Intelligent reception management method and related equipment

Abstract

The invention provides an intelligent reception management method and related equipment, and is applied to the technical field of data processing. The method comprises the steps of obtaining target customer information, target reception theme information, reception place information and a training sample set; preprocessing the training sample set to generate a training sample set with target feature data; processing a preset reception management model based on the training sample set with the target feature data to generate a target reception management model; processing the target customer information to generate state parameter information of the target customer; processing the target reception theme information matched with the target customer information to generate a reception arrangement influence factor; processing the reception place information based on the target reception management model to generate attribute information of the reception place; and processing the attribute information of the reception place, the state parameter information of the target customer and the reception arrangement influence factor based on the target reception management model to generate reception management information of the reception place.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a management method and related equipment for intelligent reception. Background Art

[0002] At present, most companies use manual reception to achieve the purpose of receiving customers. Because the previous method has many problems in the reception process, it cannot meet the company's needs to improve customer satisfaction, and at the same time causes consumption and waste of human resources, which in turn affects the company's profitability.

[0003] In the existing technology, corporate reception management usually lacks scientific planning and personalized services. The arrangement of reception venues is often arbitrary, without fully considering the specific needs of target customers, the characteristics of the reception theme, and various factors of the reception venue itself. Therefore, a new corporate reception management solution is needed to improve the reception quality and customer satisfaction.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0005] The purpose of this application is to provide a management method and related equipment for intelligent reception, which at least overcomes the problems existing in the prior art to a certain extent. By obtaining information on target customers, reception themes, reception venues, and other aspects and a preset management model, the training sample set is preprocessed to generate target feature data. The preset model is used to process the data to obtain a target reception management model, and the target customer information and reception theme information are processed in multiple steps to generate state parameter information and reception arrangement influencing factors. When processing the reception venue information, monitoring point information, grouping information, feature data sets, etc. are generated to ultimately determine the reception venue attribute information. Finally, comprehensive processing of multiple aspects of information generates reception customization values ​​and management information, realizes scientific planning and personalized services, and improves reception quality and efficiency.

[0006] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by the practice of the present invention.

[0007] According to one aspect of the present application, a management method for intelligent reception is provided, comprising: obtaining target customer information, target reception theme information matching the target customer information, reception venue information and a training sample set, wherein the reception venue information comprises location information within the reception venue, and the training sample set comprises historical environmental factor information of the reception venue and real-time environmental factor information of other venues adjacent to the reception venue; preprocessing the training sample set to generate a training sample set with target feature data, wherein the target feature data is used to characterize risk factors affecting the pre-arrangement of seats at the reception venue; obtaining a preset reception management model, wherein the preset reception management model is generated based on historical reception environmental factor information of other reception venues; and generating a training sample set with target feature data based on the training sample set with target feature data. The preset reception management model is processed based on the training sample set of the recruitment data to generate a target reception management model; the target customer information is processed to generate the target customer's state parameter information, wherein the target customer's state parameter information is used to characterize the target customer's state in the reception scenario; the target reception theme information matching the target customer information is processed to generate a reception arrangement influencing factor; the reception venue information is processed based on the target reception management model to generate the attribute information of the reception venue, wherein the attribute information of the reception venue includes the classification information of the seats; the attribute information of the reception venue, the target customer's state parameter information and the reception arrangement influencing factor are processed based on the target reception management model to generate the reception management information of the reception venue.

[0008] Another aspect of the present application is a management device for intelligent reception, characterized in that it includes: an acquisition module, used to acquire target customer information, target reception theme information matching the target customer information, reception venue information and a training sample set, wherein the reception venue information includes location information within the reception venue, and the training sample set includes historical environmental factor information of the reception venue and real-time environmental factor information of other venues adjacent to the reception venue; acquiring a preset reception management model, wherein the preset reception management model is generated based on historical reception environmental factor information of other reception venues; a processing module, used to pre-process the training sample set to generate a training sample set with target feature data, wherein the target feature data is used to characterize risk factors affecting the pre-arrangement of seats at the reception venue; based The preset reception management model is processed on the training sample set with target feature data to generate a target reception management model; the target customer information is processed to generate target customer status parameter information, wherein the target customer status parameter information is used to characterize the target customer's status in the reception scenario; the target reception theme information matching the target customer information is processed to generate a reception arrangement influencing factor; the reception venue information is processed based on the target reception management model to generate attribute information of the reception venue, wherein the attribute information of the reception venue includes classification information of seats; the attribute information of the reception venue, the target customer status parameter information and the reception arrangement influencing factor are processed based on the target reception management model to generate reception management information of the reception venue.

[0009] According to another aspect of the present application, an electronic device is provided, characterized in that it comprises: a first processor; and a memory for storing executable instructions of the first processor; wherein the first processor is configured to execute the above-mentioned intelligent reception management method by executing the executable instructions.

[0010] According to another aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a second processor, the above-mentioned intelligent reception management method is implemented.

[0011] According to another aspect of the present application, a computer program product is provided, including a computer program, characterized in that when the computer program is executed by a third processor, the above-mentioned intelligent reception management method is implemented.

[0012] The present application provides a management method and related equipment for intelligent reception. The server obtains information on target customers, reception themes, reception venues, and other aspects and a preset management model, and pre-processes the training sample set to generate target feature data. The preset model is used to process the data to obtain a target reception management model, and the target customer information and reception theme information are processed in multiple steps to generate state parameter information and reception arrangement influencing factors. When processing the reception venue information, monitoring point information, grouping information, feature data sets, etc. are generated to ultimately determine the reception venue attribute information. Finally, comprehensive processing of multiple aspects of information generates reception customization values ​​and management information to achieve scientific planning and personalized services, and improve reception quality and efficiency.

[0013] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A flowchart showing a method for managing intelligent reception provided by an embodiment of the present application is shown; Figure 2 A schematic diagram of the structure of a smart reception management device provided in one embodiment of the present application is shown. DETAILED DESCRIPTION

[0015] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0016] Combine the following Figure 1 To describe the management method of intelligent reception according to the exemplary embodiment of the present application. It should be noted that the following application scenarios are only shown to facilitate understanding of the spirit and principle of the present application, and the implementation of the present application is not limited in this regard. On the contrary, the implementation of the present application can be applied to any applicable scenario.

[0017] In one implementation, the present application also proposes a smart reception management method and related equipment. Figure 1 The following schematically shows a flow chart of a smart reception management method according to an embodiment of the present application. Figure 1 As shown, the method is applied to a server, comprising: S101, obtaining target customer information, target reception theme information matching the target customer information, reception venue information, training sample set and preset reception management model.

[0018] In one embodiment, the reception venue information includes location information within the reception venue, the training sample set includes historical environmental factor information of the reception venue and real-time environmental factor information of other venues adjacent to the reception venue, and the preset reception management model is generated based on the historical reception environmental factor information of other reception venues. In the enterprise multifunctional conference training scenario, the target customer information is diverse and targeted, and the basic information is as follows: Name: Li Si, Age: 35, Gender: Female, Department: Marketing Department, Position: Marketing Manager. Participation in past training and meeting records: Participated in many market strategy seminars, and proposed an innovative promotion plan in one of the meetings on online marketing; participated in new employee induction training (sharing experience as an old employee). The training demand information is as follows: This time, I participated in the multifunctional conference training, hoping to learn in depth the latest market trend analysis methods, especially the application of big data in market research. It is expected that through training, the team management ability will be improved, and how to motivate team members to work efficiently in a complex and changing market environment will be learned. Pay attention to the training content on cross-departmental collaboration, because the marketing department often needs to collaborate with other departments (such as R&D department and sales department) to complete projects, and hope to improve communication and collaboration efficiency. The learning style information is as follows: I prefer visual learning and have a high acceptance of training content with rich pictures, rich cases and video demonstrations. In the learning process, I like to understand the overall framework first, then go into the details, and pay attention to the systematic and logical nature of knowledge. I am good at learning in interactive communication and hope to have group discussions, brainstorming and other links to get inspiration from others' experiences and opinions.

