Portrayal generation method and device, computer readable storage medium, computer program product and computing equipment

Abstract

The invention provides a portrait generation method and apparatus, a computer readable storage medium, a computer program product and a computing device. The portrait generation method comprises the steps of obtaining multi-source data for a user; according to the multi-source data, determining a common label and a characteristic label of the user, the common label being used for indicating inherent information of the user, the characteristic label being used for indicating a personalized behavior of the user, and the characteristic label comprising at least one of the following items: a behavior characteristic label, a consumption characteristic label, a social characteristic label and a care characteristic label; and constructing an owner portrait of the user based on the general label and the characteristic label. According to the scheme, the proprietor portrait can be accurately depicted, so that property service improvement is assisted.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular, to an image generation method and device, a computer-readable storage medium, a computer program product, and a computing device. Background Art

[0002] In the traditional property management mode, service personnel usually rely on experience and manual judgment to handle owner-related affairs.

[0003] Although there are some basic property information management platforms and customer relationship management (CRM) systems on the market currently, the above platforms or systems mainly stay at the level of simple data recording and retrieval, with limited ability to deeply analyze and infer owner data, and it is difficult to accurately depict the owner portrait. Summary of the Invention

[0004] The present application provides a solution that can accurately depict the owner portrait to assist in improving property services.

[0005] To achieve the above object, the present application provides the following technical solutions:

[0006] In a first aspect, an image generation method is provided. The image generation method includes: obtaining multi-source data for a user; determining a common label and a characteristic label for the user according to the multi-source data, where the common label is used to indicate the inherent information of the user, and the characteristic label is used to indicate the personalized behavior of the user, and the characteristic label includes at least one of the following: a behavior characteristic label, a consumption characteristic label, a social characteristic label, and a care characteristic label; constructing an owner portrait of the user based on the common label and the characteristic label.

[0007] Optionally, the obtaining multi-source data includes at least one of the following steps: collecting basic information of the user; obtaining instant messaging messages of the user; capturing at least one of the user's property payment behavior information, complaint information, repair information, and activity participation information.

[0008] Optionally, the common label may include at least one of the following: an owner identity label, a housing and asset label, and a living characteristic label.

[0009] Optionally, the behavior characteristic label may include at least one of the following fields: payment behavior, payment cycle, payment method, arrears risk, and emotional expression; the consumption characteristic label includes at least one of the following fields: value-added service consumption habit, service type preference; the social characteristic label includes at least one of the following fields: community participation degree, activity preference, and instant messaging activity; the care characteristic label includes at least one of the following fields: special needs, life stage, and birthday or festival care.

[0010] Optionally, determining the common labels and characteristic labels of the user according to the multi-source data includes: matching the multi-source data with the keywords corresponding to each common label in the common label library to determine the common labels of the user; generating the characteristic labels of the user according to the multi-source data.

[0011] Optionally, generating the characteristic labels of the user according to the multi-source data includes: inputting the multi-source data into a pre-trained large language model to output the characteristic labels of the user, and the large language model is trained with training data, and the training data includes each multi-source data and its corresponding characteristic labels.

[0012] Optionally, the portrait generation method further includes: obtaining service data for the property; generating a property portrait according to the service data and the multi-source data, where the property portrait includes at least one of the following labels: service efficiency, service quality, equipment status, team status, and the service efficiency is used to represent the completion degree of property tasks, the service quality is used to represent the service response speed, the equipment status is used to represent the equipment maintenance condition, and the team status is used to represent the working condition of property personnel.

[0013] Optionally, after obtaining the multi-source data, it further includes: determining whether the multi-source data is missing; outputting a first prompt message for prompting supplementary content.

[0014] In a second aspect, the present application also discloses a portrait generation device, which includes: an acquisition module for acquiring multi-source data for a user; a label generation module for determining the common labels and characteristic labels of the user according to the multi-source data, where the common labels are used to indicate the inherent information of the user, and the characteristic labels are used to indicate the personalized behavior of the user, and the characteristic labels include at least one of the following: behavior characteristic labels, consumption characteristic labels, social characteristic labels, care characteristic labels; a portrait generation module for constructing the user's owner portrait based on the common labels and the characteristic labels.

[0015] In a third aspect, there is provided a computer-readable storage medium, on which a computer program is stored, and the computer program is run by a processing module to execute any method provided in the first aspect.

[0016] In a fourth aspect, there is provided a portrait generation device, including a storage module and a processing module, where a computer program that can run on the processing module is stored on the storage module, and the processing module runs the computer program to execute any method provided in the first aspect.

[0017] Fifth aspect, a computer program product is provided, on which a computer program is stored, and the computer program is run by a processing module to execute any one of the methods provided by the first aspect.

