Service data determination method and device for real estate service
Identify related users through video data and backtracking algorithms, and conduct quantitative analysis with the allocation model, which solves the problem of unfair performance distribution in real estate business, realizes fair distribution and performance evaluation of business data, and improves the work enthusiasm and team spirit of salesmen. .
Patent Information
- Application Number
- CN202410920357.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-10
- Publication Date
- 2024-11-12
AI Technical Summary
In real estate business, since home purchase decisions may involve multiple family members or affiliates, the performance distribution between salesmen is unfair, which may reduce the work experience and cause disputes.
By using video data and backtracking algorithms to identify related users in the home purchase decision process, and conduct quantitative analysis and performance allocation adjustments for the salespersons who receive customers based on the preset allocation model, we ensure that every salesperson involved in the home purchase decision can get reasonable performance recognition and distribution.
Accurate evaluation and fair distribution of matching user business data under the target conditions are achieved, the satisfaction of matching users is improved, disputes arising from unfair performance distribution are reduced, and the harmonious and efficient operation of the team is promoted.
Smart Images

Figure CN118941133A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing, and in particular to a method and device for determining business data for real estate business. Background Art
[0002] Business data of real estate business usually includes customer information, property information, salesperson performance, etc. For the management and analysis of salesperson performance, it is usually necessary to record detailed data such as the number of customers received, transaction volume, sales amount, etc. of each salesperson. These data are not only used to evaluate the work performance of salespersons, but also used to formulate incentive measures and allocate performance bonuses to ensure fairness and improve the overall operational efficiency of the team.
[0003] However, since the purchase decision may involve multiple family members or related parties, and the final contract may be signed by any one of them, during this process, each family member may receive a different salesperson, and they are unaware of each other. This situation may lead to unfair performance distribution among salespeople. Different salespeople may receive different participants in the purchase decision process, but the final transaction is attributed to one of the salespeople. This may not only reduce the work experience of other salespeople, but also cause disputes among salespeople.
[0004] Application Contents
[0005] The present application provides a method and device for determining business data for real estate business, so as to overcome the above-mentioned problems or at least partially solve the above-mentioned problems of unfair performance distribution that may occur and reduce the work experience of salesmen.
[0006] In a first aspect, the present application provides a method for determining business data for a real estate business, comprising:
[0007] When the target room meets the target condition, determine the target user, the first matching user and the primary service data corresponding to the target room;
[0008] Based on the preset video data, determine the associated user corresponding to the target user by using a backtracking algorithm, and determine the second matching user corresponding to the associated user;
[0009] Performing quantitative analysis on the first matching user and the second matching user using a preset allocation model to determine a first allocation ratio corresponding to the first matching user and a second allocation ratio corresponding to the second matching user;
[0010] The primary service data is adjusted based on the first allocation ratio and the second allocation ratio to determine the target service data corresponding to the target room.
[0011] In a second aspect, the present application provides a business data determination device for real estate business, including:
[0012] An information acquisition module, configured to determine a target user, a first matching user, and primary service data corresponding to the target room when the target room meets the target condition;
[0013] A second matching user determination module is configured to determine an associated user corresponding to the target user based on preset video data using a backtracking algorithm, and determine a second matching user corresponding to the associated user;
[0014] an allocation weight determination module, configured to perform quantitative analysis on the first matching user and the second matching user using a preset allocation model to determine a first allocation weight corresponding to the first matching user and a second allocation weight corresponding to the second matching user;
[0015] The target business data determination module is configured to adjust the primary business data based on the first allocation ratio and the second allocation ratio to determine the target business data corresponding to the target room.
[0016] In a third aspect, the present application provides a readable medium comprising execution instructions. When a processor of an electronic device executes the execution instructions, the electronic device executes any method described in the first aspect.
[0017] In a fourth aspect, the present application provides an electronic device, comprising a processor and a memory storing execution instructions, wherein when the processor executes the execution instructions stored in the memory, the processor executes any method described in the first aspect.
[0018] The present application provides a method and device for determining business data for real estate business, by determining the target user, the first matching user and the primary business data corresponding to the target room when the target room meets the target conditions; based on the preset video data, using the backtracking algorithm to determine the associated user corresponding to the target user, and determining the second matching user corresponding to the associated user; using the preset allocation model to quantitatively analyze the first matching user and the second matching user to determine the first allocation ratio corresponding to the first matching user and the second allocation ratio corresponding to the second matching user; adjusting the primary business data based on the first allocation ratio and the second allocation ratio to determine the target business data corresponding to the target room. The present application realizes the accurate evaluation and fair allocation of matching user business data in the process of achieving the target conditions, ensuring that each matching user can obtain reasonable business data recognition, and improving the satisfaction of matching users.
[0019] The further effects of the above-mentioned non-conventional preferred manner will be described below in conjunction with specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] In order to more clearly illustrate the embodiments of the present application or the existing technical solutions, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0021] Figure 1 A flowchart of a method for determining business data for real estate business provided in one embodiment of the present application;
[0022] Figure 2 A flowchart of another method for determining business data for real estate business provided in one embodiment of the present application;
[0023] Figure 3 A flowchart of another method for determining business data for real estate business provided in one embodiment of the present application;
[0024] Figure 4 A flowchart of another method for determining business data for real estate business provided in one embodiment of the present application;
[0025] Figure 5 It is a structural diagram of a business data determination device for real estate business provided by an embodiment of the present application;
[0026] Figure 6 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION
[0027] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.
[0028] Business data of real estate business usually includes customer information, property information, salesperson performance, etc. For the management and analysis of salesperson performance, it is usually necessary to record detailed data such as the number of customers received, transaction volume, sales amount, etc. of each salesperson. These data are not only used to evaluate the work performance of salespersons, but also used to formulate incentive measures and allocate performance bonuses to ensure fairness and improve the overall operational efficiency of the team.
[0029] However, since the purchase decision may involve multiple family members or related parties, and the final contract may be signed by any one of them, during this process, each family member may receive a different salesperson, and they are unaware of each other. This situation may lead to unfair performance distribution among salespeople. Different salespeople may receive different participants in the purchase decision process, but the final transaction is attributed to one of the salespeople. This may not only reduce the work experience of other salespeople, but also cause disputes among salespeople.