[0019] The reception theme of the enterprise multi-functional conference training should be closely centered on the training purpose and customer needs. The comprehensive training meeting of "Market Innovation and Teamwork" covers multiple aspects such as market innovation strategy sharing, teamwork skills improvement, and cross-departmental communication case analysis, aiming to comprehensively improve the comprehensive capabilities of marketing department employees and promote teamwork to cope with fierce market competition. Topics include market trend interpretation (inviting industry experts to give lectures), successful case analysis (sharing excellent cases within the company and analyzing external classic cases), team building activities (enhancing cohesion through group games and team challenges), and cross-departmental collaboration simulation (simulating real project scenarios to allow employees from different departments to work together to complete tasks). In terms of technical equipment, high-definition projectors, smart electronic whiteboards (for real-time writing and interaction), sound systems (to ensure clear sound), and high-speed and stable wireless networks (supporting online data query and real-time interaction) are required. Material resources include customized training manuals (containing detailed training content, case materials, and interactive questions), colored markers and sticky notes (for group discussion records), and team building props (such as puzzles, ropes, etc.). The training meeting lasted for three days. On the first day, there was a lecture on market trend interpretation in the morning and team building activities in the afternoon. On the second day, there was case analysis and group discussion all day. On the third day, there was cross-departmental collaboration simulation in the morning and experience summary and action plan formulation in the afternoon. Every day, the training started at 9:00 a.m., with a tea break from 10:30 to 10:45, a lunch break from 12:00 to 1:00 p.m., and continued at 1:00 p.m., with a short break from 15:00 to 15:15, and the training ended at 17:00 p.m.

[0020] The training venue is located on the second floor of the multifunctional conference center in the enterprise park. The conference center has convenient transportation, a dedicated park bus stop, and is no more than 10 minutes' walk from the main office areas of the enterprise. The conference center is surrounded by beautiful environment, with gardens and leisure areas, which are convenient for students to relax during their breaks. The multifunctional hall on the second floor is about 200 square meters and can flexibly accommodate 80-100 people for conference training activities. It can be adjusted in a variety of layouts according to different training forms (lectures, group discussions, team activities, etc.). It has a variety of tables and chairs that can be flexibly combined, and can be quickly adjusted to classroom style (suitable for lectures), group style (suitable for group discussions), island style (suitable for team collaboration activities) and other layouts according to training needs. There is a large stage in front, equipped with professional lighting and sound equipment, which can be used for lecturers to give lectures, guest speeches and results presentations. Equipped with high-definition cameras and video conferencing systems, it supports remote participation and online live broadcasts; each seat has a power socket and network interface to facilitate students to use electronic devices; an intelligent control system is installed to adjust indoor lighting, temperature and humidity. There is a dedicated service desk that provides printing, copying, binding and other services; there is an independent storage room for training materials and student belongings; the bathroom is spacious and bright, and is equipped with a full range of sanitary products.

[0021] In terms of temperature and humidity, the air conditioning system of the multifunctional hall had insufficient cooling power during hot summer weather, resulting in high indoor temperatures and affecting the students' attention and participation; in winter, humidity regulation was sometimes unstable and the air was too dry, making the students feel uncomfortable. In terms of noise control, due to its proximity to the main road of the park, the noise from passing vehicles would interfere with conference training during the day, especially during the sessions that require quiet thinking and listening. In terms of equipment use, in the past, when some high-tech trainings were held, the electronic whiteboard had a brief malfunction, which affected the normal progress of the training; the sound system occasionally made noises when the volume was high. In terms of light conditions, some window seats would affect students' viewing of the screen content when the sun was strong in the morning, and additional shading measures were required. In addition, adjacent conference rooms sometimes hold other activities at the same time, and the movement of people and the sound would be transmitted through the corridor, which would have a certain impact on the current training. The restaurant downstairs is noisy during lunch and dinner time. Although there are certain sound insulation measures, there will still be slight noise affecting the training venue. The nearby gym will have more people activities during peak hours (such as in the evening), and the sound of music and fitness equipment in use will interfere with the training environment.

[0022] The preset reception management model consists of an input layer, a hidden layer, and an output layer. Specifically, the information obtained by the input layer is as follows: target customer information, basic information: name, age, gender, department, position, etc. as separate input nodes. For example, "name: Zhang San", "age: 30", "gender: male", "department: R&D department", "position: engineer", etc. This information helps the model understand the basic background of the customer, thus providing a basis for subsequent personalized reception. Training demand information is input in the form of keywords or vectors. For example, if a customer wants to improve project management capabilities, it can be expressed as "project management training needs: high" and other similar forms. For multi-faceted training needs, such as "technical skills improvement, team collaboration enhancement", etc., it can be split into multiple input nodes or represented by appropriate encoding methods. Learning style information is also input in a classified or quantitative manner. For example, "visual learning preference: strong", "auditory learning preference: medium", "kinesthetic learning preference: weak", etc., or the learning style is mapped to a numerical vector input.

[0023] Target reception theme information, theme type: input in the form of text classification labels, such as "technical training", "management training", "market training", etc. Theme characteristics and requirements: convert information such as interactive requirements (such as "group discussion frequency: high", "case analysis requirements: medium", etc.), resource requirements ("HD projector requirements: yes", "simulation equipment requirements: no", etc.), time arrangement ("number of training days: 3", "daily training duration: 8 hours", etc.) into specific numerical values ​​or classification labels for input. Reception venue information, location information: including venue address (in text form), traffic convenience (such as "distance to bus station: 500 meters", "number of parking spaces: 50", etc.), surrounding environment ("surrounding noise source intensity: low", "surrounding leisure facilities richness: medium", etc.). Venue layout and facilities: input venue area ("area: 200 square meters"), layout type (such as "layout: classroom style", "capacity: 100", etc.), facility configuration ("HD projector: yes", "smart whiteboard: no", etc., represented by 0 / 1 or other appropriate codes).

[0024] The relevant information of the training sample set is as follows: Historical environmental factor information: such as the average temperature ("historical average temperature: 23 degrees Celsius"), humidity ("historical average humidity: 50%), noise level ("historical average noise: 40 decibels"), equipment failure rate ("projector failure rate: 5%"), etc., which are numerical inputs after statistical processing when events were held at the venue in the past. Real-time environmental factor information of adjacent venues: Similarly, the real-time noise of adjacent offices ("real-time noise of adjacent offices: 35 decibels") and the busyness of nearby restaurants ("nearby restaurant busyness index: medium") are converted into appropriate numerical or categorical inputs. This information can be used as a reference for the model to predict the current reception situation.

[0025] Hidden layers include the following: Feature processing layer (which can be regarded as part of the hidden layer), which extracts and transforms features of various information in the input layer. For example, for the training demand information of customers, the text description is converted into a numerical vector through the word vector model so that it can be better processed by the subsequent neural network. For the equipment of the venue, feature encoding is performed, such as encoding the presence or absence of multiple facilities into a binary vector, and then feature fusion and dimension transformation are performed through the fully connected layer to extract more representative high-level features. Multi-layer neural network structure (using neural network model), the first hidden layer: contains multiple neurons, such as 100 neurons (the specific number is adjusted according to actual conditions). The neurons in this layer receive all feature information from the input layer, perform nonlinear transformation through weighted summation and activation function (such as ReLU function), and begin to learn the complex relationship between input features. The second hidden layer: The number of neurons can be appropriately reduced, such as 50. This layer further abstracts and combines the features output by the first hidden layer to capture higher-level patterns. For example, information such as customer needs, theme characteristics and venue characteristics begins to be integrated to form an intermediate representation related to training effects and reception plans. The third hidden layer: further refines the feature representation, and the number of neurons can be about 30. This layer will focus on the interactive impact of some key factors, such as the matching degree between the customer's learning style and the subject interactivity requirements in a specific environment, to provide a more accurate decision-making basis for the final output.

[0026] Output layer, output a numerical value or classification result, indicating the predicted training effect. For example, a numerical value between 0 and 1 is used to indicate the predicted value of trainee satisfaction, or classification labels such as "high satisfaction", "medium satisfaction", and "low satisfaction" are output. Quantitative indicators of knowledge mastery can also be predicted, such as the expected average score of test scores. The recommended seating arrangement is output in the form of a matrix, such as "[0,1,0;1,0,1;0,1,0]" to indicate whether the seats in each row and column are recommended (1 indicates recommended, 0 indicates not recommended), or the layout type number is directly output, such as "classroom layout: 1" and "group layout: 0". The device's on status and parameter setting suggestions are output, such as specific values ​​such as "projector brightness: 80%", "audio volume: 70%", and "air conditioning temperature: 22 degrees Celsius". Environmental adjustment strategies include target values ​​of environmental factors, such as "indoor noise control target: 30 decibels", "humidity adjustment target: 45%", etc., as well as corresponding adjustment measures (such as "turn on sound insulation: yes", "increase humidifier running time: yes", etc., 0 / 1). Output tea break and lunch time suggestions (such as "tea break time: 10:30-10:45 am" and "lunch time: 12:00-13:00", etc.) and catering type recommendations ("tea break food: fruits, snacks", "lunch menu: meat and vegetable combo set meal" and other text descriptions or classification numbers). Through this structural design from the input layer to the hidden layer and then to the output layer, the model can comprehensively process various input information. After complex internal calculations and feature learning, it can provide accurate training effect predictions and reasonable reception arrangement suggestions for corporate multi-functional conference training, thereby improving the quality and efficiency of conference training.