[0018] Sixth aspect, an embodiment of the present application further provides a chip, on which a computer program is stored, and when the computer program is executed by the chip, the steps of the above method are implemented.

[0019] Seventh aspect, an embodiment of the present application further provides a system chip, which is applied to a terminal. The system chip includes at least one processing module and an interface circuit. The interface circuit and the at least one processing module are interconnected by a line, and the at least one processing module is configured to execute instructions to execute any one of the methods provided by the first aspect.

[0020] Compared with the prior art, the technical solution of the present application has the following beneficial effects:

[0021] In the technical solution of the present application, multi-source data for a user is obtained; a common label and a characteristic label of the user are determined according to the multi-source data. The common label is used to indicate the inherent information of the user, and the characteristic label is used to indicate the personalized behavior of the user; an owner portrait of the user is constructed based on the common label and the characteristic label. Through the integration and analysis of multi-source data, the technical solution of the present application generates a common label and a characteristic label to construct an owner portrait. The characteristic label can be dynamically extended and updated in real time, and can also be continuously iterated and automatically updated according to data changes and owner feedback, so that the owner portrait can be more flexible and accurate, thus laying a foundation for the subsequent property to provide better property services.

[0022] Furthermore, the technical solution of the present application can also construct a property portrait. By setting corresponding labels: service efficiency, service quality, equipment status, and team status, it is possible to comprehensively and accurately depict property services from multiple dimensions and provide assistance for the improvement of property services.

[0023] Furthermore, the technical solution of the present application realizes a complete closed loop from data cleaning to portrait generation, from portrait to policy recommendation, and then from policy execution feedback back to portrait optimization. By automatically identifying information gaps, guiding the property to supplement necessary data, and updating the portrait and policy accordingly, the accuracy and timeliness of personalized services can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a flowchart of a portrait generation method provided by an embodiment of the present application;

[0025] Figure 2 is a schematic diagram of an owner portrait provided by an embodiment of the present application;

[0026] Figure 3It is a flowchart of a portrait generation method provided by an embodiment of the present application;

[0027] Figure 4 It is a schematic diagram of a property portrait provided by an embodiment of the present application;

[0028] Figure 5 It is a schematic structural diagram of a communication device provided by an embodiment of the present application. Detailed implementation manners

[0029] As described in the background art, the deficiencies of the existing property service platforms or systems are as follows:

[0030] The data analysis granularity is relatively coarse: only basic information is recorded, lacking in-depth data mining and portrait analysis;

[0031] Lack of intelligent judgment: Even with information systems, most are processed and judged manually, unable to quickly output targeted service strategy suggestions.

[0032] The technical solution of the present application integrates and analyzes multi-source data to generate common tags and characteristic tags to construct an owner portrait. The characteristic tags can be dynamically expanded and updated in real time, and can also be continuously iterated and automatically updated according to data changes and owner feedback, so that the owner portrait can be more flexible and accurate, thus laying a foundation for providing better property services in the future. Further, a complete and accurate owner portrait can provide precise and high-value service suggestions and decision-making references for property management parties, fundamentally solving the problems of scattered information, lack of in-depth analysis and personalized services in traditional property management.

[0033] In addition, the technical solution of the present application extracts feature information from a large amount of owner-related data by using big data and artificial intelligence technologies, and generates an owner portrait through a large model, comprehensively judging information such as the living habits, demand preferences, payment behaviors, and family situations of the owners, so as to provide targeted and operable decision-making references for the property. This intelligent processing method not only improves the service accuracy, but also provides a solid data foundation for formulating subsequent property operation strategies.

[0034] The user data involved in the technical solution of the present application are all obtained with user authorization. For example, the purpose, scope, and permissions of the data are clearly informed to the user through an interactive interface, and after the user actively checks the consent or completes the electronic signature, valid authorization is obtained.

[0035] To make the above objects, features, and advantages of the present application more obvious and understandable, the following detailed description of the specific embodiments of the present application will be given with reference to the accompanying drawings.

[0036] An embodiment of the present application provides a portrait generation method, referring to Figure 1, which will be described in detail through the following specific steps.

[0037] It can be understood that, in a specific implementation, the portrait generation method can be implemented in the form of a software program, and the software program runs in a processing module integrated inside a chip or a chip module. This method can also be implemented in a way that combines software with hardware, and this application does not make any restrictions.

[0038] It should be noted that the sequence numbers of each step in this embodiment do not represent the limitation of the execution sequence of each step.

[0039] Step 101: Obtain multi-source data for the user.

[0040] In a specific implementation, the multi-source data can be data from multiple sources, and both the form and content of the data can be different. For example, the multi-source data is data from a property management system, chat information of an instant messaging tool, information manually collected by property management staff, etc.

[0041] In a specific embodiment, the multi-source data may include at least one of the following: basic information of the user, instant messaging messages, property payment behavior information, complaint information, repair information, activity participation information.