[0030] Therefore, this application provides a business data determination method for real estate business, which uses video data and backtracking algorithms to identify related users in the process of house purchase decision-making, and conducts quantitative analysis and performance distribution adjustment for salesmen who receive customers based on a preset distribution model. This can more fairly distribute the performance of salesmen, improve their work experience, and reduce disputes that may be caused by unfair performance distribution.
[0031] In this embodiment, a method for determining business data for real estate business includes:
[0032] Step 101: When a target room meets a target condition, a target user, a first matching user and primary service data corresponding to the target room are determined.
[0033] The target condition can be that the property is signed or meets certain sales criteria, which means that the property has reached the preset business indicators. The target user can refer to the buyer who finally signed the contract, that is, the customer who actually completed the purchase transaction. The first matching user refers to the salesperson who directly receives and serves the target user during the purchase process.
[0034] Primary business data includes basic information related to the transaction, such as the sales amount, transaction volume, transaction date, etc. These primary business data are the basic data for subsequent analysis and allocation. Through these data, the performance of the salesperson can be preliminarily evaluated.
[0035] After determining the target users, first matching users and primary business data, these data will be further analyzed to more fairly distribute the salesperson's performance and ensure that all salespeople involved in the home purchase decision-making process receive reasonable performance recognition and distribution.
[0036] Step 102: Based on the preset video data, a backtracking algorithm is used to determine an associated user corresponding to the target user, and a second matching user corresponding to the associated user is determined.
[0037] Family members may come into contact with different salespersons at different stages of the home purchase decision. For example, one family member may come into contact with one salesperson during the first house viewing, while another family member may come into contact with another salesperson during the final confirmation stage before signing the contract. Since these contacts and interactions directly affect the final home purchase decision, it is crucial to identify these associated users and their corresponding salespersons.
[0038] Based on the preset video data, the backtracking algorithm is used to determine the associated users corresponding to the target user, and further determine the second matching users corresponding to these associated users. This process first requires the collection and organization of all relevant video data generated during the home purchase decision-making process, such as videos of customers interacting with other personnel at the sales office. By analyzing these video data, the backtracking algorithm can track the behavior and interaction trajectory of the target user, and identify the associated users who have direct or indirect interactions with the target user during the home purchase decision-making process.
[0039] Related users can usually be considered as family members of the target user. In the process of house purchase, family members often participate in multiple stages of decision-making and discussion, so they play an important role in the entire house purchase process. In practice, the real estate industry usually regards the first family member to receive the person as a related user.
[0040] Backtracking algorithm is a systematic search algorithm that is widely used to solve combinatorial optimization problems and path finding problems. In the data determination process of real estate business, backtracking algorithm is used to track the interaction trajectory of target users and their associated users in house purchase decision-making. Its basic idea is to systematically explore all possible solution paths through step-by-step trial and backtracking to find the optimal solution.
[0041] Specifically, in the process of house purchase decision-making, the backtracking algorithm first analyzes the preset video data, and uses image recognition and behavior recognition technology to capture the target user's interaction with other people. These interaction data include the target user's communication and contact with salesmen, family members, and other related parties. The algorithm will start from the target user's initial position and gradually trace back every step of his or her actions in the house purchase decision, recording each interaction and contact.
[0042] When encountering a branch choice, the backtracking algorithm will evaluate each possible choice path, record the status after the choice and move on. If a path cannot meet the preset target conditions, the algorithm will go back to the previous step and try other possible paths until a path that meets the conditions is found or all possibilities are exhausted. In this way, the algorithm can comprehensively and systematically track the target user's home purchase decision-making process and identify all key interaction nodes and related users.
[0043] After identifying these related users, the next step is to determine the second matching users of the related users. The second matching users are the salespersons who receive and serve these related users in the process of house purchase decision-making.
[0044] In this way, all key participants and their corresponding salespersons in the entire home purchase decision chain can be fully tracked, thus providing accurate data support for subsequent performance distribution. This method can not only ensure that every salesperson involved in the reception receives due performance recognition, but also more accurately reflect the actual contribution of each salesperson in the complex home purchase decision process, thereby achieving fairness and transparency in performance distribution.
[0045] Step 103: Quantitatively analyze the first matching user and the second matching user using a preset allocation model to determine a first allocation ratio corresponding to the first matching user and a second allocation ratio corresponding to the second matching user.
[0046] The preset allocation model is built based on historical data, empirical rules and specific algorithms to quantify and evaluate the contribution of each salesperson in the home purchase decision-making process. The model takes into account multiple factors, such as the number of customer receptions, the depth of interaction, customer feedback, and the professional ability of the salesperson.
[0047] In specific operations, the information obtained from the video data and the backtracking algorithm is first input into the allocation model. The model will analyze the first matching user and the second matching user. Through quantitative analysis, the model will evaluate the role played by each salesperson in the entire home purchase decision-making process and its impact on the final transaction.
[0048] The allocation model usually includes a series of weights and parameters to evaluate the various contributions of salespeople. For example, the model may give higher weights to salespeople who receive customers for the first time, because they often have an important influence on the customer's initial impression and decision. At the same time, the model will also give corresponding weights to salespeople who provide important information or help at key decision nodes. Through these weights and parameters, the model can calculate the contribution score of each salesperson.
[0049] Then, based on these contribution scores, the model determines the first allocation weight for the first matching user and the second allocation weight for the second matching user. The first allocation weight reflects the share of the performance that the salesperson who directly receives the target user deserves, while the second allocation weight reflects the share of the performance that the salesperson who receives the associated user deserves. In this way, the model ensures that the performance distribution is not only based on the final signing result, but also fully considers the actual contribution of each salesperson in the entire decision-making process.
[0050] This quantitative analysis method not only makes performance distribution more fair and transparent, but also motivates sales staff to work harder during the reception process, thereby improving overall service quality and customer satisfaction.
[0051] Step 104: Adjust the primary service data based on the first allocation ratio and the second allocation ratio to determine the target service data corresponding to the target room.
[0052] Primary business data includes basic information related to real estate transactions, such as sales amount, etc. These data are the basis for performance distribution. By adjusting them, we can accurately reflect the contribution of each salesperson in the process of house purchase decision-making.