[0027] S102, preprocessing the training sample set to generate a training sample set with target feature data.

[0028] In one implementation, feature extraction is performed on the training sample set to generate an original feature library. Assume that the enterprise has multiple multifunctional conference and training venues, and collects relevant data on various conference and training activities held in these venues in the past year as a training sample set. It includes conference types (such as internal training, customer seminars, industry summits, etc.), information on participants (department, position, number of people, etc.), meeting time (weekdays / weekends, morning / afternoon / evening, etc.), venue environmental data (temperature, humidity, noise, etc.), equipment usage (whether the projector, audio, etc. are working properly) and feedback on the effect of the meeting (participant satisfaction score, knowledge mastery evaluation, etc.). Extract classification features of the conference type, such as "internal training" marked as 1, "customer seminar" marked as 2, "industry summit" marked as 3, etc. At the same time, extract some derivative features based on the characteristics of the conference type, such as for internal training, extract training topics (such as technical training, management training, etc.), training instructor level (senior / intermediate / primary), etc.; for customer seminars, extract features such as the industry to which the customer belongs and the size of the customer.

[0029] For the departments of the participants, one-hot encoding is performed. For example, if the company has departments such as sales, marketing, and R&D, if the participants are from the sales department, the code position corresponding to the sales department is 1, and the code position for other departments is 0. Positions are also processed similarly, such as manager level, supervisor level, employee level, etc. The number of people is directly extracted as a numerical feature. The meeting time is divided into weekdays (Monday to Friday) and weekends, marked as 0 and 1 respectively. For the time of day, it is divided into morning (9:00-12:00), afternoon (13:00-18:00) and evening (19:00-22:00), marked as 1, 2, and 3 respectively. At the same time, the meeting duration is considered and taken as a numerical feature in hours. Temperature, humidity, and noise are directly extracted as numerical features. In addition, according to the comfortable range of temperature and humidity (such as temperature 22-25 degrees Celsius, humidity 40%-60% is comfortable), a classification feature of "environmental comfort" is generated, which is divided into comfortable, relatively comfortable, and uncomfortable categories. For noise, depending on whether it exceeds a certain decibel value (such as 50 decibels), a classification feature of "noise interference level" is generated, which is divided into low interference, medium interference, and high interference.

[0030] For equipment such as projectors and audio equipment, normal operation is marked as 1 and failure is marked as 0. At the same time, the equipment usage time is recorded as a numerical feature. The satisfaction score of participants (0-10 points) and the knowledge mastery level assessment (0-100 points) are directly extracted as numerical features. Through the above feature extraction process, all features are combined to form an original feature library. For example, a sample is represented as [meeting type code, participant department code, participant position code, number of participants, meeting time code, meeting duration, temperature, humidity, environmental comfort code, noise, noise interference level code, projector status, audio status, equipment usage time, satisfaction score, knowledge mastery level score]. The original feature library is processed to generate a training set and a validation set. The original feature library is divided into a training set and a validation set by random sampling. Assuming that there are 1,000 samples in the original feature library, 70% (700 samples) are used as the training set and 30% (300 samples) are used as the validation set. In the division process, it is necessary to ensure that the distribution of the training set and the validation set on each feature is as consistent as possible. For example, for the meeting type feature, the proportion of various types of meetings such as "internal training", "customer seminars", and "industry summits" in the training set should be similar to that in the original feature library; for the environmental comfort feature, the proportions of comfortable, relatively comfortable, and uncomfortable categories should also remain similar. This can ensure the effectiveness and accuracy of the model during training and verification.

[0031] Based on the classifier, each validation set divided by the original feature library is predicted to generate prediction results. A decision tree classifier is selected to predict the validation set. The decision tree classifier can build a decision tree model based on the relationship between features and classify and predict samples. Use the training set to train the decision tree classifier. During the training process, the decision tree classifier learns the relationship between features and classification results and builds a decision tree model based on the characteristics of the samples in the training set and the corresponding meeting effect feedback (such as the high and low classification of satisfaction scores, which can be set as 7 points and above for high satisfaction and below 7 points for low satisfaction). For example, it finds that when the meeting type is internal training, the number of participants is small, the environmental comfort is comfortable, and the equipment is working properly, the probability of a high satisfaction score is greater, thereby constructing the corresponding decision branch.

[0032] The samples in the validation set are input into the trained decision tree model. The model makes decisions on the decision tree based on the characteristics of the samples and predicts the satisfaction score classification (high satisfaction or low satisfaction) of each sample. For example, for a sample in the validation set [the meeting type is coded as 1 (internal training), the department of the participants is coded as the sales department, the position of the participants is coded as the supervisor level, the number of participants is 20, the meeting time is coded as weekday morning, the meeting duration is 3 hours, the temperature is 23 degrees Celsius, the humidity is 50%, the environmental comfort is coded as comfortable, the noise is 40 decibels, the noise interference level is coded as low interference, the projector status is 1 (normal), the audio status is 1 (normal), the equipment usage time is 2 hours, the satisfaction score is unknown, and the knowledge mastery score is unknown], the model predicts its satisfaction score classification according to the rules of the decision tree. After predicting all the samples in the validation set, a series of prediction results are obtained, such as [sample 1: predicted as high satisfaction, sample 2: predicted as low satisfaction, sample 3: predicted as high satisfaction...]. These prediction results will be used for comparative analysis with the subsequent validation set prediction results.

[0033] Based on the preset algorithm, each training set divided by the original feature library is trained to generate the prediction results of the validation set. The random forest algorithm is selected to train the training set. The random forest algorithm is an ensemble learning algorithm based on decision trees. It improves the accuracy and stability of the model by building multiple decision trees and combining their prediction results. When training the random forest algorithm, multiple sub-sample sets (for example, 100 sub-sample sets) are randomly extracted from the training set with replacement. For each sub-sample set, a decision tree model is constructed. When constructing a decision tree, each time a node is split, a part of the features (for example, features are selected, k is the total number of features) is randomly selected for consideration instead of using all the features, which can increase the diversity of the model. After the random forest model training is completed, the validation set samples are input into the random forest model. Each decision tree predicts the validation set samples and obtains a prediction result (such as satisfaction score classification). Then the prediction results of all decision trees are combined, for example, by majority voting (if most decision trees predict high satisfaction, the final prediction result is high satisfaction), to obtain the final prediction result of the validation set samples, that is, the validation set prediction result. This result is similar to the result of direct prediction of the validation set based on the decision tree classifier, but is more accurate and stable due to the ensemble nature of the random forest algorithm.

[0034] The prediction results and the validation set prediction results are processed to generate target feature data, which is used to characterize the risk factors that affect the pre-arrangement of seats in the reception venue. Compare the prediction results based on the decision tree classifier with the prediction results based on the random forest algorithm. Calculate the accuracy, that is, the proportion of correctly predicted samples to the total number of samples. For example, if the decision tree classifier correctly predicts 150 of the 200 validation set samples, the accuracy is 150 / 200=75%; the random forest algorithm correctly predicts 180, and the accuracy is 180 / 200=90%. At the same time, other evaluation indicators such as recall rate and F1 value can also be calculated to comprehensively evaluate the performance of the two prediction results. Observe the samples with inconsistent prediction results and analyze the reasons. For example, it is found that for some samples with long meeting time and average environmental comfort, the decision tree classifier predicts low satisfaction, while the random forest algorithm predicts high satisfaction. Further analysis revealed that the decision tree classifier relied too much on the meeting duration feature, while the random forest algorithm made different predictions by combining the results of multiple decision trees and taking into account the impact of other features (such as good device usage).

