Operation strategy processing method and device for hotel OTA business

Abstract

The application provides a hotel OTA business operation strategy processing method and device, the method comprises the following steps: obtaining a preset discount quota and a plurality of preset influencing factors, generating at least one initial operation strategy based on the preset discount quota and the preset influencing factors through a pre-trained planning large model; replaying and evaluating each initial operation strategy through a pre-built risk pre-check platform to obtain a plurality of strategy risk groups; in the case that the operation income loss corresponding to the strategy risk group is less than a first threshold value, determining the initial operation strategy corresponding to the maximum operation income as a target operation strategy, and executing the target operation strategy, wherein the operation income is determined according to platform income, hotel income and user discount. The method solves the problem that in the prior art, when an operation strategy is formulated, there is a lack of access capability for market, weather and discount dimensions, resulting in that the actual online effect does not meet the expected result.

Description

Technical Field

[0001] This invention relates to the field of business operation technology, and more specifically, to a method, apparatus, computer-readable storage medium, and electronic device for processing operational strategies for hotel OTA business. Background Technology

[0002] In existing technologies, the generation, verification, and deployment of operational strategies for hotel OTA businesses rely on manual configuration and offline calculations, making it difficult to achieve multi-dimensional systematic perception and structured access to the market. When there are many factors to consider, the process cost of manual configuration is high and there is a great deal of subjective intention, making it difficult to guarantee large-scale and refined operational iterations. Summary of the Invention

[0003] The main objective of this application is to provide a method, apparatus, computer-readable storage medium, and electronic device for processing operational strategies for hotel OTA businesses, so as to at least solve the problem that the existing technology lacks the ability to access outputs of dimensions such as market, weather, and discounts when formulating operational strategies, resulting in the actual online effect not meeting expectations.

[0004] To achieve the above objectives, according to one aspect of this application, a method for processing operational strategies for hotel OTA business is provided, comprising: obtaining a preset discount amount and multiple preset influencing factors; generating at least one initial operational strategy based on the preset discount amount and the preset influencing factors using a pre-trained planning model; performing playback evaluation on each of the initial operational strategies using a pre-built risk pre-testing platform to obtain multiple strategy risk groups; and, if the operational revenue loss corresponding to the strategy risk group is less than a first threshold, determining the initial operational strategy with the maximum corresponding operational revenue as the target operational strategy, and launching the target operational strategy for execution, wherein the operational revenue is determined based on platform revenue, hotel revenue, and user discounts.

[0005] Optionally, before generating at least one initial operating strategy based on the preset discount amount and the preset influencing factors using a pre-trained planning model, the method further includes: obtaining historical transaction data, historical discount amounts, and historical operating strategies from the logs of the OTA platform, and determining the corresponding values ​​of the preset influencing factors in the logs to obtain historical influencing factors, wherein the preset influencing factors include regional traffic, weather changes, news events, and holiday data; determining historical platform revenue, historical hotel revenue, and historical user discounts based on the historical transaction data, and performing a weighted calculation on the historical platform revenue, the historical hotel revenue, and the historical user discounts to obtain historical operating revenue; training a large model using the historical discount amount and historical influencing factors as input data, the operating strategy as output, and the predicted operating revenue as output data, and using the historical operating strategy corresponding to the maximum historical operating revenue as label data to obtain the planning model.

[0006] Optionally, a pre-built risk assessment platform is used to replay and evaluate each of the initial operational strategies to obtain multiple strategy risk groups, including: obtaining historical transaction data and historical operational strategies from the logs of the OTA platform, and determining the corresponding values ​​of the preset influencing factors in the logs to obtain historical influencing factors; training a first simulation model based on the historical transaction data, the historical influencing factors, and the historical operational strategies, the first simulation model being used to simulate the market transaction process under the influence of different combinations of the influencing factors and operational strategies; inputting the initial operational strategies and the preset influencing factors into the first simulation model to obtain a predicted transaction process; and determining the risk points existing in the initial operational strategies based on the predicted user behavior and the predicted transaction process to obtain strategy risk groups, the risk points including user churn, order loss, and revenue decline.