[0053] In specific operations, the system will multiply the primary business data by the first allocation ratio to calculate the first business data corresponding to the first matching user. For example, if the primary business data is 100,000 yuan and the first allocation ratio is 0.6, then the first business data of the first matching user will be 60,000 yuan. Similarly, the system will multiply the primary business data by the second allocation ratio to calculate the second business data corresponding to the second matching user. For example, if the second allocation ratio is 0.4, then the performance data of the second matching user will be 40,000 yuan.
[0054] This adjustment process not only takes into account the contribution of the salesperson who directly serves the target user, but also fully reflects the efforts made by the salesperson who serves the related users in the decision-making process. In this way, it ensures that all salespeople involved in the home purchase decision-making process can receive due performance recognition and rewards.
[0055] Ultimately, the target business data corresponding to the target room is the sum of the business data of each matched user. These adjusted target business data will serve as the final basis for performance distribution, ensuring the fairness and transparency of the distribution process. This method not only improves the accuracy of performance distribution, but also enhances the work enthusiasm and teamwork spirit of salesmen, reduces conflicts and disputes caused by unfair distribution, and thus promotes the harmony and efficient operation of the entire team.
[0056] Through the above technical solution, it can be seen that the beneficial effects of this embodiment are:
[0057] The embodiment of the present application develops a method and device for determining business data for real estate business. When the target room meets the target conditions, the target user, the first matching user and the primary business data corresponding to the target room are determined; based on the preset video data, the associated user corresponding to the target user is determined by the backtracking algorithm, and the second matching user corresponding to the associated user is determined; the first matching user and the second matching user are quantitatively analyzed by the preset allocation model to determine the first allocation ratio corresponding to the first matching user and the second allocation ratio corresponding to the second matching user; the primary business data is adjusted based on the first allocation ratio and the second allocation ratio to determine the target business data corresponding to the target room. The embodiment of the present application realizes the accurate evaluation and fair allocation of matching user business data in the process of achieving the target conditions, ensures that each matching user can get reasonable business data recognition, and improves the satisfaction of matching users.
[0058] Figure 1 What is shown is only a basic embodiment of a method for determining business data for real estate business of the present application. By making certain optimizations and extensions thereon, other preferred embodiments of a method for determining business data for real estate business can be obtained.
[0059] like Figure 2 As shown, another specific embodiment of a method for determining business data for real estate business according to the present application.
[0060] In this embodiment, a method for determining business data for real estate business includes the following steps:
[0061] Step 201: When a target room meets a target condition, a target user, a first matching user, and primary service data corresponding to the target room are determined.
[0062] Step 202: Based on the preset video data, a backtracking algorithm is used to determine an associated user corresponding to the target user, and a second matching user corresponding to the associated user is determined.
[0063] Step 203: Analyze the visual data to determine the interactive user corresponding to the target user.
[0064] First, the video data needs to be preprocessed, including video decoding and video frame extraction. Each frame of the video image needs to be processed by image recognition technology to identify and mark each individual appearing in the video. This step usually uses face recognition technology, posture detection algorithms and other machine learning models to ensure the accuracy and comprehensiveness of recognition.
[0065] Next, the behavior recognition and tracking of the individuals marked in each frame of the image are performed. Behavior recognition technology can analyze the movements, postures, and movement paths of individuals to help determine which individuals are interacting with the target user. These interactions can be face-to-face conversations, joint viewing of real estate information, and visiting model homes together. Through behavior recognition, the details of these interactions can be captured, and information such as the time, frequency, and duration of the interactions can be recorded.
[0066] After identifying all individuals appearing in the video and annotating their behaviors, the next step is to match these individuals with the target user’s interactions. Interacting users are all individuals who have direct or indirect communication with the target user in the video. These interactions can take various forms, such as the target user consulting with a salesperson, discussing with a family member, or exchanging opinions with other potential buyers.
[0067] Parsing video data also includes the application of speech recognition and natural language processing technologies. By processing the audio in the video, the conversation content can be transcribed and analyzed to identify the roles and contributions of the interactive users in the conversation. For example, through speech recognition technology, the conversation can be converted into text, and then analyzed through natural language processing technology to identify the users who made key suggestions or made important decisions.
[0068] In summary, parsing video data is a complex and sophisticated process. Through the comprehensive application of multiple technical means such as image recognition, behavior recognition, voice recognition and natural language processing, we can comprehensively capture and analyze the interaction of target users and accurately determine the interactive users corresponding to the target users. This step lays a solid foundation for the subsequent identification of associated users and allocation of business data, ensuring that every matching user who contributes to the home purchase decision-making process can receive due performance recognition.
[0069] Step 204: Based on the video data and each interactive user, determine the associated users among the interactive users using a preset associated identification algorithm.
[0070] First, by processing and analyzing the video data, all users who have interacted with the target user have been identified. Next, the preset association recognition algorithm will filter out related users from these interactive users through a series of data analysis and model calculations.
[0071] The core of the association identification algorithm is to evaluate the strength and importance of the relationship between the interactive user and the target user through a variety of analysis methods. The algorithm can analyze the interaction frequency and interaction duration. The interaction frequency refers to the number of times the target user communicates with a certain interactive user, and the interaction duration refers to the duration of each interaction.
[0072] Social graph analysis is also an important part of the association identification algorithm. Based on the video data, the interaction between the target user and the interactive users is analyzed to build the target user's social graph; based on the social graph, the graph analysis algorithm is used to identify the target user's family members among the interactive users; based on the time priority principle, the family members who appear first in the video data are determined as the associated users.
[0073] By constructing a social network graph of the target user, the algorithm can intuitively display the strength and level of the relationship between the target user and each interactive user. In this graph, nodes represent users, edges represent interactions, and the weight of the edges reflects the frequency and intensity of interactions. Through graph analysis, the algorithm can identify the core social circle of the target user and determine the associated users within it.
[0074] By analyzing the interaction between the target user and each interactive user through video data, the social graph of the target user is constructed. A social graph is a visual network diagram in which each node represents a user and the lines between nodes represent the interactive relationship between users. By analyzing the interactive behavior in the video data, the algorithm can identify who has communicated with the target user, the frequency and intensity of the interaction, and the specific content of the interaction.