[0035] According to the prediction results and analysis, the features that have a greater impact on the meeting effect are selected as target feature data. These features also indirectly represent the risk factors that affect the pre-arrangement of seats in the reception venue. For example, it is found that the features such as environmental comfort, equipment status (projector and audio), and meeting type have a greater impact on the prediction results. If the environmental comfort is poor, the equipment often fails, and the meeting type is an important customer seminar, the probability of a low satisfaction score is higher, then these features become target feature data. Quantify and represent the target feature data. For example, environmental comfort can be represented by a value of 1-3 (1 for comfortable, 2 for relatively comfortable, and 3 for uncomfortable); equipment status is represented by 0 / 1 for normal or faulty; and meeting type is represented by classification code. These quantified target feature data are used in the subsequent reception management model. For example, when pre-arranging seats, better seating areas are arranged for important customer seminars with high environmental comfort and good equipment status to reduce the risk of low satisfaction.

[0036] S103: Process the preset reception management model based on the training sample set with target feature data to generate a target reception management model.

[0037] In one implementation, some key target feature data are determined from the previous processing of the training sample set, such as "environmental comfort" (divided into three levels: comfortable, relatively comfortable, and uncomfortable, represented by 1, 2, and 3 respectively), "equipment failure rate" (the ratio of the number of equipment failures to the total number of times used in the past period of time), "meeting importance level" (divided into three levels: high, medium, and low according to factors such as the level of participants in the meeting and the purpose of the meeting, represented by 1, 2, and 3 respectively), etc. A training sample contains these target feature data and the corresponding actual reception arrangement results (such as the actual seat layout effect score, feedback on the actual use of the equipment, etc.). For example, a sample is [environmental comfort: 1, equipment failure rate: 0.05, meeting importance level: 1, seat layout effect score: 8 (out of 10 points), smoothness of equipment use: high].

[0038] The training samples with target feature data are input into the input layer of the preset reception management model in sequence. The data is forward propagated in the model according to the connection structure of the neural network, and is calculated and converted through the neurons in the hidden layer. The neurons in the hidden layer perform weighted summation and nonlinear transformation on the input data according to the preset weights and activation functions, and gradually extract the characteristic patterns in the data. After the prediction result of the model is obtained in the output layer, the error between the prediction result and the actual reception arrangement result is calculated. For example, the difference between the predicted seat layout effect score and the actual score. Then, according to the error, the connection weights between neurons are adjusted from the output layer to the hidden layer and the input layer in sequence through the back propagation algorithm. The back propagation algorithm updates the weights according to a certain learning rate based on the partial derivative of the error with respect to the weight, so that the model can reduce the error in the next prediction. For example, if it is found that the seat layout effect score predicted by the model is higher than the actual score, it means that the current weight setting of the model leads to over-optimism for certain features (such as the weight of the influence of environmental comfort on the seat layout). The relevant weights are reduced through the back propagation algorithm so that the model can more accurately reflect the actual situation.

[0039] Repeat the above process of data input, forward propagation, error calculation and back propagation weight adjustment for multiple iterative training. As the number of iterations increases, the weight of the model is gradually optimized, and the prediction ability of the training samples with target feature data is continuously improved. For example, after 100 iterative trainings, the prediction accuracy of the model for the seat layout effect score increased from the initial 60% to 80%. After a sufficient number of iterative trainings (such as reaching the preset stop condition, such as the error is less than a certain threshold or the number of iterations reaches the upper limit), the weights and parameters of the model have been adjusted to be able to better adapt to the training samples with target feature data. At this time, the model obtained is the target reception management model. This target reception management model can more accurately generate more reasonable reception arrangement suggestions based on the input target feature data (such as environmental comfort, equipment failure rate, meeting importance level, etc.), such as more accurate seat layout recommendations, more complete equipment preparation plans, etc., to meet the diverse needs of corporate multi-functional conference training activities and improve the quality and efficiency of reception management.

[0040] S104: Process the target customer information to generate state parameter information of the target customer.

[0041] In one implementation, the target customer information is processed to generate initial customer image information and target customer identity information, wherein the initial customer image information is the face information of the target customer within a preset time period. In the sign-in phase of the enterprise multifunctional conference training, a high-definition camera and identity recognition equipment are installed. When the target customer (assuming that it is employee Xiao Wang) comes to sign in, the system starts working. The camera continuously captures images within 30 seconds before and after Xiao Wang signs in, and obtains a series of image frames containing Xiao Wang's face. These image frames are the initial customer image information, which records Xiao Wang's facial appearance near the sign-in time, including facial contours, facial features, etc. Xiao Wang uses an employee card to sign in, and the identity recognition device reads the information in the employee card to obtain his name, employee number, department (such as R&D department), position (such as senior engineer) and other identity information, and records relevant data such as sign-in time.

[0042] Process the initial customer image information to generate the influencing factors of the target customer's expression characteristics. Use image processing technology and facial expression recognition algorithms to analyze the initial customer image information. For example, by detecting Xiao Wang's facial features such as the position and shape of his eyebrows (such as whether they are raised or frowned), the degree of eye opening (such as wide eyes indicating possible curiosity or surprise, squinting may indicate confusion or thinking), and the curvature of the corners of his mouth (raised to indicate happiness, drooping to indicate displeasure). At the same time, consider environmental factors when collecting images, such as light intensity and angle. If the light is dim and uneven, it may affect the accuracy of facial expression recognition, and the light intensity and uniformity are recorded as factors affecting expression characteristics. In addition, the background noise level at the check-in site may also affect Xiao Wang's expression. For example, excessive noise may cause him to frown slightly, etc., which is also included in the factors affecting expression characteristics.

[0043] The influencing factors of the target customer's expression characteristics are processed to generate the target customer's emotional information, wherein the target customer's emotional information is used to characterize the target customer's expression change information. The emotional information of Xiao Wang is inferred based on the influencing factors of the expression characteristics obtained by analysis. Assuming that Xiao Wang's eyebrows are slightly raised, his eyes are open normally, his mouth corners are slightly smiling, and the light is sufficient and the environment is quiet, these characteristics can be combined to infer that Xiao Wang's emotions at this time are relatively positive, and he may have expectations for this conference training. If Xiao Wang's eyebrows are slightly wrinkled, his eyes are half-closed, there is no obvious arc at the corners of his mouth, and the light is dim and the noise is loud, it may mean that he is in a slightly anxious or uneasy emotional state, and he may be worried that the content of the conference training is too difficult or disturbed by the environment. In the subsequent conference training process, if Xiao Wang's face image is collected again, his expression changes will be continuously tracked. For example, when the conference training reaches the interactive stage, Xiao Wang actively participates in the discussion, his expression becomes more vivid, his smile increases, and his eyes become brighter. At this time, his emotional information is updated to active involvement, excitement, etc. to reflect his expression changes at different stages.

[0044] Process the target customer's identity information to generate the target customer's real-time status information. Using Xiao Wang's identity information, obtain his real-time health data from the company's internal health management system (assuming that Xiao Wang wears a smart health bracelet and the data is synchronized with the system). If the data shows that his heart rate is normal, his blood pressure is stable, his sleep quality is good, and he has no recent sick leave records, it means that his physical condition is basically good. However, if it is found that his heart rate is slightly higher than the normal range, and he has a record of mild cold symptoms that morning, this will be included as part of the real-time status information, which may affect his participation and attention in the meeting training. Combined with his position and recent work tasks, analyze his work pressure. For example, as a senior engineer, if he is responsible for multiple important projects recently and is approaching the delivery deadline, he has a lot of work pressure, and this information is also included in the real-time status information. At the same time, check his activity level in the company's internal communication tools, the progress of the projects he has recently participated in, etc., and comprehensively judge his work status, such as whether he is busy, efficient, or may need more support.

[0045] The real-time status information of the target customer and the influencing factors of the target customer's facial expression characteristics are processed to generate the target customer's status parameter information, which is used to characterize the target customer's status in the reception scene. The real-time status information of Xiao Wang (such as physical condition, work pressure, etc.) and the influencing factors of facial expression characteristics (such as emotional state) are processed comprehensively. If Xiao Wang is in good physical condition but has high work pressure and his facial expression shows some fatigue, then the status parameter information may indicate that he needs a relatively comfortable and quiet seating environment so that he can better concentrate during the conference training. At the same time, some relaxation and adjustment links may need to be properly arranged in the conference process to relieve his pressure. Based on these comprehensive factors, a set of status parameter information is generated for Xiao Wang, such as using numerical values ​​to indicate his degree of need for a comfortable environment (such as 0.8, indicating a high need), willingness to participate in interactive links (such as 0.6, indicating a moderate willingness, because work pressure may be slightly tired), and tolerance for the difficulty of training content (such as 0.7, indicating that he can withstand a certain degree of difficulty, but if it is too complicated, he may feel strenuous), etc. These status parameter information will be used in the subsequent reception management model to provide Xiao Wang with personalized reception arrangements, such as choosing a seat for him that is close to the air conditioner and relatively quiet, guiding him to participate appropriately in the interactive session, adjusting the pace of explaining the training content according to his tolerance, etc., to improve his experience and effect in the conference training.