[0007] Optionally, after the target operation strategy is deployed and executed, the method further includes: real-time monitoring of the preset influencing factors and the preset discount amount; if the fluctuation of the preset influencing factors and / or the preset discount amount is greater than a second threshold, determining the fluctuation amount of the preset influencing factors and / or the preset discount amount to obtain a first deviation amount; inputting the first deviation amount into a pre-trained correction model to obtain a corrected operation strategy, wherein the correction model is used to generate correction rules for the target operation strategy based on the changes of the influencing factors and / or the preset discount amount; deploying and executing the corrected operation strategy, and updating the version number of the target operation strategy.

[0008] Optionally, after the modified operation strategy is deployed and implemented, the method further includes: monitoring the operation revenue within a first preset time period after the modified operation strategy is deployed and implemented, and obtaining the monitored operation revenue; if the deviation between the monitored operation revenue and the predicted operation revenue of the target operation strategy is greater than or equal to a third threshold, performing a rollback operation on the target operation strategy according to the current version number of the target operation strategy.

[0009] Optionally, after determining the initial operational strategy that corresponds to the maximum operational revenue as the target operational strategy, the method further includes: determining monitoring indicators under the user dimension, hotel dimension, and market dimension respectively to obtain a monitoring dictionary; determining the quality inspection rules for each monitoring indicator in the monitoring dictionary to obtain a target rule set; determining the benchmark values ​​of the quality inspection rules in the target rule set based on the preset discount amount and the preset influencing factors through multiple baseline prediction models to obtain multiple benchmark value groups, wherein each benchmark value group corresponds one-to-one with the baseline prediction model; determining the historical transaction data under the action of the target operational strategy when the target operational strategy exceeds its effective time limit; calculating the first deviation rate between the historical transaction data and each of the benchmark value groups respectively, wherein the first deviation rate is the sum of the second deviation rates between the historical transaction data and each benchmark value in the benchmark value group; determining the abnormal impact level of the target operational strategy based on all the first deviation rates; and generating an alarm message when the abnormal impact level is greater than or equal to a fourth threshold.

[0010] Optionally, generating alarm information includes: calculating the first impact and first contribution of each monitoring indicator in the monitoring dictionary to the operating revenue; determining the corresponding monitoring indicator as an abnormal indicator when the first impact and / or the first contribution is greater than or equal to a fifth threshold; randomly combining the abnormal indicator and its associated indicators to obtain multiple indicator combinations; calculating the second impact and second contribution of each indicator combination to the operating revenue; determining the corresponding indicator combination as an abnormal combination when the second impact and / or the second contribution is greater than or equal to a sixth threshold; and generating the alarm information based on the abnormal indicators and the abnormal combinations.

[0011] According to another aspect of this application, an operational strategy processing device for hotel OTA business is provided. The device includes: a first acquisition unit, configured to acquire a preset discount amount and multiple preset influencing factors, and generate at least one initial operational strategy based on the preset discount amount and the preset influencing factors using a pre-trained planning model; a first calculation unit, configured to perform playback evaluation on each of the initial operational strategies through a pre-built risk assessment platform to obtain multiple strategy risk groups; and a first determination unit, configured to determine the initial operational strategy with the maximum corresponding operational revenue as the target operational strategy when the operational revenue loss corresponding to the strategy risk group is less than a first threshold, and to put the target operational strategy on the platform for execution, wherein the operational revenue is determined based on platform revenue, hotel revenue, and user discounts.

[0012] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device on which the computer-readable storage medium is located to perform any of the methods described.

[0013] According to another aspect of this application, an electronic device is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including methods for performing any one of the methods described.

[0014] Applying the technical solution of this application, the above-mentioned method for processing operational strategies for hotel OTA business involves the following steps: First, obtaining preset discount limits and multiple preset influencing factors, and generating at least one initial operational strategy based on these preset discount limits and influencing factors using a pre-trained planning model. Then, evaluating each initial operational strategy through a pre-built risk pre-verification platform to obtain multiple strategy risk groups. Finally, if the operational revenue loss corresponding to a strategy risk group is less than a first threshold, the initial operational strategy with the highest corresponding operational revenue is determined as the target operational strategy, and the target operational strategy is launched and executed. The operational revenue is determined based on platform revenue, hotel revenue, and user discounts. This application incorporates multiple preset influencing factors and preset discount limits into operational considerations through a large model to generate operational strategies, avoiding the subjective influence of human intervention. Simultaneously, the risk pre-verification platform verifies the operational strategies to be launched, ensuring the accuracy of the expected strategy. This addresses the problem in existing technologies where the lack of access to outputs from dimensions such as market, weather, and discounts during operational strategy formulation leads to actual launch results that do not meet expectations. Attached Figure Description