[0042] Among them, the basic information of the user may include the user's personal basic information and the user's family relationship information. For example, the user's personal basic information includes name, mobile phone number, age, height, occupation, etc. Also, for example, the family relationship information may include the number of people in the owner's family, the names of each family member, whether there are elderly people or children, etc. Specifically, property management staff can obtain the basic information of the owner by visiting the owner and upload and store it in the property information management platform. Then, in the specific implementation of step 101, the basic information of the owner can be directly retrieved from the property information management platform.

[0043] Among them, the instant messaging messages of the user can be messages sent or received by the user in an instant messaging software, and this message can be used to construct an owner portrait. For example, in the scenario of providing property services for the user, property management staff will establish a WeChat group for the owner, and the owner can send information in the WeChat group.

[0044] Among them, the property payment behavior information of the user indicates whether the user pays the property fee, when the user pays the property fee, and the amount of the property fee paid.

[0045] Among them, the complaint information of the user indicates whether the user complains about the property service and the content of the complaint.

[0046] Among them, the repair information of the user indicates whether the user repairs the equipment, the content of the repair, and the feedback of the repair.

[0047] Among them, the user's activity participation information includes whether the user participates in the activities organized by the property, the frequency of participation, the time of participation, and the content of the activities participated in.

[0048] Specifically, instant messaging messages (such as chat records) in instant messaging tools (such as WeChat groups) belong to unstructured data. It is necessary to adopt a suitable data collection method to convert the instant messaging messages into structured data, and then perform text analysis, classification modeling, etc.

[0049] Exemplarily, keywords in instant messaging messages can be crawled by performing optical character recognition (OCR) on screenshots, such as activity registration, complaint content, property feedback, etc.

[0050] Exemplarily, multi-source data can be obtained by performing keyword extraction, sentiment analysis, classification annotation, etc. on the locally exported chat records through text parsing.

[0051] Exemplarily, WeChat robots can also be used to monitor WeChat group messages and extract key data.

[0052] It should be noted that obtaining instant messaging messages in the embodiments of the present application is authorized by the user.

[0053] In a specific embodiment, activity participation information can be extracted from instant messaging messages, such as information on activity registration, time, location, participants, etc. Specifically, the above data can be extracted through keyword matching (such as term frequency-inverse document frequency (TF-IDF), latent Dirichlet allocation (LDA) topic model), regular expressions (such as matching "activity time", "registration"), and intent recognition models (bidirectional encoder representations from transformers (BERT), fastText model).

[0054] In a specific embodiment, complaint information can be extracted from instant messaging messages, such as complaint content, complaint type (facility / environment / service), emotional state, etc. Exemplarily, complaint emotions (rational communication / easy to get emotional) can be identified through sentiment analysis, such as Long Short-Term Memory (LSTM), BERT. Exemplarily, complaint categories can be automatically classified through text classification, such as TF-IDF combined with Support Vector Machine (SVM). Exemplarily, key information can be extracted through keyword matching, such as regular expressions, such as elevator breakdown, noise problem, etc.

[0055] Furthermore, multi-source data can be stored in a database. Even further, the multi-source data can be standardized and the standardized data can be stored in the database.

[0056] In a specific embodiment, the following method is used to standardize the data: First, perform data cleaning on the multi-source data, then perform field standardization, unified encoding, standardized data structure, and finally unified data format.

[0057] Noise data can be removed through data cleaning. Specifically, duplicate records can be deleted, for example, the same repair order is submitted by the same owner multiple times. Specifically, incorrect data can be excluded, for example, the payment amount is negative.

[0058] Missing values can be processed through data cleaning. Exemplarily, for payment data, if the payment record for a certain month is missing, historical mean interpolation can be used for filling. For complaint data, if the specific description is missing, the default category (such as "other") can be used. For example, if key fields (such as owner name, house number) are missing, try to complete them through other information. For non-key fields (such as community activity participation records), the default value can be selected for filling or marked as "unknown".

[0059] Outliers can be corrected through data cleaning. Specifically, data beyond a reasonable range can be identified, such as "monthly property fee exceeds 10,000 yuan", and the source needs to be checked and corrected.

[0060] Through field standardization, field naming and field format can be unified. For example, unified use of Chinese and English naming. For example, "owner name" is unified as owner_name, and "house number" is unified as unit_number. Date fields are unified in the ISO8601 format (such as 2025-01-14). Amount fields are unified in the decimal(10,2) format, retaining two decimal places. Text fields remove extra spaces or special characters.

[0061] By means of unified coding, it is possible to represent key entities using standardized characters. For example, the property building code is COMMUNITY01 - BUILDING03. Fixed values can also be managed by establishing a dictionary table. For example: payment cycle: 1 = monthly payment, 2 = quarterly payment, 3 = annual payment; payment status: 1 = on time, 2 = overdue. For example, numerical data such as payment amount, overdue days, complaint frequency, etc. are normalized; data such as owner type, payment method, community participation degree, etc. are one - hot encoded.