[0046] S105, processing the target reception theme information that matches the target customer information to generate reception arrangement influencing factors.

[0047] In one implementation, the target reception theme information that matches the target customer information is processed to generate type information of the target reception theme. For the target reception theme of "Enterprise Digital Transformation Strategic Planning and Implementation Path", the type information is determined by an in-depth analysis of its covered content and goals. This theme focuses on the planning and implementation of digital transformation at the strategic level of the enterprise, involving collaborative decision-making among multiple departments, and aims to promote the adjustment of the enterprise's long-term development strategy. It involves organizational structure changes, business process reshaping, and technological application innovation. Therefore, its type information is determined to be "Enterprise Strategic Planning and Change Management (Digital Transformation Direction)". This type clarifies the core scope of the theme, that is, focusing on the overall strategic layout of the enterprise, using digital means to carry out in-depth changes and innovations, providing a basic framework for subsequent processing.

[0048] The type information of the target reception theme is processed to generate the customer's historical reception arrangement information, where the customer's historical reception arrangement information is used to characterize the reception resource information allocated to different customers when they experience the target reception theme of this type. Based on the determined "Enterprise Strategic Planning and Change Management (Digital Transformation Direction)" type information, the reception arrangement records of past similar theme meetings are queried in the enterprise's conference management system and archives, and the reception situation of different customer groups (such as middle and senior managers, department heads, IT technical backbones, etc.) is summarized and analyzed. The conference table is arranged in a sitting-around style to facilitate face-to-face communication and discussion. A special main seat area is set up for middle and senior managers to highlight their decision-making status. At the same time, ensure that each seat is equipped with a comfortable seat, sufficient lighting, and convenient power and network interfaces to meet the needs of participants to use electronic devices to display materials and take notes. The conference room environment is kept quiet and tidy, with appropriate temperature and humidity to create a good discussion atmosphere. Equipped with a high-definition large-screen projector and an intelligent electronic whiteboard, it is used to display technology trend reports related to digital transformation, industry case analysis PPTs, and digital simulation diagrams of enterprise business processes, etc. Provide each participating group with a high-performance laptop computer pre-installed with data analysis software, project management tools, digital transformation case library and other related software resources, so that data can be easily accessed and analyzed at any time. In addition, ensure that the network in the conference room is high-speed and stable, and support multiple people to access internal enterprise data and external industry resources online at the same time. Prepare detailed paper materials for participants, including enterprise digital transformation status analysis reports, industry digital trend research reports, successful case studies of similar enterprises, and background information and preliminary planning drafts for the topics discussed at this meeting. At the same time, establish a special digital transformation conference material sharing folder on the internal network platform of the enterprise, update the relevant meeting materials in real time, and facilitate participants to download and view the latest information at any time.

[0049] Provide exquisite tea break service, including various types of coffee, tea, fresh fruits, snacks, etc. The tea break time is arranged in the intervals between the discussion of key topics of the meeting, so that participants can relax and replenish their energy appropriately. Lunch is served in the form of a buffet, providing a rich and diverse selection of dishes to meet the needs of different tastes. At the same time, a special dining area is set up to ensure that participants can dine in a comfortable environment. The meal time is arranged reasonably to avoid being too long or too short to affect the progress of the meeting. There are multiple interactive sessions in the meeting, such as group discussions, brainstorming, case studies, etc. In the group discussion session, participants are divided into groups according to their departments and responsibilities to ensure that each group has people from different perspectives to participate and promote cross-departmental exchanges and cooperation. Each group is equipped with a facilitator who is responsible for recording the key points of the discussion and guiding the direction of the discussion. The brainstorming session encourages participants to freely propose creative ideas and thoughts on digital transformation without restrictions to stimulate innovative thinking. The case study session selects representative digital transformation success and failure cases, organizes participants to conduct in-depth analysis and discussion, and draws lessons from them. The reception arrangement information for past meetings with similar themes is summarized to form the customer's historical reception arrangement information, which comprehensively reflects the reception resources and services enjoyed by different customers when participating in corporate strategic planning and change management (digital transformation direction) type meetings.

[0050] The historical reception arrangement information of customers is processed to generate the reception method selection influencing factors and the weight information corresponding to the reception method selection influencing factors. Strategic decision support degree: Since the conference theme involves corporate strategic planning and change, it is necessary to provide sufficient decision support information and a good discussion environment for the middle and senior management personnel attending the conference. Therefore, this factor reflects the effectiveness of the reception arrangement in promoting strategic decision-making. For example, the comprehensiveness of the information, the support ability of the equipment for data display and analysis, and the effect of the interactive link on the collision and integration of different viewpoints all affect the degree of strategic decision support. Technology application display demand: Digital transformation is inseparable from the application of various technologies. The conference needs to display and discuss different digital technologies and their application scenarios in the enterprise. Therefore, factors such as the advanced equipment, the richness of technical resources, and the design of the technical demonstration link in the reception arrangement constitute the influencing factor of technology application display demand. Convenience of cross-departmental collaboration: Participants come from different departments, and one of the important goals of the meeting is to promote cross-departmental collaboration. Whether the seating layout is conducive to communication between personnel from different departments, whether the interactive link helps to break down departmental barriers, and whether the sharing of information and materials is convenient all determine the degree of convenience of cross-departmental collaboration. Knowledge transfer and learning effect: The conference is a platform for knowledge transfer and learning. Participants expect to learn the latest knowledge of digital transformation, industry best practices, and experiences of other departments through the conference. Therefore, factors such as the quality of training materials, the professional level of the instructors, the depth and breadth of case sharing, and learning opportunities in the interactive session jointly affect the factor of knowledge transfer and learning effect.

[0051] The weight of each influencing factor is determined by evaluating the effects of similar meetings in the past (such as the improvement in the quality of enterprise digital transformation strategic decisions after the meeting, the degree of mastery of digital transformation knowledge by participants, the launch and advancement speed of cross-departmental collaboration projects, etc.) and the experience of organizing meetings. After analysis, it is found that the degree of support for strategic decision-making has the greatest impact on the final results of the meeting, and the weight is set to 0.4. Because accurate and effective strategic decisions are the key to the success of enterprise digital transformation, reception arrangements should primarily serve this goal. The weight of the demand for technology application display is 0.25. Digital transformation relies on technological innovation. Good technology display helps participants better understand and apply related technologies. The weight of cross-departmental collaboration convenience is 0.2. Cross-departmental collaboration is an important guarantee for promoting the full implementation of digital transformation in enterprises. Its convenience directly affects the efficiency and effect of collaboration. The weight of knowledge transfer and learning effect is 0.15. Although knowledge transfer is also important, its direct impact on the core goals of the meeting is slightly lower than that of the first three factors.

[0052] Based on the preset reception mode selection influencing feature set, the reception mode selection influencing factors and the weight information corresponding to the reception mode selection influencing factors are processed to generate reception arrangement influencing factors. According to the characteristics and goals of the digital transformation strategy seminar, the enterprise pre-sets the reception mode selection influencing feature set, which includes the following key features and rules: Strategic fit: Evaluate whether the reception arrangement is closely centered on the enterprise's digital transformation strategic goals, including whether the conference content, material preparation, equipment support, etc. are helpful for participants to discuss in depth and formulate a digital transformation plan that conforms to the enterprise strategy. For example, whether the provided cases are consistent with the enterprise's industry characteristics and strategic direction, whether the equipment can meet the needs of displaying and analyzing strategic-related data, etc. Resource synergy: Consider the degree of coordination between reception resources, such as whether the seating layout, equipment, materials, catering and other resources are coordinated with each other to jointly create a good meeting environment and promote the smooth progress of the meeting process. For example, whether the seating layout is convenient for participants to use equipment and obtain materials, and whether the catering arrangement reasonably considers the meeting time and personnel energy requirements. Innovation promotion: Determine whether the reception arrangement is conducive to stimulating the innovative thinking and innovative behavior of participants and promoting innovative practices in digital transformation. For example, whether the design of the interactive session encourages free discussion and creative ideas, whether the technical resources provide the possibility of exploring new digital solutions, etc. Cost-benefit balance: Under the premise of ensuring the quality and effectiveness of the meeting, consider the balance between the cost input of the reception arrangement and the expected benefits. For example, whether the cost of equipment rental and material preparation is reasonable, whether the results of the meeting (such as the improvement of the quality of strategic decision-making, the strengthening of cross-departmental collaboration, etc.) can bring corresponding economic benefits or long-term benefits, etc.