[0015] Figure 1A hardware structure block diagram of a mobile terminal for processing the above-described operational strategy for hotel OTA business, according to an embodiment of this application, is shown.

[0016] Figure 2 A flowchart illustrating an operational strategy processing method for hotel OTA business according to an embodiment of this application is shown.

[0017] Figure 3 A structural block diagram of an operation strategy processing apparatus for hotel OTA business provided according to an embodiment of this application is shown. Detailed Implementation

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0020] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0021] As described in the background section, the generation, verification, and deployment of operational strategies for hotel OTA businesses in the prior art rely on manual configuration and offline calculations, making it difficult to achieve multi-dimensional systematic perception and structured access to the market. When there are many application factors to consider, the process cost of manual configuration is high and there is a great deal of subjective intention, making it difficult to guarantee large-scale and refined operational iterations. In order to solve the problem that the prior art lacks the ability to access the output of dimensions such as market, weather, and discounts when formulating operational strategies, resulting in the actual online effect not meeting expectations, the embodiments of this application provide an operational strategy processing method, apparatus, computer-readable storage medium, and electronic device for hotel OTA businesses.

[0022] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0023] The methods and embodiments provided in this application can be executed on a mobile terminal, computer terminal, or similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for a hotel OTA business operation strategy processing method according to an embodiment of the present invention. Figure 1 As shown, a mobile terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0024] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the hotel OTA business operation strategy processing method in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above-described networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the above-described networks may include wireless networks provided by the mobile terminal's communication provider. In one example, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0025] This embodiment provides an operational strategy processing method for hotel OTA business that runs on a mobile terminal, computer terminal or similar computing device. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although the logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0026] Figure 2 This is a flowchart of an operational strategy processing method for hotel OTA business according to an embodiment of this application. Figure 2 As shown, the method includes the following steps:

[0027] Step S201: Obtain the preset discount amount and multiple preset influencing factors, and generate at least one initial operation strategy based on the preset discount amount and preset influencing factors using a pre-trained planning model;

[0028] Understandably, large models, especially pre-trained language models, can learn from massive amounts of data to understand complex business logic and market dynamics. In this scenario, the model is trained to consider factors such as discount amounts, business district traffic, seasonal demand fluctuations, and competitor strategies in the hotel OTA business. Based on these factors and a certain discount amount, it predicts the possible returns under different strategies and market reactions, thereby generating the optimal or preliminary operational strategy.

[0029] In practice, when a new hotel joins the platform or the market environment changes, the model will automatically generate a series of initial operating strategies based on the latest data input, including preset discount amounts and influencing factors.

[0030] Step S202: The initial operational strategies are replayed and evaluated through a pre-built risk pre-verification platform to obtain multiple strategy risk groups;

[0031] Understandably, risk assessment platforms use historical data playback to simulate the effects of initial operational strategies, thereby evaluating the risks and benefits of those strategies. By comparing changes in returns and potential losses under different strategies, high-risk strategies and safe, profitable strategies can be identified.

[0032] In practice, a simulation environment is built based on historical transaction data, user behavior data, and market environment data to simulate possible scenarios after strategy implementation. Each initial strategy is applied to the replay environment to calculate its impact on hotel revenue, platform revenue, and user experience, generating a detailed evaluation report. The evaluation report should include strategy effectiveness indicators, revenue / loss predictions, and user experience feedback predictions. The risk points that arise during the pre-testing process for each initial operational strategy are compiled into groups, resulting in the aforementioned strategy risk groups.

[0033] Step S203: If the loss of operating revenue corresponding to the strategy risk group is less than the first threshold, the initial operating strategy with the maximum corresponding operating revenue is determined as the target operating strategy, and the target operating strategy is put on the platform for execution. The operating revenue is determined based on platform revenue, hotel revenue and user discounts.

