of the job type of the object to be evaluated can be the job requirements corresponding to the job grade of the object to be evaluated. For example, the sales quotas corresponding to sales employees and sales managers are different, and the work content is also different. Therefore, when evaluating the object to be evaluated, it is necessary to implement it in combination with the job requirements of the corresponding position. According to the job requirements of the job grade, assign corresponding weights to each evaluation dimension and indicator. The weight reflects the relative importance of each dimension and indicator in the object evaluation. Using weighted summation or other comprehensive scoring methods, the scores of each dimension and indicator are summarized to obtain the comprehensive performance score of the object to be evaluated.

[0092] Before generating the object evaluation result, it is necessary to first determine the work type of the object to be evaluated, wherein the work type corresponds to the fusion rule that matches it, and the fusion rule is determined based on the fusion rule.

[0093] Specifically, first, clarify the position of the object to be evaluated and its corresponding work type, such as sales, R&D, customer service and other positions. Fusion rule formulation: For each type of work, determine the specific fusion rule (the fusion rule is formulated in advance). The fusion rule can cover multiple dimensions such as work attitude, task completion, work efficiency, teamwork, etc., and clarify the weight and scoring criteria of each dimension. According to the fusion rule, determine the fusion method of each type of data to be evaluated. It can be a variety of methods such as weighted average, fuzzy comprehensive evaluation, hierarchical analysis method, etc. The analysis results of the positivity representation, target work state representation and target emotion representation are fused according to the fusion rule to generate the object evaluation result describing the object to be evaluated. The object evaluation result should clearly indicate whether the work attitude is positive or negative, and specific reasons and basis can be given. According to the type of work, the task completion of the object to be evaluated can also be quantitatively evaluated, including the number, quality, progress, etc. of completed tasks. Combine the evaluation information with the task completion, and calculate the final object evaluation result according to the weight and scoring criteria determined in the fusion rule.

[0094] Example 1: The job type of the object to be evaluated is sales. The corresponding fusion rules are: work attitude (30%), task completion (70%). Work attitude includes communication willingness (15%) and cooperation attitude (15%); task completion includes sales (50%) and customer satisfaction (20%). The analysis results of communication willingness and cooperation attitude are integrated with the data of sales and customer satisfaction to generate the evaluation information of the salesperson. Based on the evaluation information and task completion, the object evaluation result of the salesperson is calculated.

[0095] Example 2: The work type of the object to be evaluated is R&D. The corresponding fusion rules are: work attitude (40%), task completion (60%). Work attitude includes innovative thinking (20%) and teamwork (20%); task completion includes project progress (40%) and code quality (20%). The analysis results of innovative thinking and teamwork are integrated with the data of project progress and code quality to generate the evaluation information of R&D personnel. Based on the evaluation information and task completion, the object evaluation results of R&D personnel are calculated.

[0096] In the process of determining the fusion rules, the fusion rules for the analysis results corresponding to different data types are different. Among them, according to the fusion rules, the positivity representation, the target work state representation and the target emotion representation are fused to obtain the overall evaluation information, including: generating the first evaluation information based on the positivity representation according to the evaluation sub-rule of the positivity representation dimension in the fusion rules; generating the second evaluation information based on the target work state representation according to the evaluation sub-rule of the target work state representation dimension in the fusion rules; generating the third evaluation information based on the target emotion representation according to the evaluation sub-rule of the target emotion representation dimension in the fusion rules; merging the first evaluation information, the second evaluation information and the third evaluation information to obtain the overall evaluation information.

[0097] Among them, the object evaluation result is generated according to the overall evaluation information and the task completion amount of the object to be evaluated, including: generating the work completion parameters of the object to be evaluated based on the task completion amount of the object to be evaluated and the preset workload standard of the work type of the object to be evaluated; generating the object evaluation result based on the overall evaluation information and the work completion parameters.