[0053] Based on this preset reception mode selection influencing feature set, the reception mode selection influencing factors and corresponding weight information are comprehensively processed. For the strategic decision support degree influencing factor, it is assumed that in this meeting, based on the meeting preparation and resource allocation, its satisfaction degree is quantified as 0.8 (full score 1.0). Combined with its weight of 0.4, its impact value on the overall reception arrangement is calculated to be 0.4×0.8=0.32. Similarly, for the technology application display demand factor, if the quantitative satisfaction degree is 0.7, its impact value is 0.25×0.7=0.175; the satisfaction degree of the cross-departmental collaboration convenience factor is 0.85, and the impact value is 0.2×0.85=0.17; the satisfaction degree of the knowledge transfer and learning effect factor is 0.75, and the impact value is 0.15×0.75=0.1125.

[0054] These impact values ​​are weighted summed or comprehensively evaluated according to the rules of the preset feature set. For example, according to the importance of each factor in the feature set, the impact value of strategic decision support related to strategic fit is given a higher weight (assuming it is 0.4), the weights of cross-departmental collaboration convenience and knowledge transfer and learning effect related to resource synergy are 0.3 and 0.2 respectively (the sum of the two is 0.5, which constitutes an overall weight distribution of 1.0 with the strategic fit weight), and the weight of the impact value of technology application display demand related to innovation promotion is 0.1. Reception arrangement impact factor = 0.4 × 0.32 + 0.3 × 0.17 + 0.2 × 0.1125 + 0.1 × 0.175 = 0.128 + 0.051 + 0.0225 + 0.0175 = 0.219. This reception arrangement impact factor will be used as an important indicator to guide the optimization of reception arrangements for this digital transformation strategy seminar. For example, if the impact factor of reception arrangements is low, it may be necessary to re-examine whether the seating layout is conducive to strategic discussions, whether the technical equipment meets the display requirements, and whether the interactive sessions promote cross-departmental collaboration and knowledge transfer. This will help adjust the reception method, improve the meeting effect, and ensure that the meeting can provide strong support for the strategic planning and implementation of the company's digital transformation.

[0055] S106: Process the reception venue information based on the target reception management model to generate attribute information of the reception venue.

[0056] In one implementation, the reception venue information is processed based on the target reception management model to generate monitoring point information, wherein the monitoring point information is the seat or reception area information located in the reception venue. Assume that the reception venue is a multifunctional conference center of a large enterprise, which is used to hold digital transformation strategy seminars. The conference center covers an area of ​​1,000 square meters and has different types of seats and reception areas, including a podium, an auditorium, a group discussion area, etc. According to the target reception management model, the reception venue information is processed to generate monitoring point information. For example, each seat and important reception area (such as the center of the group discussion area) in the conference center are set as monitoring points. The monitoring point information specifically includes the number of each seat, the location coordinates (such as the row and column of the conference center) and the basic characteristics of the seat (such as seat comfort, whether it is close to a power socket, etc.). For the reception area, the monitoring point information includes the area name, the location range (such as the area boundary expressed in coordinate form) and the functional characteristics of the area (such as the group discussion area has movable tables and chairs, whiteboards and other facilities).

[0057] The monitoring point information is processed in groups to generate several group information, wherein the group information includes the monitoring point information at different distances from the reception core area. The monitoring point information is processed in groups according to the different distances from the reception core area (assuming that the rostrum or the main speech area is the reception core area). It can be divided into three groups: the close distance group (monitoring points within 0-5 meters from the core area), the medium distance group (monitoring points within 5-10 meters) and the long distance group (monitoring points within a range of more than 10 meters). The close distance group: may include the front row VIP seats, the recorder seats near the rostrum, etc. These monitoring points usually have better vision and hearing effects, and can interact with the speaker more directly. The medium distance group: mainly the middle part of the auditorium, which is neither too close to the core area to be disturbed, nor too far to affect the viewing effect. The monitoring points in this group are more balanced in vision and hearing, and are suitable for most participants. The long distance group: may be the edge area, corner position or back row seats of the conference center. These monitoring points are relatively far away from the core area and may be affected by certain line of sight obstruction or sound attenuation, but they may also be quieter and suitable for participants who like to think independently or do not need to interact with speakers frequently.

[0058] The group information is processed with features to generate feature data sets of different groups. The monitoring point information of each group is processed with features. For example, for the close-distance group: Seat comfort features: Extract information such as the material, hardness, and backrest angle of the seat to evaluate the comfort of the seat. If the seat is made of high-grade leather, has moderate hardness, and the backrest angle is adjustable, the comfort is high. Visual field effect features: Analyze factors such as the angle between the monitoring point and the rostrum and whether there are any obstructions to determine the visual field effect. If there is no obstruction and the angle with the rostrum is moderate, the visual field effect is good. Interactive convenience features: Consider the distance between the monitoring point and the speaker and whether there are microphones and other equipment to evaluate the interactive convenience. If the distance is close and equipped with a microphone, the interactive convenience is high. Close-distance group feature data set: Contains multiple feature vectors, such as [high comfort, good visual effect, high interactive convenience], etc., reflecting the characteristics of the monitoring point in this group. Medium-distance group feature data set: May contain feature vectors such as [medium comfort, good visual effect, medium interactive convenience]. Long-distance group feature data set: for example, feature vectors such as [lower comfort, general visual effect, low interaction convenience].

[0059] The feature data sets of different groups are processed to obtain the abnormal reception area features and correlation co-occurrence frequencies. The feature data sets of different groups are processed to determine the abnormal reception area features by comparing the actual monitoring data with the expected standards. For example, if the seat damage at a certain monitoring point causes a significant decrease in comfort, or the noise in a certain area suddenly increases and affects the hearing effect, these are all abnormal features. At the same time, the co-occurrence frequencies of different feature abnormalities are counted, that is, the probability that certain features are abnormal at the same time. For example, in the long-distance group, the frequency of low seat comfort and general visual effect occurring at the same time is high, which is a correlation co-occurrence frequency. By analyzing these abnormal reception area features and correlation co-occurrence frequencies, we can understand the places where problems are prone to occur in different groups and the correlation between the problems. For example, if it is found that the co-occurrence of unstable network signals and insufficient light brightness often occurs in the medium-distance group, it means that these two problems may have a certain correlation, which may be caused by reasons such as equipment layout or power supply in the area.

[0060] The abnormal reception area characteristics and the correlation co-occurrence frequency are processed to generate correlation information, wherein the correlation information is used to characterize the correlation between the reception area characteristics at different distances from the reception core area and the reception situation at the reception venue. The abnormal reception area characteristics and the correlation co-occurrence frequency are processed to generate correlation information. For example, the correlation between the reception area characteristics at different distances from the reception core area and the reception situation at the reception venue is analyzed. If it is found that the interactive convenience characteristics of the close-distance group are strongly positively correlated with the discussion activity of the meeting, it means that the close-distance seating arrangement plays an important role in promoting the active participation of participants in the discussion. The quietness characteristics of the long-distance group are positively correlated with the satisfaction of participants who need to think independently, indicating that the long-distance area may be more suitable for specific types of participants. This correlation information can help conference organizers better understand the characteristics and applicable populations of different areas, so as to be more scientific and reasonable in arranging seats and allocating resources. For example, for an important digital transformation strategy seminar, those middle and senior managers who actively speak and like to interact can be arranged in the close-distance group based on the correlation information so that they can better participate in the discussion; some IT technical backbones who need to think quietly can be arranged in the long-distance group so that they can focus on analyzing problems.