[0062] Through a standardized data structure, data can be recorded and presented more clearly. Specifically, it is designed by dividing tables according to functional modules. For example, design the following multiple tables: owner information table, payment record table, complaint record table. Ensure that the master - slave relationship is clear. For example: owner table: master table (owner_id), payment record table: slave table (payment_id).

[0063] By unifying the data format, data can be represented clearly and orderly. Specifically, in terms of time format, all time fields are stored as Coordinated Universal Time (UTC), and are converted according to the time zone where the owner is located when displayed; in terms of address format, it is uniformly recorded as "province - city - district - detailed address". For example, "Beijing - Chaoyang District - Wangjing East Road No. 10".

[0064] Furthermore, the dimensions of the data can also be unified, that is, map the data to a unified range. For example, the property management fee is based on the monthly cost: monthly payment: directly use the amount; quarterly payment: the amount is divided by 3 and then mapped to the monthly cost; annual payment: the amount is divided by 12 and then mapped to the monthly cost. The values of the data can also be unified, that is, different expressions of the same field are uniformly processed. For example, "normal payment" and "on - time payment" are classified as "on time".

[0065] In the embodiments of this application, for data from different channels, formats and structures (including text, numerical values, logs, social media information), through an efficient data extraction, transformation and loading (Extract - Transform - Load, ETL) process and algorithm strategy, that is, through specific data cleaning, outlier removal, text processing and semantic analysis methods, multi - source data is unified into a standardized data set that can be processed by subsequent models, realizing efficient data fusion and refinement.

[0066] Continue to refer to Figure 1 , in step 102, common labels and characteristic labels of users are determined according to multi - source data. The common labels are used to indicate the inherent information of users, and the characteristic labels are used to indicate the personalized behaviors of users.

[0067] In this embodiment, the common labels are used to indicate the user's inherent information, which generally does not change over time. Specifically, the common labels may include at least one of the following:

[0068] Owner identity label;

[0069] House and asset label;

[0070] Residence characteristic label.

[0071] Among them, the owner identity label may include at least one of the following fields: name, gender, age range, occupation type. The house and asset label may include at least one of the following fields: house use, house area, parking space situation, house status. The residence characteristic label may include at least one of the following fields: residence status, family structure.

[0072] For example, the user's information is as follows: Name: Zhang San; Gender: Male; Age range: 36 - 55 years old; Occupation type: enterprise employee; House use: self - occupied house; House area: 100㎡; Parking space situation: has a parking space; House status: normal residence; Residence status: small family; Family structure: a family of three. Then, the common labels of this user are: middle - aged man, self - occupied medium - sized house, family with a car, a family of three.

[0073] In this embodiment, the characteristic labels are used to indicate the user's personalized behaviors. The types of personalized behaviors are more and more complex and can reflect the user's personality. Specifically, the characteristic labels may include at least one of the following:

[0074] Behavior characteristic labels (such as payment behavior, complaint behavior, emotional expression);

[0075] Consumption characteristic labels (such as value - added service preference, price sensitivity);

[0076] Social characteristic labels (such as community participation, WeChat group activity);

[0077] Care characteristic labels (such as special needs, birthday reminder);

[0078] Residence mode label;

[0079] Parking space usage situation label;

[0080] Property service satisfaction label.

[0081] Among them, the behavior characteristic tags may include at least one of the following fields: payment behavior, payment cycle, payment method, arrears risk, emotional expression. For example, payment behavior includes on-time payment, early payment, late payment, and risk of refusal to pay. Specifically, for payment behavior, behavior characteristic tags such as payment habits, payment cycle, and arrears risk can be extracted according to payment records. Exemplarily, through time series analysis algorithms (such as Autoregressive Integrated Moving Average Model, LSTM), the future payment behavior of property owners is predicted to identify users with high arrears risks. Exemplarily, through clustering analysis algorithms (such as K-Means), the payment habits of users are determined, such as "early payment", "on-time payment", and "delayed payment".

[0082] For complaint behavior, the complaint categories can be identified through text classification algorithms (such as BERT, TF-IDF combined with SVM), such as services, facilities, and environment. Then, through clustering algorithms (such as Density-Based Spatial Clustering of Applications with Noise, DBSCAN), the characteristics of complaint behaviors with different complaint frequencies are analyzed and determined. Specifically, the behavior characteristic tags can be generated according to the rules shown in Tables 1 to 7.