[0098] Specifically, the work completion parameters include the work completion type and the work completion degree, and the overall evaluation information includes the evaluation type; if the work completion type is not meeting the preset performance standard and the evaluation type is a negative attitude, a first object evaluation result is generated based on the preset performance standard and the work completion degree;

[0099] If the work completion type is failure to meet the preset performance standard and the evaluation type is positive attitude, generating a second object evaluation result based on the preset performance standard and the work completion degree;

[0100] If the work completion type is meeting the preset performance standard and the evaluation type is positive attitude, a third object evaluation result is generated based on the preset performance standard and the work completion degree.

[0101] Among them, the evaluation results of the first object represent suggestions for optimizing the work content; the evaluation results of the second object represent suggestions for adjusting the work content; and the evaluation results of the third object represent suggestions for adjusting the work type.

[0102] For example, let's take one of the situations as an example. For example, the person to be evaluated is a sales representative who is mainly responsible for customer development and sales performance improvement. The preset performance standards include completing a certain number of new customer development per month, achieving a certain sales volume, and maintaining good customer feedback.

[0103] Evaluation sub-rules of the positive representation dimension: willingness to communicate, cooperative attitude, and customer orientation. The positive representation is that the employee is relatively passive in communication and lacks the willingness to actively develop new customers. The first evaluation information is that the employee is not good at communicating willingness and lacks a positive attitude towards customer orientation.

[0104] Evaluation sub-rules for the target work status representation dimension: work efficiency, task completion, and customer maintenance. The target work status is that the number of new customers developed by the employee this month is far below the preset standard, and the sales volume has not reached the expected target. The second evaluation information is that the employee's performance in work efficiency and new customer development is insufficient, the task completion is low, and the corresponding target work status is not good.

[0105] Evaluation sub-rules of the target emotion representation dimension: emotional stability, positive emotion ratio. The target emotion is that the employee's recent emotional fluctuations are large, and the negative emotion ratio is high. The third evaluation information is that the employee's emotional stability is not good, and negative emotions affect work performance.

[0106] Object evaluation results: Based on the evaluation information of the above three dimensions, the employee has not met the preset performance standards in many aspects, and the overall evaluation information is negative. Based on the preset performance standards and differences, the first object evaluation results are generated. This result not only points out the specific performance indicators that the employee has not achieved (such as the number of new customers developed and sales), but also analyzes in detail the employee's shortcomings in communication willingness, work efficiency, emotional stability, etc.

[0107] Suggestions for optimizing work content: Strengthen communication skills training, enhance customer-oriented awareness, and actively find and develop new customers. Develop detailed work plans and goals, improve work efficiency, and ensure that tasks are completed on time. Seek support for emotional management, learn to regulate and stabilize emotions, and maintain a positive work attitude. Regularly provide feedback and communicate with superiors and colleagues, and adjust work strategies and methods in a timely manner.

[0108] In summary, the disclosed embodiments obtain the data to be evaluated of the object to be evaluated. Since the data to be evaluated includes at least one of the text, audio and image of the object to be evaluated, the problem of single evaluation index is avoided to a certain extent. Furthermore, for any data to be evaluated, an analysis model corresponding to the data type of the data to be evaluated is selected, and the data to be evaluated is analyzed to obtain an analysis result, thereby avoiding the problem that the evaluation result is affected by subjective factors due to manual evaluation. Finally, the obtained analysis results are fused, and the fused result is used as the object evaluation result of the object to be evaluated, thereby improving the accuracy of the evaluation.