[0075] Next, based on the constructed social graph, the graph analysis algorithm is used to identify the target user's family members among the interactive users. The graph analysis algorithm can effectively identify which interactive users are the target user's family members by evaluating the relationship strength and pattern between nodes. These family members usually have more frequent and deeper interactions with the target user, and the content of the interactions is more decisive. The graph analysis algorithm can accurately filter out the target user's family members from many interactive users through these characteristics.
[0076] Finally, based on the time priority principle, the family member who first appears in the video data is determined as the associated user. The application of this principle is based on a common practice in the real estate industry, that is, the first salesperson to receive the family member is usually regarded as an associated user. In the process of house purchase decision-making, family members may interact and consult with multiple salespeople, but the first salesperson to receive the family member often plays a key role in the whole process. The first salesperson is usually the first person to understand the customer's needs, provide information and advice, and establish trust and relationship in the initial interaction.
[0077] Therefore, the time priority principle emphasizes that the family member who appears first in the video data is often the individual with the closest relationship and the greatest influence on the target user. Through video data analysis, the family member who first interacts with the target user in the entire home buying process can be accurately identified. This first-appearing family member usually plays an important role in the target user's home buying decision because they may be the first to make opinions, suggestions, or make decisions.
[0078] This principle ensures that the identified associated users have a significant influence in the home purchase decision process, thereby ensuring the fairness of business data distribution. In this way, the problem of unfair performance distribution caused by multiple interactions and the participation of multiple salesmen can be avoided. Ultimately, this method not only improves the work enthusiasm and satisfaction of salesmen, but also improves the overall work efficiency and customer satisfaction of the team. In some cases, the target user itself can also be an associated user. That is, the target user is the first member of the family that appears in the video data.
[0079] The algorithm can also use semantic analysis and natural language processing technology to deeply analyze the conversation content in the video. The natural language algorithm is used to perform semantic analysis on the video data to determine the keywords of the conversation between the target user and the interactive user; based on the keywords, the target user's family members are determined among the interactive users; based on the video data, the associated users are determined among the family members.
[0080] Through natural language processing technology, the algorithm can analyze the voice and text data in the video and extract key information from the conversation. First, the natural language algorithm is used to perform semantic analysis on the video data. The algorithm converts the audio in the video into text and performs word segmentation, part-of-speech tagging, and named entity recognition on the text content to extract keywords in the conversation. These keywords may include the characteristics, location, price, etc. of the house mentioned by the user, as well as the names and relationship information between users, such as "dad", "mom", "son", etc.
[0081] Based on these keywords, the algorithm can further analyze the relationship between the interactive users to determine the target user's family members. For example, when keywords such as "dad" and "mom" appear in the video, the algorithm can determine that these interactive users are the target user's family members. Combined with contextual information, the algorithm can further confirm the role and importance of these family members in the home purchase decision. In this way, the algorithm can accurately identify which interactive users are the target user's relatives and mark these relatives as potential related users.
[0082] Similarly, based on the time priority principle, the family member who first appears in the video data can be identified as the associated user. This means that in the process of house purchase decision-making involving multiple family members, the system will prioritize identifying the first family member who appears in the video data and interacts with the salesperson as the main associated user.
[0083] Step 205: Determine the user characteristics of the associated user using feature recognition technology.
[0084] The process of determining the user characteristics of associated users using feature recognition technology aims to accurately identify the identity and characteristics of associated users from video data. Feature recognition technology includes a variety of technical means such as facial recognition, voice recognition, and behavioral pattern analysis. By analyzing and processing video data, unique feature information of each user can be extracted.
[0085] Facial recognition technology is an important part of feature recognition. It analyzes the facial images in the video and extracts the user's facial features, such as the relative position and shape of the eyes, nose, mouth and other parts, as well as the overall facial contour. By comparing with the facial features in the database, the identity of each user in the video can be accurately identified. The advantages of facial recognition technology are its high accuracy and high reliability. It can accurately identify in complex environments and maintain a high recognition rate even in poor lighting or when the user's facial angle changes.
[0086] Speech recognition technology extracts the user's voice features, such as pitch, timbre, speaking speed and intonation, by analyzing the sound data in the video. Each person's voice features are unique, and through speech recognition technology, these features can be converted into digital information and matched with the voice features in the database to confirm the user's identity. In addition, speech recognition technology can also understand the role and importance of the user in the interaction by analyzing the content of the conversation.
[0087] Behavioral pattern analysis is a method of identifying users' characteristics by observing their behaviors and interactions. By analyzing the user's movements, postures, and interactions in the video, information reflecting the user's identity and characteristics can be extracted. For example, a family member may exhibit different behavioral patterns from other members during the house purchase process, such as frequently checking the details of the house and actively communicating with the salesperson. By analyzing these behavioral patterns, the user's identity and role can be further confirmed.
[0088] By comprehensively utilizing the above feature recognition technologies, comprehensive and accurate user feature information can be extracted from video data. This feature information provides a basis for determining the user information of associated users in the user database in the future, ensuring the accuracy and comprehensiveness of the recognition process. This multi-dimensional feature recognition method can effectively overcome the recognition difficulties caused by incomplete information or complex environment, and provide reliable data support for subsequent business data allocation.
[0089] Step 206: Determine user information of the associated user in the user database based on the user characteristics.
[0090] Through the previously extracted user features, specific user information matching these features is found in the user database, so as to accurately confirm the identity and detailed information of the associated user. The user database usually contains detailed information of all potential customers and actual customers, including but not limited to name, contact information, ID number, home purchase requirements, past interaction records, etc. This information is constantly updated and maintained in daily business operations to ensure its accuracy and comprehensiveness.
[0091] When these user features are input into the database, the system will use specific algorithms for comparison. These algorithms can include facial recognition algorithms, voice matching algorithms, and behavior pattern analysis algorithms. Each algorithm will analyze the input feature information and compare it with the corresponding features stored in the database. For example, the facial recognition algorithm will compare the extracted facial feature points with the facial images in the database one by one to find the most matching record; the voice matching algorithm will determine the most suitable user by analyzing the voice features.