[0083] Table 1

[0084]

[0085] Table 2

[0086]

[0087] Table 3

[0088]

[0089] Table 4

[0090]

[0091] Table 5

[0092]

[0093] Table 6

[0094]

[0095] Table 7

[0096]

[0097] Among them, the consumption characteristic tags include at least one of the following fields: value-added service consumption habits, service type preferences. For example, value-added service consumption habits include high consumption, medium consumption, and low consumption. Specifically, the consumption characteristic tags can be generated according to the rules shown in Tables 8 to 9.

[0098] Table 8

[0099]

[0100] Table 9

[0101]

[0102] Among them, the social characteristic tags include at least one of the following fields: community engagement, activity preferences, instant messaging activity. Community engagement can reflect the user's community activity level and predict participation preferences. Specifically, the Apriori algorithm for association rule mining can be used to discover the user's preferences for different types of activities; then, time series (such as LSTM, Extreme Gradient Boosting (XGBoost)) can be used to predict the user's future community participation probability. Instant messaging activity can reflect the user's communication preferences. Specifically, text classification and clustering algorithms, such as TF-IDF combined with K-Means, can be used to identify the keywords of instant messaging and classify the types of user messages, such as property matters, casual life chats, pure emojis, etc.; then, natural language topic modeling algorithms, such as the LDA topic model, can be used to extract the main topics of chat records and determine the instant messaging activity. Specifically, the consumption characteristic tags can be generated according to the rules shown in Tables 10 to 12.

[0103] Table 10

[0104]

[0105] Table 11

[0106]

[0107] Table 12

[0108]

[0109] Among them, the care characteristic tags include at least one of the following fields: special needs, life stage, birthday or festival care. For example, special needs include: elderly care needs, children's service needs, pet-friendly needs. Specifically, the care characteristic tags can be generated according to the rules shown in Tables 13 to 15.

[0110] Table 13

[0111]

[0112] Table 14

[0113]

[0114] Table 15

[0115]

[0116] Among them, the living mode label is used to distinguish whether the owner lives in the property for a long time or rents it out, helping the property management to provide more targeted services. Specifically, it can be divided into the following fields: self-occupied owners, rental owners, and short-term rental owners. Self-occupied owners indicate that the owner lives in the house for a long time, and the multi-source data involved are property registration information and payment records; rental owners indicate that the owner owns the property but does not live there, and the multi-source data involved are information of the water and electricity bill payers and tenant registration information; short-term rental owners indicate that the owner's house is used for short-term rental, and the multi-source data involved are short-term high-frequency rental record information.

[0117] Among them, the parking space usage label is used to clarify the owner's parking space needs, optimize parking management, and improve the utilization rate of parking space resources. Specifically, it can be divided into the following fields: long-term fixed parking spaces, temporary parking users, and carless owners. The multi-source data involved are property registration data and data in the property parking system.

[0118] Among them, the property service satisfaction label is used to measure the owner's satisfaction with property management and optimize the property service experience. Specifically, it can be divided into the following fields: highly satisfied owners, moderately satisfied owners, and lowly satisfied owners. The multi-source data involved are instant messaging messages and complaint information.

[0119] Continue to refer to Figure 1 , in step 103, an owner portrait of the user is constructed based on the common labels and characteristic labels.

[0120] In one implementation, the owner portrait may include common labels and characteristic labels. For example, the owner portrait of Owner 1 is: a family of three, forgetful about payments, hesitant to participate, and concerned about quality.

[0121] In another implementation, the owner portrait may include common labels, characteristic labels, and descriptive content, and the descriptive content is generated based on multi-source data, common labels, and characteristic labels. As Figure 2 shown, the content within the box 201 is the descriptive content in the owner portrait, and the content within the box 202 is the common labels and characteristic labels in the owner portrait.

[0122] In a non-limiting embodiment, the common labels can be generated in the following manner: matching the multi-source data with the keywords corresponding to each common label in the common label library to determine the common labels of the user.

[0123] In this embodiment, the common label library can be pre-set. The common label library includes multiple common labels and their corresponding keywords. When generating the common labels of a user, the multi-source data can be matched with each common label and its corresponding keyword, and the matched common label is the common label of the user.

[0124] In a non-limiting embodiment, the characteristic labels can be generated in the following way: generate the characteristic labels of the user according to the multi-source data.

[0125] Further, the multi-source data is input into a pre-trained large language model to output the characteristic labels of the user. The large language model is trained with training data, and the training data includes each multi-source data and its corresponding characteristic label.

[0126] For example, for the property payment behavior of the property owner, the input of the large language model is {"payment_date": "2025-01-03", "payment_amount": 900, "monthly_fee": 300}, and the large language model outputs: {"payment_behavior": "Pay quarterly on time", "payment_cycle": "Quarterly payment", "payment_status": "On time"}.

[0127] Among them, payment_date represents the payment date, payment_amount represents the payment amount, monthly_fee represents the monthly payment amount, payment_behavior represents the payment behavior, payment_cycle represents the payment cycle, and payment_status represents the payment status.