[0109] correspond Figure 3 The present disclosure also provides an object evaluation device, which can be deployed in Figure 1 The server 10 is used to perform the operations of the object evaluation method performed by the server in any of the above embodiments, such as Figure 4 As shown, the device comprises:

[0110] An acquisition module 401 is used to acquire data to be evaluated of the object to be evaluated, wherein the data type of the data to be evaluated includes at least one of text, audio and image;

[0111] A text analysis module 402 is used for, if the data type of the data to be evaluated is text, calling a preset large language model to perform a first operation on the data to be evaluated to obtain a positive representation, wherein the first operation is to extract semantic features of the data to be evaluated and determine the positive representation based on the semantic features;

[0112] The image analysis module 403 is used for, if the data type of the data to be evaluated is an image, calling the image recognition model to perform a second operation on the data to be evaluated to obtain a target working state representation, wherein the second operation is to identify the facial features of the object to be evaluated in the data to be evaluated, and determine the target working state representation based on the mapping relationship between the facial features and the preset working state;

[0113] The audio analysis module 404 is used to, if the data type of the data to be evaluated is audio, call the speech recognition model to perform a third operation on the data to be evaluated to obtain a target emotion representation, wherein the third operation is to extract audio features of the data to be evaluated, and determine the target emotion representation corresponding to the audio features based on a mapping relationship between the audio features and the preset emotion representation;

[0114] The determination module 405 is used to perform a fusion operation on the positivity representation, the target working state representation and the target emotion representation, and use the fusion result as the object evaluation result of the object to be evaluated.

[0115] In the embodiment of the present disclosure, the determining module includes:

[0116] A type rule determination unit, used to determine the work type of the object to be evaluated; and determine a fusion rule matching the work type;

[0117] a fusion unit, configured to perform a fusion operation on the positivity representation, the target work state representation, and the target emotion representation according to the fusion rule to obtain overall evaluation information, wherein the overall evaluation information is a positive attitude or a negative attitude;

[0118] A result generating unit is used to generate the object evaluation result according to the overall evaluation information and the completion amount of the task of the object to be evaluated.

[0119] In the embodiment of the present disclosure, the fusion unit includes:

[0120] A first evaluation subunit, configured to generate first evaluation information based on the positivity representation according to the evaluation subrule of the positivity representation dimension in the fusion rule;

[0121] A second evaluation subunit, configured to generate second evaluation information based on the target working state representation according to the evaluation subrule of the target working state representation dimension in the fusion rule;

[0122] A third evaluation subunit, configured to generate third evaluation information based on the target emotion representation according to the evaluation subrule of the target emotion representation dimension in the fusion rule;

[0123] The merging subunit is used to merge the first evaluation information, the second evaluation information and the third evaluation information to obtain the overall evaluation information.

[0124] In the embodiment of the present disclosure, the result generating unit is further used to:

[0125] Generate work completion parameters of the object to be evaluated based on the task completion amount of the object to be evaluated and the preset workload standard of the work type of the object to be evaluated;

[0126] An object evaluation result is generated based on the overall evaluation information and the work completion parameter.

[0127] In the embodiment of the present disclosure, the work completion parameters include the work completion type and the work completion degree, the overall evaluation information includes the evaluation type, and the result generating unit is further used to:

[0128] If the work completion type is failure to meet the preset performance standard and the evaluation type is negative attitude, generating a first object evaluation result based on the preset performance standard and the work completion degree;

[0129] If the work completion type is not meeting the preset performance standard and the evaluation type is positive attitude, generating a second object evaluation result based on the preset performance standard and the work completion degree;

[0130] If the work completion type is satisfying the preset performance standard and the evaluation type is positive attitude, a third object evaluation result is generated based on the preset performance standard and the work completion degree.

[0131] The object evaluation device provided in the above-mentioned embodiment of the present disclosure and the object evaluation method provided in the embodiment of the present disclosure are based on the same inventive concept, and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0132] The present disclosure also provides an electronic device to perform the object evaluation method. Figure 5 , which shows a schematic diagram of an electronic device provided by some embodiments of the present disclosure. Figure 5 As shown, the electronic device 5 includes: a processor 500, a memory 501, a bus 502 and a communication interface 503, wherein the processor 500, the communication interface 503 and the memory 501 are connected via the bus 502; the memory 501 stores a computer program that can be run on the processor 500, and when the processor 500 runs the computer program, the object evaluation method provided in any of the aforementioned embodiments of the present disclosure is executed.