[0092] Through this comparison process, the system can quickly and accurately find user records that match user characteristics. The matching results will not only show the user's basic identity information, but also include the user's detailed home purchase needs, historical interaction records, preferences, etc. This information is of great significance for subsequent analysis and decision-making.
[0093] In summary, by determining the user information of the associated user in the user database based on the user characteristics, the identity of the associated user can be accurately confirmed through an efficient feature matching algorithm. This not only provides a reliable basis for the subsequent business data determination and performance distribution, but also improves the efficiency and accuracy of the entire identification process. This method can better serve customer management and business data distribution in real estate business, and promote fair, just and efficient business operations.
[0094] Step 207: Based on the user information, a matching algorithm is used to determine a second matching user corresponding to the associated user.
[0095] Based on user information, the matching algorithm is used to determine the second matching user corresponding to the associated user. The purpose is to identify other salesmen who have important interactions with the associated user in the home purchase process, namely the second matching users, through the analysis of user information. This step ensures that all salesmen who have contributed to the home purchase decision process can be accurately identified and quantified, thereby achieving fair performance distribution.
[0096] First, the detailed user information of the associated users determined in the previous step, such as name, contact information, home purchase needs, and past interaction records, will be used as input data for the matching algorithm. The matching algorithm is designed to screen out those salesmen who have had important interactions with the associated users in the home purchase decision-making process from a large amount of user data. This process needs to consider multiple factors, including interaction frequency, interaction content, interaction time, etc.
[0097] The matching algorithm will first analyze the interaction records of the associated user and identify all the salespersons who have interacted with them. These interaction records may include telephone conversations, interview records, email exchanges, etc. The algorithm will classify and count these records to determine which salespersons have had key interactions with the associated user in the home purchase decision process. For example, if a salesperson accompanies a user to view houses many times and answers various questions of the user in detail, then the salesperson's contribution to the entire decision process is obviously significant.
[0098] Next, the matching algorithm will conduct a quantitative analysis of these interactions. By calculating indicators such as interaction frequency, duration, and content depth, the algorithm can evaluate the role and contribution of each salesperson in the entire home buying process. For example, a salesperson may have had multiple long communications with associated users, and these communications may cover the entire process from initial consultation to final decision-making. In this case, the salesperson's contribution will be rated higher by the algorithm. In addition, the algorithm will also consider the chronological order of interactions, and prioritize identifying salespeople who begin to interact with associated users in the early stages of decision-making, because they often have a greater impact on the user's final decision.
[0099] By comprehensively analyzing these indicators, the matching algorithm can ultimately determine the contribution of each salesperson and identify the second matching user corresponding to the associated user. This process not only takes into account the number of interactions, but also comprehensively evaluates the quality and influence of the interactions, ensuring the fairness and accuracy of the identification results.
[0100] Step 208: Perform quantitative analysis on the first matching user and the second matching user using a preset allocation model to determine a first allocation ratio corresponding to the first matching user and a second allocation ratio corresponding to the second matching user.
[0101] Step 209: Adjust the primary service data based on the first allocation weight and the second allocation weight to determine the target service data corresponding to the target room.
[0102] Through the above technical solution, it can be seen that the beneficial effect of this embodiment is that through this multi-level and multi-dimensional analysis, the preset association identification algorithm can accurately identify the associated users and second matching users in the house purchase decision-making process. This not only ensures the fairness and transparency of business data distribution, but also improves the work enthusiasm of matching users, motivating them to provide better services during the reception process, thereby improving the operational efficiency and customer satisfaction of the entire team.
[0103] like Figure 3 FIG. 1 is another specific embodiment of a method for determining business data for real estate business of the present application. This embodiment is further described on the basis of the above-mentioned embodiment.
[0104] In this embodiment, a method for determining business data for real estate business includes the following steps:
[0105] Step 301, determine the initial model; determine the historical business data corresponding to the real estate business; construct a training data set for the initial model based on the historical business data; the training data set includes data features of each training matching user.
[0106] In the context of real estate business, the selection of the initial model may include a variety of machine learning algorithms, such as linear regression, decision tree, random forest, etc. Each algorithm has its own unique advantages and applicable scenarios. Choosing the right algorithm can improve the accuracy and stability of the model.
[0107] The historical business data corresponding to the real estate business contains detailed performance records of each salesperson in the past, such as the number of customers received, transaction volume, workload, etc. These data not only reflect the historical performance of the salesperson, but also reveal the overall trends and rules of business operations. By comprehensively collecting and organizing these data, a rich source of information can be provided for the model, enabling it to fully learn the distribution pattern and influencing factors of the salesperson's business data during the training process.
[0108] The training data set is the basis for model learning and training, which includes detailed data features of each training matching user. Data features can cover information in multiple dimensions, such as basic information of customers, properties, interaction records of salesmen and sales performance. The selection and processing of these features directly affect the prediction ability and accuracy of the model. By properly designing and constructing the training data set, it can be ensured that the model can capture key patterns and potential relationships in the data during the learning process.
[0109] The training data set includes the data features of each training matching user. These features have been carefully selected and processed to effectively reflect the user's behavior and contribution in the real estate business. For example, customer information may include the customer's housing purchase needs, budget and intended housing sources, and housing source information may include the location, area, price, etc. of the house, while the salesperson's business data covers the number of customers he has received, the number of houses sold, the total amount of sales, etc. These multi-dimensional data features provide comprehensive input for the model, enabling it to fully learn and optimize during the training process.
[0110] Through this series of steps, a high-quality initial model can be constructed, laying a solid foundation for subsequent forecasting analysis and performance distribution. This not only helps to improve the work enthusiasm and team collaboration efficiency of sales staff, but also ensures the fairness and rationality of performance distribution and improves the efficiency and benefits of overall business operations.
[0111] Step 302: Determine the annotation information corresponding to the training features; the annotation information includes the actual allocation proportion corresponding to each training matching user.
[0112] The labeling information is the target of model learning, which reflects the distribution ratio of each training matching user in the actual business. In the context of real estate business, the labeling information can include the actual performance distribution ratio of the salesperson, which is based on their contribution in the historical business data. By matching the training features with the labeling information, the model can learn how to accurately predict the performance distribution ratio of the salesperson based on the input features.