[0128] In this embodiment, the understanding ability of the large language model can be used to generate characteristic labels. At this time, the training data of the large language model includes each multi-source data and its corresponding characteristic label (pre-labeled). Further, the large language model can also be used to generate a portrait of the property owner. At this time, the training data is each multi-source data and its corresponding portrait of the property owner (pre-labeled).

[0129] In this embodiment, since the characteristic labels are generated instantaneously based on the collected multi-source data, the characteristic labels can more flexibly reflect the characteristics of the property owner, so as to accurately depict the portrait of the property owner. In addition, the sources of the multi-source data are more comprehensive, which also ensures the perfection of the portrait of the property owner.

[0130] In a non-limiting embodiment, for multiple characteristic labels, the characteristic labels can be selected according to the numerical size of the weights corresponding to the characteristic labels to participate in constructing the portrait of the property owner, or the characteristic labels can be displayed according to the numerical size of the weights corresponding to the characteristic labels.

[0131] For example, there are the following five types of characteristic tags: payment behavior, complaint behavior, participation in activities, value-added service preferences, and special needs, with weights of 5, 4, 3, 2, and 1 respectively; when four characteristic tags are selected in the user settings, the characteristic tags with weights of 5, 4, 3, and 2 are selected to construct the portrait of the property owner.

[0132] Another example is that there are the following five types of characteristic tags: payment behavior, complaint behavior, participation in activities, value-added service preferences, and special needs, with weights of 5, 4, 3, 2, and 1 respectively. The portrait of the property owner is generated according to the above five types of characteristic tags, and the corresponding descriptions are output in descending order of weights.

[0133] In practical applications, the portrait of the property owner includes multi-dimensional tags: common tags and characteristic tags, which can assist property management units to provide better property services. Specifically, it can include the following services: precise payment reminder, personalized activity push, and intelligent community operation.

[0134] Specifically, based on the overdue payment risk field in the characteristic tags, the property can send reminders in advance to high-risk property owners with overdue payments, or provide an installment payment plan for property owners with a high risk of refusal to pay.

[0135] Specifically, based on the family activity preference field in the characteristic tags, the property sends invitations for parent-child activities to property owners; based on the pet-friendly need field in the characteristic tags, the property pushes preferential information of pet hospitals to property owners.

[0136] Specifically, based on the social characteristic tags in the characteristic tags, high-social-activity property owners can be invited to be Key Opinion Leaders (KOLs) to improve the sense of participation of property owners; or invitation to join the group can be pushed to low-active property owners to enhance the community atmosphere.

[0137] In a non-limiting embodiment, in addition to generating the portrait of the property owner, a portrait of the property management can also be generated, through which the property services can be understood more comprehensively, providing a reference for subsequent improvement of property services.

[0138] Refer to Figure 3 , Figure 3 which shows a portrait generation method. The portrait generation method may include step 301 and step 302.

[0139] In step 301, service data for the property is obtained.

[0140] In specific implementation, the service data may include at least one of the following data: working status, maintenance status, and personnel status. More specifically, the service data may be sourced from the property management system of the property company, such as the task management system, customer service system, work order management system, equipment inspection report, repair system, historical maintenance record, and attendance system.

[0141] In step 302, a property portrait is generated based on the service data and multi-source data.

[0142] In specific implementation, the operation situation of the property company is decomposed into multiple key dimensions, such as service efficiency, service quality, equipment status, and team status. The corresponding service data is extracted respectively, and at least one of the following labels is generated: service efficiency, service quality, equipment status, and team status. Among them, service efficiency is used to represent the completion degree of property tasks, service quality is used to represent the service response speed, equipment status is used to represent the equipment maintenance condition, and team status is used to represent the working condition of property personnel.

[0143] Exemplarily, the service data is that 10 tasks need to be completed every day, and 8 tasks are actually completed. The service efficiency (or task completion rate) is 80%.

[0144] Exemplarily, the service quality reflects the service response duration. For example, the average response time from the owner's repair request to dispatching workers is 30 minutes.

[0145] Exemplarily, the service quality reflects the task delay rate. For example, 50% of the work orders are delayed by more than 48 hours.

[0146] Exemplarily, the service quality reflects the timely repair handling rate. For example, the proportion of repairs completed within 24 hours is 85%.

[0147] Exemplarily, the equipment status reflects the equipment intact rate. For example, among all the equipment, the proportion of equipment in normal operation is 90%.

[0148] Exemplarily, the equipment status reflects the secondary repair rate. For example, in the past month, the repeated repair rate for the same problem is 10%.

[0149] Exemplarily, the team status reflects the personnel attendance rate. For example, the ratio of the number of days an employee is present to the number of days they should be present is 95%.