[0133] It should be pointed out that Figure 5 The electronic device shown can be implemented as Figure 1 In the server device or client device, when the electronic device is implemented as a server device, the aforementioned monitoring model can also be stored in the memory 501. When the electronic device is implemented as a client device, the electronic device can also include a display.

[0134] The memory 501 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk storage. The communication connection between the virtual devices of the system is realized through at least one communication interface 503 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0135] The bus 502 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The memory 501 is used to store programs, and the processor 500 executes the programs after receiving execution instructions. The object evaluation method disclosed in any implementation of the aforementioned embodiment of the present disclosure may be applied to the processor 500, or implemented by the processor 500.

[0136] The processor 500 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit in the processor 500 or the instruction in the form of software. The above processor 500 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 gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present disclosure can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in conjunction with the embodiments of the present disclosure can be directly embodied as a hardware decoding processor to be executed, or the hardware and software modules in the decoding processor can be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 501, and the processor 500 reads the content in the memory 501 and completes the steps of the above method in combination with its hardware.

[0137] The electronic device provided by the embodiment of the present disclosure and the object evaluation method provided by the embodiment of the present disclosure are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by them.

[0138] The present disclosure also provides a computer-readable storage medium corresponding to the object evaluation method provided in the above embodiment. Figure 6 The computer-readable storage medium shown is a CD 30 on which a computer program (ie, a program product) is stored. When the computer program is run by a processor, the object evaluation method provided by any of the aforementioned embodiments is executed.

[0139] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical or magnetic storage media, which are not listed here one by one.

[0140] The computer-readable storage medium provided by the above-mentioned embodiments of the present disclosure and the object evaluation method provided by the embodiments of the present disclosure are based on the same inventive concept, and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0141] Although the optional embodiments of the present disclosure have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present disclosure.

[0142] The specific implementation methods described above further illustrate the purpose, technical solutions and beneficial effects of the present disclosure. It should be understood that the above description is only the specific implementation method of the present disclosure and is not intended to limit the protection scope of the present disclosure. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solutions of the present disclosure should be included in the protection scope of the present invention.

Claims

1. An object evaluation method, characterized in that: The method comprises: Acquire data to be evaluated of the object to be evaluated, wherein the data type of the data to be evaluated includes at least one of text, audio, and image; If the data type of the data to be evaluated is text, calling a preset large language model to perform a first operation on the data to be evaluated to obtain a positive representation, wherein the first operation is to extract semantic features of the data to be evaluated and determine the positive representation based on the semantic features; If the data type of the data to be evaluated is an image, calling the image recognition model to perform a second operation on the data to be evaluated to obtain a target working state representation, wherein the second operation is to identify facial features of the object to be evaluated in the data to be evaluated, and determine the target working state representation based on a mapping relationship between the facial features and the preset working state; If the data type of the data to be evaluated is audio, calling the speech recognition model to perform a third operation on the data to be evaluated to obtain a target emotion representation, wherein the third operation is to extract audio features of the data to be evaluated, and based on a mapping relationship between the audio features and a preset emotion representation, determine a target emotion representation corresponding to the audio features; A fusion operation is performed on the positivity representation, the target work state representation, and the target emotion representation, and a fusion result is used as an object evaluation result of the object to be evaluated.

2. The method according to claim 1, characterized in that The performing a fusion operation on the positivity representation, the target work state representation, and the target emotion representation, and using the fusion result as the object evaluation result of the object to be evaluated, includes: Determining the work type of the object to be evaluated; and determining a fusion rule matching the work type; Performing a fusion operation on the positivity representation, the target work state representation, and the target emotion representation according to the fusion rule to obtain overall evaluation information; An object evaluation result is generated according to the overall evaluation information and the task completion amount of the object to be evaluated.