[0113] The process of determining the annotation information needs to take into account many factors. The annotation information usually comes from the real business data allocation records in the historical business data. These records reflect the actual work performance of the salesperson. Through careful analysis and processing of these records, it can be ensured that the annotation information can accurately represent the actual contribution of the salesperson.
[0114] At the same time, the annotation information needs to be reasonably matched with the training features. The training features include customer information, property information, salesperson interaction records and sales performance, etc. These features constitute the input of the model. When determining the annotation information, it is necessary to ensure that each training feature corresponds to an accurate actual allocation ratio. In this way, the model can learn the relationship between features and allocation ratios through these matching data during the training process.
[0115] By accurately determining the annotation information corresponding to the training features, we can provide the model with reliable learning goals, so that it can fully learn the salesperson’s business data distribution pattern and influencing factors during the training process. This not only improves the model’s prediction ability and accuracy, but also provides a scientific basis for subsequent performance distribution, ensuring the fairness of the distribution process and improving the team’s overall operational efficiency and work enthusiasm.
[0116] Step 303: input the training features into the initial model so that the initial model outputs the predicted weight.
[0117] Inputting training features into the initial model is the process of model training, which enables the initial model to learn and predict based on the input data. In the context of real estate business, training features include customer information, property information, interaction records of salespersons, etc. These features constitute the input data of the model. By inputting these training features into the initial model, the model can calculate based on these data and output the predicted allocation ratio for each salesperson.
[0118] In the specific operation, all training features need to be preprocessed and standardized first. This step includes data cleaning, missing value processing, feature selection and feature scaling. The preprocessed data will be more standardized and consistent, which is conducive to the training and learning process of the model. Standardized features ensure that data of different dimensions can be reasonably processed in the model, avoiding excessive impact of certain features on the model due to dimensional differences.
[0119] Next, the preprocessed training features are input into the initial model. The initial model can be one of a variety of machine learning algorithms, such as linear regression, decision tree, random forest, neural network, etc. Different model algorithms have different characteristics and advantages. Choosing the right algorithm can improve the prediction accuracy and stability of the model. After the training features are input, the model will calculate based on these features to generate the predicted allocation ratio for each salesperson.
[0120] During the calculation process, the model uses its internal parameters and structure to process and analyze the input training features. For example, for a linear regression model, the calculation process includes a linear combination of features and model weights; for a neural network model, the calculation process includes the application of multiple layers of nonlinear activation functions. The result of these calculation steps is the predicted value generated by the model for each input sample, that is, the predicted allocation ratio of the salesperson.
[0121] By inputting the training features into the initial model, the model can learn and predict based on the input data, thereby outputting the predicted distribution ratio for each salesperson. These prediction results reflect the model's estimate of the salesperson's performance distribution based on their performance and feature relationships in the training data. By continuously inputting new training features and making adjustments and optimizations, the model can gradually improve its prediction accuracy, and ultimately provide a scientific and reasonable reference basis for the performance distribution of real estate business.
[0122] Step 304: Determine the loss index of the initial model based on a preset loss function using the predicted weight and the annotation information.
[0123] The loss function is used to measure the difference between the model's prediction results and the actual results. By minimizing the value of the loss function, the model's prediction accuracy can be improved. In the context of real estate business, the initial model outputs the predicted allocation ratio for each salesperson based on the input training features. These predicted ratios reflect the model's estimate of the salesperson's performance distribution. However, in order to ensure the accuracy of these predictions, they need to be compared with the actual allocation ratio. The actual allocation ratio comes from the annotation information and reflects the salesperson's true contribution in the historical business data.
[0124] The loss function is a bridge between the prediction weight and the annotation information. Common loss functions include mean square error, mean absolute error, cross entropy loss, etc. By calculating the loss function value, the loss index of the initial model can be obtained. The smaller the loss index, the closer the model's prediction results are to the actual situation, and the higher the accuracy of the model. Conversely, the larger the loss index, the greater the deviation between the model's prediction results and the actual situation, and the model needs to be further optimized.
[0125] During model training, the calculation of the loss function is an iterative process. In each iteration, the model generates a prediction weight based on the current parameters and structure, and calculates the loss index. According to the size of the loss index, the model adjusts its internal parameters to minimize the value of the loss function. Common optimization algorithms include gradient descent and stochastic gradient descent. These algorithms continuously adjust model parameters to gradually reduce the loss function value, thereby improving the prediction accuracy of the model.
[0126] Step 305: When the loss index does not meet the preset training conditions, adjust the hyperparameters of the initial model.
[0127] Hyperparameters are parameters that need to be manually set during model training. They are set before model training begins and directly affect the training efficiency and prediction performance of the model. Common hyperparameters include learning rate, batch size, regularization parameters, number of network layers and number of nodes, etc.
[0128] In the context of real estate business, the initial model is trained by inputting training features and annotation information, outputting the predicted weight, and calculating the loss index. If the loss index does not meet the preset training conditions, it indicates that the prediction performance of the model is not ideal and the model needs to be optimized by adjusting the hyperparameters.
[0129] In practice, hyperparameters are usually adjusted using methods such as grid search, random search, or Bayesian optimization. These methods systematically explore different hyperparameter combinations to find the best parameter configuration. After each adjustment, the model needs to be retrained and evaluated until the loss index meets the preset training conditions.
[0130] By adjusting the hyperparameters of the initial model, the model’s training effect and prediction performance can be significantly improved. This step not only helps the model better learn data features during training, but also improves the model’s accurate prediction of salesperson performance distribution, providing a solid foundation for fair distribution and efficient operation of real estate business.
[0131] Step 306: When the loss index meets the training condition, the initial model is determined to be the allocation model.
[0132] When the loss index meets the training conditions, determining the initial model as the allocation model is the final step of model training, which indicates that the model has reached the expected performance standard and has the ability to make predictions in practical applications. In the context of real estate business, this process ensures that the model can accurately and fairly allocate salesperson performance, improving overall operational efficiency and team work experience.