[0061] Based on the correlation information, the feature data sets of different groups are processed to generate the attribute information of the reception venue, and the attribute information of the reception venue includes the classification information of the seats. Based on the correlation information, the feature data sets of different groups are processed to generate the attribute information of the reception venue. For example, the seats are classified according to the correlation between different groups and the reception situation. Interactive seats: mainly distributed in the close-distance group, with good comfort and interactive convenience, suitable for participants who actively participate in the discussion. Balanced seats: located in the medium-distance group, relatively balanced in comfort, visual effect and interactive convenience, suitable for most participants. Thinking seats: arranged in the long-distance group, relatively quiet, may be slightly less comfortable, but suitable for participants who need to think independently. These attribute information can provide conference organizers with more detailed reception venue planning basis. In future conference arrangements, different types of seats can be reasonably selected and allocated according to different conference themes and participants' needs to improve the utilization efficiency of reception venues and the satisfaction of participants.

[0062] S107 , processing the attribute information of the reception place, the state parameter information of the target customer and the reception arrangement influencing factor based on the target reception management model to generate reception management information of the reception place.

[0063] In one implementation, the attribute information of the reception venue is processed based on the target reception management model to generate a first reception customization value of the target reception area. The target reception management model also includes a calculation formula for obtaining the first reception customization value, and the calculation formula includes: ; ;in, represents the first reception customization value; m represents the number of reception area groups in the reception venue; represents the number of reception units (such as seats, display areas, etc.) in the i-th group; Represents the first The pre-arranged value of reception units in the group; k represents the number of reception unit features in each group; Representative In the group The weight of each feature; Represents the first The reception unit is The value of a feature; Used to adjust the jth feature weight in the i-th group The degree of influence of Used to adjust the first The reception unit is The value of the feature degree of impact.

[0064] Suppose our reception venue is a digital transformation seminar venue for an enterprise. The venue is divided into several area groups according to different functions. Suppose it is divided into three area groups (m=3), namely the core discussion area, the general audience area and the independent thinking area. The core discussion area has 8 round discussion tables as reception units ( =8), the general audience area has 50 seats ( =50), the independent thinking area has 12 single workstations ( =12). Consider five characteristics (k=5), namely, technical equipment completeness ( )、Space Comfort( ), ease of interaction with speakers ( ), environmental quietness ( ) and visual field width ( In the core discussion area, the weight of technical equipment completeness ( ) is relatively high, at 0.3; the spatial comfort weight ( ) is 0.2; the weight of the convenience of interacting with the speaker ( ) is 0.25; the weight of environmental quietness ( ) is 0.1; the weight of the visual field width ( ) is 0.15. At the same time, assume that the parameters α=2, β=1.5.

[0065] Calculate the pre-scheduled value of a single reception unit. Take a circular discussion table in the core discussion area as an example. The characteristics of this reception unit are evaluated as follows: The completeness of technical equipment is very high ( =9 points, full score 10 points), good space comfort ( =8 points, full score 10 points), easy to interact with the speaker ( =8 points, full score 10 points), the environment is moderately quiet ( =7 points, full score 10 points), good field of vision ( =8 points out of 10).

[0066] According to the formula , we can get the pre-arranged value of the discussion table in the core discussion area: , the specific value can be obtained by calculation.

[0067] According to the formula , calculate the first reception customization value of the entire seminar venue, first calculate the sum of the pre-arranged values ​​of all reception units in each area, that is, the sum of the core discussion area is , the total number of general audience areas is , the sum of independent thinking areas is Then add the sum of the three regions to get the value of the numerator S = + + , the denominator is the total number of reception units n= + + The final calculation , get the first reception customization value.

[0068] Based on the target reception management model, the target customer's status parameter information and reception arrangement influencing factors are processed to generate the second reception customization value for the target reception area. Based on the target reception management model, the target customer's status parameter information and reception arrangement influencing factors are processed to generate the second reception customization value for the target reception area. Target customer status parameter information: Assume that the target customers are the company's middle and senior management and IT department technical backbones who attend the seminar. Their status parameter information may include a high degree of demand for technical depth, high expectations for interactive communication, etc. Determine the reception arrangement influencing factors by considering factors such as the importance of the seminar and the professional level of the participants. For example, since this seminar is a key meeting for the company's digital transformation, its importance is high and the impact factor is correspondingly large.

[0069] Assumptions , According to the formula ,in, Represents the second reception customization value; The factors that influence the reception of the target customer's status parameter information; Factors affecting the arrangements for reception of delegates; Represents the weight of state parameter information; Represents the weight of the factors affecting reception arrangements; parameter Used to adjust the weight of state parameter information and reception arrangement impact factor weight The degree of influence of Used to adjust the impact factor of target customer status parameter information on reception and reception arrangement impact factor Assuming γ = 1.5, δ = 2, the influence factor of the target customer's state parameter information on reception is , reception arrangement impact factor .but , the second reception customization value can be obtained by calculation.

[0070] The first reception customization value of the target reception area and the second reception customization value of the target reception area are processed to generate reception management information of the reception venue. The first reception customization value of the target reception area and the second reception customization value of the target reception area are processed to generate reception management information of the reception venue. Combining the first reception customization value and the second reception customization value, it is possible to fully understand the various aspects of the reception venue and the degree of matching with the needs of the target customers. For example, if the first reception customization value shows that the technical equipment completeness of a certain area is high, but the second reception customization value shows that the target customers have higher expectations for the convenience of interaction, then in the reception management, it is possible to consider adjusting the layout of the area, adding interactive equipment or optimizing the settings of the interactive links. Determine the resource allocation strategy based on the customization value. For example, for areas with higher first reception customization values, more technical support personnel and high-quality service resources can be allocated; for areas where the second reception customization value shows that customer needs are concentrated, such as specific catering needs or rest area settings, resources can be invested and optimized in a targeted manner. Provide personalized reception services based on customization values. For example, according to the status parameter information of the target customers, arrange seating areas that better meet their needs or provide special technical explanation services for specific customer groups. At the same time, according to the first reception customization value of the reception venue, the overall environment is optimized to ensure that each area can provide customers with a good experience. Through the above steps, using the first reception customization value, the second reception customization value and the final reception management information generated by the target reception management model, a scientific and accurate reception plan can be provided for activities such as corporate digital transformation seminars to improve reception quality and customer satisfaction.

[0071] This application focuses on corporate reception management, aiming to improve reception quality and customer satisfaction. First, collect various information, including target customer information, target reception theme information, reception venue information, training sample sets, and preset reception management models. Among them, the reception venue information covers the location, the training sample set includes the historical environmental factors of the reception venue and the real-time environmental factors of the adjacent venues, and the preset model is generated based on the historical data of other venues. Next, preprocess the training sample set, establish the original feature library through feature extraction, and divide the training set and the validation set. Use the classifier and the preset algorithm to predict the validation set and train the training set respectively to generate target feature data to characterize the risk factors of seat pre-arrangement. Afterwards, use the sample set with target feature data to process the preset model to obtain the target reception management model. At the same time, the target customer information is gradually processed to generate the initial image and identity information, and then the influencing factors of the expression features, customer emotional information and real-time status information are introduced to comprehensively obtain the status parameter information.

[0072] For the target reception theme information, the generation type information, customer historical reception arrangement information, influencing factors and weight information are processed in turn, and finally the reception arrangement influencing factors are generated based on the preset feature set. When processing the reception venue information, the monitoring point information is first generated, and the information of different distances is obtained after grouping. The features are extracted to generate a data set, and the data set is processed to obtain abnormal features and correlation co-occurrence frequencies, and then the correlation information is generated, and finally the reception venue attribute information and seat classification are determined. Finally, based on the target reception management model, various information is processed to generate the first and second reception customization values, and then the reception management information of the reception venue is obtained. This method comprehensively considers various factors, realizes scientific planning and personalized services, and is of great significance to improving the level and efficiency of enterprise reception management.