3. The method according to claim 2, characterized in that The performing a fusion operation on the positivity representation, the target work state representation, and the target emotion representation according to the fusion rule to obtain overall evaluation information includes: According to the evaluation sub-rule of the positivity representation dimension in the fusion rule, generating first evaluation information based on the positivity representation; According to the evaluation sub-rule of the target working state representation dimension in the fusion rule, second evaluation information is generated based on the target working state representation; According to the evaluation sub-rule of the target emotion representation dimension in the fusion rule, generating third evaluation information based on the target emotion representation; The first evaluation information, the second evaluation information and the third evaluation information are combined to obtain the overall evaluation information.

4. The method according to claim 2, characterized in that: The generating of the object evaluation result according to the overall evaluation information and the task completion amount of the object to be evaluated includes: Generate work completion parameters of the object to be evaluated based on the task completion amount of the object to be evaluated and the preset workload standard of the work type of the object to be evaluated; An object evaluation result is generated based on the overall evaluation information and the work completion parameter.

5. The method according to claim 4, characterized in that The work completion parameters include the work completion type and the work completion degree, the overall evaluation information includes the evaluation type, and the object evaluation result is generated based on the overall evaluation information and the work completion parameters, including: If the work completion type is failure to meet the preset performance standard and the evaluation type is negative attitude, generating a first object evaluation result based on the preset performance standard and the work completion degree; If the work completion type is not meeting the preset performance standard and the evaluation type is positive attitude, generating a second object evaluation result based on the preset performance standard and the work completion degree; If the work completion type is satisfying the preset performance standard and the evaluation type is positive attitude, a third object evaluation result is generated based on the preset performance standard and the work completion degree.

6. An object evaluation device, characterized in that The device comprises: An acquisition module is used to acquire the data to be evaluated of the object to be evaluated, wherein the data type of the data to be evaluated includes at least one of text, audio and image; A text analysis module, for, if the data type of the data to be evaluated is text, calling a preset large language model to perform a first operation on the data to be evaluated to obtain a positive representation, wherein the first operation is to extract semantic features of the data to be evaluated and determine the positive representation based on the semantic features; an image analysis module, for, if the data type of the data to be evaluated is an image, calling an image recognition model to perform a second operation on the data to be evaluated to obtain a target working state representation, wherein the second operation is to identify facial features of the object to be evaluated in the data to be evaluated, and determine the target working state representation based on a mapping relationship between the facial features and a preset working state; An audio analysis module, for, if the data type of the data to be evaluated is audio, calling a speech recognition model to perform a third operation on the data to be evaluated to obtain a target emotion representation, wherein the third operation is to extract audio features of the data to be evaluated, and determine the target emotion representation corresponding to the audio features based on a mapping relationship between the audio features and a preset emotion representation; The determination module is used to perform a fusion operation on the positivity representation, the target working state representation and the target emotion representation, and use the fusion result as the object evaluation result of the object to be evaluated.

7. The device according to claim 6, characterized in that The determining module comprises: A type rule determination unit, used to determine the work type of the object to be evaluated; and determine a fusion rule matching the work type; A fusion unit, configured to perform a fusion operation on the positivity representation, the target work state representation, and the target emotion representation according to the fusion rule to obtain overall evaluation information; A result generating unit is used to generate an object evaluation result according to the overall evaluation information and the task completion amount of the object to be evaluated.

8. The device according to claim 7, characterized in that The fusion unit comprises: A first evaluation subunit, configured to generate first evaluation information based on the positivity representation according to the evaluation subrule of the positivity representation dimension in the fusion rule; A second evaluation subunit, configured to generate second evaluation information based on the target working state representation according to the evaluation subrule of the target working state representation dimension in the fusion rule; A third evaluation subunit, configured to generate third evaluation information based on the target emotion representation according to the evaluation subrule of the target emotion representation dimension in the fusion rule; The merging subunit is used to merge the first evaluation information, the second evaluation information and the third evaluation information to obtain the overall evaluation information.

9. An electronic device comprising a memory, a processor and a computer program stored in the memory, wherein the processor implements the method according to any one of claims 1 to 5 when executing the computer program.

10. A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images