[0133] During the model training process, the initial model will go through multiple iterations. In each iteration, the model generates a prediction weight based on the input training features and annotation information, and calculates the loss index. The loss index reflects the difference between the model's prediction results and the actual results. By continuously adjusting the model's hyperparameters and structure, the model's loss index gradually decreases. When the loss index is reduced to the preset training conditions, that is, when the pre-set performance indicators are met, it means that the model has been able to fit the training data well and has a strong generalization ability.
[0134] Step 307: When the target room meets the target condition, determine the target user, the first matching user and the primary service data corresponding to the target room.
[0135] Step 308: Based on the preset video data, determine the associated user corresponding to the target user using a backtracking algorithm, and determine the second matching user corresponding to the associated user.
[0136] Step 309: Perform quantitative analysis on the first matching user and the second matching user using a preset allocation model to determine a first allocation ratio corresponding to the first matching user and a second allocation ratio corresponding to the second matching user.
[0137] Step 310: Adjust the primary service data based on the first allocation ratio and the second allocation ratio to determine the target service data corresponding to the target room.
[0138] It can be seen from the above technical solutions that the beneficial effect of this embodiment is: by constructing an accurate and effective performance allocation model, it provides strong technical support for business data management and allocation in real estate business.
[0139] like Figure 4 As shown, another specific embodiment of a method for determining business data for real estate business according to the present application.
[0140] In this embodiment, a method for determining business data for real estate business includes the following steps:
[0141] Step 401: When a target room meets a target condition, a target user, a first matching user and primary service data corresponding to the target room are determined.
[0142] Step 402: Based on the preset video data, a backtracking algorithm is used to determine an associated user corresponding to the target user, and a second matching user corresponding to the associated user is determined.
[0143] Step 403: Quantitatively analyze the first matching user and the second matching user using a preset allocation model to determine a first allocation weight corresponding to the first matching user and a second allocation weight corresponding to the second matching user.
[0144] Step 404: Adjust the primary service data based on the first allocation ratio and the second allocation ratio to determine the target service data corresponding to the target room.
[0145] Step 405: Determine the first service data corresponding to the first matching user based on the product of the first allocation ratio and the primary service data.
[0146] Primary business data represents the overall results of the entire business activity, such as total sales amount, total transaction volume, etc., while the first matching user refers to the salesperson who initially receives the target user and has an important influence on his or her home purchase decision.
[0147] Specifically, primary business data needs to be distributed according to the contribution of each salesperson. In order to achieve this goal, it is necessary to first determine the distribution ratio of the first matching user, that is, the first distribution ratio. The first distribution ratio reflects the proportion of the salesperson in the entire business activity. This ratio can be determined in many ways, such as a comprehensive evaluation based on factors such as the salesperson's contact time with the customer, communication frequency, and customer satisfaction.
[0148] Once the first allocation ratio is determined, the first business data corresponding to the first matching user can be calculated by multiplying it with the primary business data. For example, if the primary business data is a total sales amount of 100,000 yuan, and the first allocation ratio is 0.6, then the business data of the first matching user is 60,000 yuan. This means that the salesperson contributed 60,000 yuan of business data in the entire sales process.
[0149] The advantage of this allocation method is that it can objectively and fairly reflect the actual contribution of the salesperson, avoiding the limitation of allocating all performance based solely on the salesperson who finally signs the order. Through quantitative analysis, each salesperson's efforts and contributions can be recognized and rewarded as they should be, thereby motivating the salesperson to actively participate and fulfill their duties throughout the sales process.
[0150] Step 406: Determine the second service data corresponding to the second matching user based on the product of the second allocation ratio and the primary service data.
[0151] The second matching user is usually another salesperson who interacts with the associated user and has a significant impact on the home purchase decision process. In order to ensure that the contributions of all salespeople involved in the business activities are recognized and distributed fairly and impartially, the primary business data must be reasonably distributed according to the distribution ratio of different salespeople.
[0152] The allocation ratio of the second matching user, i.e., the second allocation ratio, needs to be determined by a preset allocation model. This ratio reflects the contribution ratio of the second matching user in the entire business activity. When determining the second allocation ratio, multiple factors can be considered, such as the frequency of interaction between the salesperson and the customer, the depth of the communication content, and the degree of influence on the customer's decision.
[0153] Once the second allocation ratio is determined, the second business data corresponding to the second matching user can be calculated by multiplying it with the primary business data. For example, assuming the primary business data is a total sales amount of 100,000 yuan, and the second allocation ratio is 0.4, the business data of the second matching user is 40,000 yuan. This means that the second matching user contributed 40,000 yuan of business data in the entire sales process.
[0154] This calculation process ensures that the actual contribution of each salesperson can be quantified and reasonably allocated, avoiding the unfair phenomenon caused by the salesperson who finally signs the order monopolizing all the results. In this way, the efforts of all salespeople can be recognized and rewarded, thus motivating them to actively participate and contribute throughout the business process.
[0155] Step 407: Determine target business data according to the first business data and the second business data.
[0156] The target business data is determined based on the first business data and the second business data. This step combines the above two allocation results to form the final target business data. The target business data reflects the specific contributions of different salesmen in the entire business activity. This method ensures that the allocation of business data is scientifically based and fair, and avoids unfair allocation caused by a single salesman signing a contract.
[0157] This method not only helps the system to formulate effective incentive policies, but also provides a reliable data basis for business operations, supporting decision-making and strategic adjustments. In short, through a scientific and fair business data allocation method, enterprises can better motivate team members and optimize resource allocation, thereby improving overall business operation efficiency and competitiveness.
[0158] Through the above technical solutions, it can be seen that the beneficial effect of this embodiment is that it not only takes into account the specific workload of each matching user at different stages, but also uses a scientific allocation model for quantitative analysis, thereby achieving fair and just business data allocation. This not only improves the work enthusiasm of matching users, but also helps to improve teamwork and overall operational efficiency.