[0150] Exemplarily, the team status reflects the personnel training coverage rate. For example, the proportion of employees who have participated in the annual training to the total number of employees is 60%.

[0151] Exemplarily, the team status reflects the personnel satisfaction. For example, the employee satisfaction score is 4.2 (out of 5).

[0152] In a practical application scenario, service efficiency can specifically include the following tags: High - efficiency execution means the task completion rate ≥ 90%; Medium - efficiency execution means the task completion rate is between 70% - 90%; Low - efficiency execution means the task completion rate < 70%.

[0153] In a practical application scenario, service quality can specifically include the following tags: High - quality service means the service response time ≤ 30 minutes and the repair reporting timeliness rate ≥ 90%; Medium service means the response time is 31 - 60 minutes and the timeliness rate is 70% - 90%; Low - quality service means the response time > 60 minutes and the timeliness rate < 70%.

[0154] In a practical application scenario, the device status can specifically include the following tags: The device is running well means the device integrity rate ≥ 95% and the secondary repair rate ≤ 5%; The device needs to be optimized means the integrity rate is between 80% - 95% and the secondary repair rate is between 5% - 10%; The device has serious problems means the integrity rate < 80% and the secondary repair rate > 10%.

[0155] In a practical application scenario, the team status can specifically include the following tags: Excellent team: attendance rate ≥ 95%, training coverage rate ≥ 80%, satisfaction ≥ 4.5; Medium team: attendance rate is between 85% - 95%, coverage rate is between 60% - 80%, satisfaction is between 4.0 - 4.5; Problem team: attendance rate < 85%, coverage rate < 60%, satisfaction < 4.0.

[0156] The embodiments of the present application can comprehensively and accurately depict property services from multiple dimensions, providing assistance for the improvement of property services.

[0157] In one implementation, the property portrait can include the above - mentioned tags and description content, and the description content is generated based on service data and the above - mentioned tags. As Figure 4 shown Figure 4 shows a property portrait for describing the status of the property's equipment. The content in box 401 is the description of the equipment in the property portrait, and the content in box 402 is the tags in the property portrait regarding the equipment: integrity rate, secondary repair rate, etc.

[0158] Furthermore, the property services can be scored according to each tag in the property portrait, the total score can be calculated, and rectification suggestions can be generated.

[0159] Please refer to Figure 5 , Figure 5 shows a portrait generation device 50. The portrait generation device 50 can include:

[0160] An acquisition module 501, configured to acquire multi - source data for users;

[0161] A label generation module 502 is configured to determine common labels and characteristic labels of a user based on multi-source data. The common labels are used to indicate the inherent information of the user, and the characteristic labels are used to indicate the personalized behaviors of the user.

[0162] A portrait generation module 503 is configured to construct a homeowner portrait of the user based on the common labels and the characteristic labels.

[0163] In a specific implementation, the above portrait generation device 50 may correspond to a chip with portrait generation function in a computing device, such as a System-On-a-Chip (SOC), a baseband chip, etc.; or correspond to a chip module including a chip with portrait generation function in a computing device; or correspond to a chip module with a chip having data processing function, or correspond to a computing device.

[0164] Other relevant descriptions of the portrait generation device 50 may refer to the relevant descriptions in the foregoing embodiments, and will not be elaborated here.

[0165] Regarding each device and product described in the above embodiments and the respective modules / units included therein, they may be software modules / units, or hardware modules / units, or may also be partly software modules / units and partly hardware modules / units. For example, for each device and product applied to or integrated into a chip, the respective modules / units included therein may all be implemented in a hardware manner such as circuits, or at least some of the modules / units may be implemented in a software program manner, and the software program runs on a processing module integrated inside the chip, and the remaining (if any) part of the modules / units may be implemented in a hardware manner such as circuits; for each device and product applied to or integrated into a chip module, the respective modules / units included therein may all be implemented in a hardware manner such as circuits, and different modules / units may be located in the same component (such as a chip, a circuit module, etc.) or different components of the chip module, or at least some of the modules / units may be implemented in a software program manner, and the software program runs on a processing module integrated inside the chip module, and the remaining (if any) part of the modules / units may be implemented in a hardware manner such as circuits; for each device and product applied to or integrated into a computing device, the respective modules / units included therein may all be implemented in a hardware manner such as circuits, and different modules / units may be located in the same component (such as a chip, a circuit module, etc.) or different components inside the computing device, or at least some of the modules / units may be implemented in a software program manner, and the software program runs on a processing module integrated inside the computing device, and the remaining (if any) part of the modules / units may be implemented in a hardware manner such as circuits.

[0166] The embodiment of the present application also discloses a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is run, the steps of the method shown in the above embodiment can be executed. The storage medium may include a read-only memory module (ROM), a random access memory module (RAM), a disk or an optical disk, etc. The storage medium may also include a non-volatile storage module (non-volatile) or a non-transitory storage module, etc.