[0073] In one embodiment, if Figure 2 As shown, the present application also provides a smart reception management device, including: The acquisition module 201 is used to acquire target customer information, target reception theme information matching the target customer information, reception venue information and a training sample set, wherein the reception venue information includes location information within the reception venue, and the training sample set includes historical environmental factor information of the reception venue and real-time environmental factor information of other venues adjacent to the reception venue; acquire a preset reception management model, wherein the preset reception management model is generated based on historical reception environmental factor information of other reception venues; The processing module 202 is used to pre-process the training sample set to generate a training sample set with target feature data, wherein the target feature data is used to characterize the risk factors that affect the pre-arrangement of seats at the reception venue; based on the training sample set with target feature data, the preset reception management model is processed to generate a target reception management model; the target customer information is processed to generate state parameter information of the target customer, wherein the state parameter information of the target customer is used to characterize the state of the target customer in the reception scenario; the target reception theme information that matches the target customer information is processed to generate a reception arrangement influencing factor; based on the target reception management model, the reception venue information is processed to generate attribute information of the reception venue, wherein the attribute information of the reception venue includes classification information of seats; based on the target reception management model, the attribute information of the reception venue, the state parameter information of the target customer and the reception arrangement influencing factor are processed to generate the reception management information of the reception venue.

[0074] The computer-readable storage medium provided in the above-mentioned embodiments of the present application and the intelligent reception management method provided in the embodiments of the present application 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.

[0075] Each embodiment in the present application is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the management method, electronic device, electronic device, and readable storage medium embodiment for evaluating intelligent reception, since they are basically similar to the management method embodiment of intelligent reception described above, the description is relatively simple, and the relevant parts can be referred to the partial description of the management method embodiment of intelligent reception described above.

[0076] Although the present application is disclosed as above, the present application is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present application. Therefore, the protection scope of the present application shall be subject to the scope defined by the claims.

Claims

1. A method for managing intelligent reception, characterized in that: include: Obtaining target customer information, target reception theme information matching the target customer information, reception venue information, a training sample set, and a preset reception management model, wherein the reception venue information includes location information within the reception venue, the training sample set includes historical environmental factor information of the reception venue and real-time environmental factor information of other venues adjacent to the reception venue, and the preset reception management model is generated based on the historical reception environmental factor information of other reception venues; Preprocessing the training sample set to generate a training sample set with target feature data, wherein the target feature data is used to characterize risk factors that affect the pre-arrangement of seats at a reception venue; Processing the preset reception management model based on the training sample set with target feature data to generate a target reception management model; Processing the target customer information to generate state parameter information of the target customer, wherein the state parameter information of the target customer is used to characterize the state of the target customer in the reception scenario; Process the target reception theme information that matches the target customer information to generate reception arrangement influencing factors; Processing the reception venue information based on the target reception management model to generate attribute information of the reception venue, wherein the attribute information of the reception venue includes classification information of seats; Based on the target reception management model, the attribute information of the reception venue, the status parameter information of the target customer and the reception arrangement influencing factor are processed to generate reception management information of the reception venue.

2. The method according to claim 1, characterized in that Preprocessing the training sample set to generate a training sample set with target feature data includes: Extract features from the training sample set to generate an original feature library; Processing the original feature library to generate a training set and a verification set; Based on the classifier, each validation set divided by the original feature library is predicted to generate a prediction result; Based on the preset algorithm, each training set divided by the original feature library is trained to generate the prediction results of the validation set class; The prediction results and the validation set prediction results are processed to generate target feature data, where the target feature data is used to characterize risk factors that affect the advance arrangement of seats at reception venues.

3. The method according to claim 1, characterized in that The target customer information is processed to generate state parameter information of the target customer, including: Processing the target customer information to generate initial customer image information and target customer identity information, wherein the initial customer image information is face information of the target customer within a preset time period; Processing the initial customer image information to generate influencing factors of the target customer's facial expression characteristics; Processing the influencing factors of the target customer's expression characteristics to generate the target customer's emotion information, wherein the target customer's emotion information is used to represent the target customer's expression change information; Processing the identity information of the target customer to generate real-time status information of the target customer; The real-time status information of the target customer and the influencing factors of the target customer's expression characteristics are processed to generate status parameter information of the target customer.

4. The method according to claim 3, characterized in that Process the target reception theme information that matches the target customer information to generate reception arrangement influencing factors, including: Processing the target reception theme information that matches the target customer information to generate type information of the target reception theme; Processing the type information of the target reception theme to generate customer historical reception arrangement information, wherein the customer historical reception arrangement information is used to represent the reception resource information allocated to different customers when they experience the target reception theme of this type; Processing the historical reception arrangement information of the customer to generate reception method selection influencing factors and weight information corresponding to the reception method selection influencing factors; The reception mode selection influencing factor and the weight information corresponding to the reception mode selection influencing factor are processed based on a preset reception mode selection influencing feature set to generate a reception arrangement influencing factor.

5. The method according to claim 4, characterized in that The reception venue information is processed based on the target reception management model to generate attribute information of the reception venue, including: Processing the reception venue information based on the target reception management model to generate monitoring point information, wherein the monitoring point information is seat or reception area information located in the reception venue; The monitoring point information is grouped to generate a plurality of group information, wherein the group information includes monitoring point information at different distances from the reception core area; Perform feature extraction on the grouping information to generate feature data sets for different groups; Process the feature data sets of different groups to obtain the abnormal reception area features and related co-occurrence frequencies; Processing the abnormal reception area features and the correlation co-occurrence frequency to generate correlation information, wherein the correlation information is used to characterize the correlation between reception area features at different distances from the reception core area and the actual reception situation at the reception venue; The characteristic data sets of different groups are processed based on the correlation information to generate attribute information of the reception venue.

6. The method according to claim 5, characterized in that The target reception management model is used to process the attribute information of the reception venue, the state parameter information of the target customer and the reception arrangement influencing factor to generate reception management information of the reception venue, including: Processing the attribute information of the reception venue based on the target reception management model to generate a first reception customization value of the target reception area; Processing the state parameter information of the target customer and the reception arrangement influencing factor based on the target reception management model to generate a second reception customization value for the target reception area; The first reception customization value of the target reception area and the second reception customization value of the target reception area are processed to generate reception management information of the reception place.

7. The method according to claim 6, characterized in that The attribute information of the reception place, the state parameter information of the target customer and the reception arrangement influencing factor are processed based on the target reception management model to generate reception management information of the reception place, further comprising: The target reception management model also includes a calculation formula for obtaining the first reception customization value, and the calculation formula includes: ； ； in, represents the first reception customization value; m represents the number of reception area groups in the reception venue; represents the number of reception units in the i-th group; Represents the first The pre-arranged value of reception units in the group; k represents the number of reception unit features in each group; Representative In the group The weight of each feature; Represents the first The reception unit is The value of a feature; parameter Used to adjust the weight of the jth feature in the i-th group The degree of influence of Used to adjust the first The reception unit is The value of the feature the extent of the impact; The target reception management model also includes a calculation formula for obtaining the second reception customization value, and the calculation formula includes: ； in, Represents the second reception customization value; The factors that influence the reception of the target customer's status parameter information; Factors affecting the arrangements for reception of delegates; Represents the weight of state parameter information; Represents the weight of the factors affecting reception arrangements; parameter Used to adjust the weight of state parameter information and reception arrangement impact factor weight The degree of influence of Used to adjust the impact factor of target customer status parameter information on reception and reception arrangement impact factor degree of impact.

8. An intelligent reception management device, characterized in that: The device comprises: An acquisition module is used to acquire target customer information, target reception theme information matching the target customer information, reception venue information and a training sample set, wherein the reception venue information includes location information within the reception venue, and the training sample set includes historical environmental factor information of the reception venue and real-time environmental factor information of other venues adjacent to the reception venue; acquire a preset reception management model, wherein the preset reception management model is generated based on historical reception environmental factor information of other reception venues; A processing module is used to pre-process the training sample set to generate a training sample set with target feature data, wherein the target feature data is used to characterize the risk factors that affect the pre-arrangement of seats at a reception venue; based on the training sample set with target feature data, the preset reception management model is processed to generate a target reception management model; the target customer information is processed to generate state parameter information of the target customer, wherein the state parameter information of the target customer is used to characterize the state of the target customer in the reception scenario; the target reception theme information that matches the target customer information is processed to generate a reception arrangement influencing factor; based on the target reception management model, the reception venue information is processed to generate attribute information of the reception venue, wherein the attribute information of the reception venue includes classification information of seats; based on the target reception management model, the attribute information of the reception venue, the state parameter information of the target customer and the reception arrangement influencing factor are processed to generate reception management information of the reception venue.

9. An electronic device, characterized in that: include: a first processor; and a memory for storing executable instructions of the first processor; Wherein, the first processor is configured to execute the intelligent reception management method described in any one of claims 1 to 7 by executing the executable instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the second processor, the intelligent reception management method described in any one of claims 1 to 7 is implemented.

Images