[0159] like Figure 5 As shown in the figure, a specific embodiment of a business data determination device for real estate business of the present application is shown in the figure. Figures 1 to 4 A physical device of a method for determining business data for real estate business. Its technical solution is essentially consistent with the above embodiment, and the corresponding description in the above embodiment is also applicable to this embodiment. In this embodiment, a device for determining business data for real estate business includes:
[0160] The information acquisition module 501 is configured to determine the target user, the first matching user and the primary service data corresponding to the target room when the target room meets the target condition;
[0161] The second matching user determination module 502 is configured to determine the associated user corresponding to the target user by using a backtracking algorithm based on the preset video data, and determine the second matching user corresponding to the associated user;
[0162] The allocation weight determination module 503 is configured to perform quantitative analysis on the first matching user and the second matching user using a preset allocation model to determine a first allocation weight corresponding to the first matching user and a second allocation weight corresponding to the second matching user;
[0163] The target service data determination module 504 is configured to adjust the primary service data based on the first allocation weight and the second allocation weight to determine the target service data corresponding to the target room.
[0164] Figure 6 It is a structural diagram of an electronic device provided in an embodiment of the present application. At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. Among them, the memory may include a memory, such as a high-speed random access memory (Random-Access Memory, RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage, etc. Of course, the electronic device may also include hardware required for other services.
[0165] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.
[0166] The memory is used to store execution instructions. Specifically, the execution instructions are computer programs that can be executed. The memory may include internal memory and non-volatile memory, and provides execution instructions and data to the processor.
[0167] In a possible implementation, the processor reads the corresponding execution instruction from the non-volatile memory into the memory and then runs it, and can also obtain the corresponding execution instruction from other devices to form a business data determination device for real estate business at the logical level. The processor executes the execution instruction stored in the memory to implement a business data determination method for real estate business provided in any embodiment of the present application through the executed execution instruction.
[0168] The above application Figure 5The method for executing a business data determination device for real estate business provided in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or instructions in the form of software. The above processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The various methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.
[0169] The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in a memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.
[0170] The embodiment of the present application also proposes a readable medium, which stores an execution instruction. When the stored execution instruction is executed by a processor of an electronic device, the electronic device can execute a business data determination method for a real estate business provided in any embodiment of the present application, and is specifically used to execute the following: Figure 1 or Figure 2 or Figure 3 or Figure 4 The method shown.
[0171] The electronic device in each of the aforementioned embodiments may be a computer.
[0172] Those skilled in the art should understand that the embodiments of the present application can be provided as methods or computer program products. Therefore, the present application can adopt a complete hardware embodiment, a complete software embodiment, or a combination of software and hardware.
[0173] Each embodiment in this application is described in a progressive 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 device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0174] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0175] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.
Claims
1. A method for determining business data for real estate business, characterized in that: include: When the target room meets the target condition, determining the target user, the first matching user and the primary service data corresponding to the target room; Based on the preset video data, determine the associated user corresponding to the target user by using a backtracking algorithm, and determine the second matching user corresponding to the associated user; Performing quantitative analysis on the first matching user and the second matching user using a preset allocation model to determine a first allocation weight corresponding to the first matching user and a second allocation weight corresponding to the second matching user; The primary service data is adjusted based on the first allocation ratio and the second allocation ratio to determine the target service data corresponding to the target room.
2. The method according to claim 1, characterized in that The determining the associated user corresponding to the target user by using a backtracking algorithm based on the preset video data includes: Parsing the visual data to determine the interactive user corresponding to the target user; Based on the video data and each of the interactive users, the associated users are determined among the interactive users using a preset associated identification algorithm.
3. The method according to claim 2, characterized in that The determining the associated user among the interactive users by using a preset associated identification algorithm based on the video data and each of the interactive users comprises: Based on the video data, analyzing the interaction mode between the target user and the interactive user to construct a social graph of the target user; Based on the social graph, identifying family members of the target user among the interactive users using a graph analysis algorithm; Based on the time priority principle, the family member who appears earliest in the video data is determined as the associated user.
4. The method according to claim 2, characterized in that: The determining a second matching user corresponding to the associated user comprises: Determining user characteristics of the associated user using feature recognition technology; Determining user information of the associated user in a user database based on the user characteristics; Based on the user information, a matching algorithm is used to determine the second matching user corresponding to the associated user.
5. The method according to claim 1, characterized in that Also includes: Determine the initial model; Determining historical business data corresponding to the real estate business; constructing a training data set for the initial model based on the historical business data; The training data set includes data features of each training matching user; Determining annotation information corresponding to the training feature; The annotation information includes the actual allocation proportion corresponding to each of the training matching users; Inputting the training features into the initial model so that the initial model outputs a predicted weight; Determining the loss index of the initial model based on a preset loss function using the predicted weight and the labeled information; When the loss index does not meet the preset training conditions, adjusting the hyperparameters of the initial model; When the loss index satisfies the training condition, the initial model is determined to be the allocation model.
6. The method according to claim 5, characterized in that The adjusting the primary service data based on the first allocation weight and the second allocation weight to determine the target service data corresponding to the target room includes: determining first service data corresponding to the first matching user based on a product of the first allocation weight and the primary service data; determining second service data corresponding to the second matching user based on the product of the second allocation proportion and the primary service data; The target business data is determined according to the first business data and the second business data.
7. The method according to claim 2, characterized in that: The determining the associated user among the interactive users by using a preset associated identification algorithm based on the video data and each of the interactive users comprises: Performing semantic analysis on the video data using a natural language algorithm to determine keywords of the conversation between the target user and the interactive user; Based on the keywords, determining family members of the target user among the interactive users; Based on the video data, the associated users are determined among the family members.
8. A business data determination device for real estate business, characterized in that: include: An information acquisition module, configured to determine a target user, a first matching user, and primary service data corresponding to the target room when the target room meets the target condition; A second matching user determination module is configured to determine an associated user corresponding to the target user based on preset video data by using a backtracking algorithm, and determine a second matching user corresponding to the associated user; an allocation weight determination module, configured to perform quantitative analysis on the first matching user and the second matching user using a preset allocation model to determine a first allocation weight corresponding to the first matching user and a second allocation weight corresponding to the second matching user; The target business data determination module is configured to adjust the primary business data based on the first allocation weight and the second allocation weight to determine the target business data corresponding to the target room.
9. A computer-readable storage medium storing a computer program, wherein the computer program is used to execute the business data determination method for real estate business as described in any one of claims 1 to 7.
10. An electronic device, comprising: processor; a memory for storing instructions executable by the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the business data determination method for real estate business described in any one of claims 1-7.