[0167] It should be understood that the term "and / or" in this article is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article indicates that the associated objects before and after are in an "or" relationship.

[0168] The "plurality" appearing in the embodiments of the present application refers to two or more.

[0169] The first, second, etc. descriptions appearing in the embodiments of the present application are only used for illustration and distinction of the description objects. There is no order, nor do they indicate any special limitation on the number of devices in the embodiments of the present application, and cannot constitute any limitation on the embodiments of the present application.

[0170] The "connection" that appears in the embodiments of the present application refers to various connection methods such as direct connection or indirect connection to achieve communication between devices, and the embodiments of the present application do not impose any limitations on this.

[0171] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless means.

[0172] It should be understood that in various embodiments of the present application, the sequence numbers of the above processes do not imply the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0173] In several embodiments provided by the present application, it should be understood that the disclosed methods, devices and systems can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for example, the division of the units is only a logical function division, and there can be other division methods in actual implementation; for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces, and the indirect coupling or communication connection of the device or unit can be in electrical, mechanical or other forms.

[0174] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0175] In addition, in each embodiment of the present application, the functional units can be integrated into one processing unit, or each unit can be physically included separately, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware, or in the form of hardware plus software functional units.

[0176] The above integrated units implemented in the form of software functional units can be stored in a computer-readable storage medium. The above software functional units stored in a storage medium include several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute some steps of the methods described in each embodiment of the present application.

[0177] 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 should be subject to the scope defined by the claims.

Claims

1. An image generation method, characterized in that, Including: Obtain multi-source data for the user; Determine the common tags and characteristic tags of the user according to the multi-source data, where the common tags are used to indicate the inherent information of the user, and the characteristic tags are used to indicate the personalized behavior of the user. The characteristic tags include at least one of the following: behavior characteristic tags, consumption characteristic tags, social characteristic tags, care characteristic tags; Construct an owner portrait of the user based on the common tags and the characteristic tags.

2. The portrait generation method according to claim 1, wherein The obtaining of the multi-source data includes at least one of the following steps: Collect the basic information of the user; Obtain the instant messaging messages of the user; Scrape at least one of the property payment behavior information, complaint information, repair information, and activity participation information of the user.

3. The portrait generation method according to claim 1, wherein The common tags may include at least one of the following: owner identity tag, housing and asset tag, living characteristic tag.

4. The portrait generation method according to claim 1, wherein The behavior characteristic tags may include at least one of the following fields: payment behavior, payment cycle, payment method, arrears risk, emotional expression; the consumption characteristic tags include at least one of the following fields: value-added service consumption habit, service type preference; the social characteristic tags include at least one of the following fields: community participation degree, activity preference, instant messaging activity; The care characteristic tags include at least one of the following fields: special needs, life stage, birthday or festival care.

5. The portrait generation method according to claim 1, characterized in that The determining of the common tags and characteristic tags of the user according to the multi-source data includes: Match the multi-source data with the keywords corresponding to each common tag in the common tag library to determine the common tags of the user; Generate the characteristic tags of the user according to the multi-source data.

6. The portrait generation method according to claim 5, wherein The generating of the characteristic tags of the user according to the multi-source data includes: Input the multi-source data into a pre-trained large language model to output the characteristic tags of the user. The large language model is trained with training data, and the training data includes each multi-source data and its corresponding characteristic tags.

7. The portrait generation method according to claim 1, wherein Also including: Obtain service data for the property; Generate a property portrait according to the service data and the multi-source data, where the property portrait includes at least one of the following tags: service efficiency, service quality, equipment status, team status. The service efficiency is used to represent the completion degree of property tasks, the service quality is used to represent the service response speed, the equipment status is used to represent the equipment maintenance condition, and the team status is used to represent the working condition of property personnel.

8. An image generation device, characterized in that, Including: An obtaining module, configured to obtain multi-source data for the user; A tag generation module, configured to determine the common tags and characteristic tags of the user according to the multi-source data, where the common tags are used to indicate the inherent information of the user, and the characteristic tags are used to indicate the personalized behavior of the user. The characteristic tags include at least one of the following: behavior characteristic tags, consumption characteristic tags, social characteristic tags, care characteristic tags; A portrait generation module, configured to construct an owner portrait of the user based on the common tags and the characteristic tags.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processing module, it executes the steps of the portrait generation method according to any one of claims 1 to 7.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processing module, the steps of the image generation method according to any one of claims 1 to 7 are implemented.

11. A computing device, comprising a storage module and a processing module, wherein a computer program that can run on the processing module is stored on the storage module, characterized in that When the processing module runs the computer program, the steps of the image generation method according to any one of claims 1 to 7 are executed.

Images