[0034] Understandably, the system employs a strategy selection mechanism based on the principles of maximizing profits and controlling risks. By setting a first threshold, i.e., the maximum allowable percentage of profit loss, the system can filter out strategies that are both profitable and risk-controlled.

[0035] In practice, based on business objectives and risk tolerance, a reasonable revenue / loss threshold is set, namely the aforementioned first threshold. Then, from the initial operational strategies that meet the criteria, the strategy that can bring the greatest expected revenue is further selected and determined as the final target operational strategy. Once the target operational strategy is determined, the system automatically deploys and executes it, monitoring the actual effect to ensure it matches the pre-assessed results. If any deviation is detected, an adjustment process is immediately initiated.

[0036] In this embodiment, firstly, a preset discount amount and multiple preset influencing factors are obtained. A pre-trained planning model generates at least one initial operational strategy based on these factors. Then, a pre-built risk assessment platform is used to replay and evaluate each initial operational strategy, resulting in multiple strategy risk groups. Finally, if the operational revenue loss corresponding to a strategy risk group is less than a first threshold, the initial operational strategy with the highest corresponding operational revenue is determined as the target operational strategy, and this strategy is then implemented. The operational revenue is determined based on platform revenue, hotel revenue, and user discounts. This application incorporates multiple preset influencing factors and preset discount amounts into operational considerations through a large model to generate operational strategies, avoiding the subjective influence of human intervention. Simultaneously, the risk assessment platform verifies the operational strategies to be launched, ensuring the accuracy of the expected strategies. This addresses the problem in existing technologies where the lack of access to outputs from dimensions such as market, weather, and discounts during operational strategy formulation leads to unsatisfactory actual results.

[0037] To train the aforementioned planning model, in one optional implementation, before generating at least one initial operating strategy based on preset discount amounts and preset influencing factors using the pre-trained planning model, the method further includes:

[0038] Step S301: Obtain historical transaction data, historical discount amount and historical operation strategy from the logs of the OTA platform, and determine the corresponding values ​​of preset influencing factors in the logs to obtain historical influencing factors. The preset influencing factors include regional traffic, weather changes, news events and holiday data.

[0039] Understandably, based on the feature engineering principles of big data processing and machine learning, and through NLP and data mining techniques, historical log data from OTA platforms is collected and cleaned to extract transaction data (such as order volume and transaction amount), discount amounts, and operational strategy information related to promotional strategies. Secondly, the expression of preset influencing factors, such as regional traffic, weather changes, news events, and holidays, in the logs is identified and transformed into numerical or categorical features.

[0040] In practical implementation, big data technologies such as Hadoop or Spark are used to batch pull historical data from the OTA platform's log system, including but not limited to transaction records, user behavior records, hotel information, and weather API interface data. The collected data is cleaned using SQL queries or tools such as Python / Pandas to remove invalid or abnormal data, while unstructured data such as dates and weather descriptions are transformed into structured features. Features are constructed based on preset influencing factors such as regional traffic, weather changes, news events, and holiday data. For example, regional traffic is quantified as the number of user visits or page views within a certain time window; weather changes are transformed into numerical features such as temperature, humidity, and precipitation; news events are transformed into positive, negative, or neutral sentiment scores through sentiment analysis; and holiday data is transformed into a binary feature indicating whether it is a holiday.

[0041] Step S302: Determine historical platform revenue, historical hotel revenue, and historical user discounts based on historical transaction data, and calculate historical operating revenue by weighting the historical platform revenue, historical hotel revenue, and historical user discounts.

[0042] Understandably, weighted calculations are used to balance the importance and influence of each revenue component, and the optimal operational revenue strategy is found by comprehensively considering the revenue of the platform, hotels, and users.

[0043] In practice, based on historical transaction data, the platform's total revenue, the hotel's total revenue, and the total discounts received by users over a past period are calculated separately. The weights of platform revenue, hotel revenue, and user discounts are set according to business strategy and market conditions. For example, during promotional periods, the weight of user discounts may be increased to attract more customers; while during off-seasons, the weight of hotel revenue may be emphasized to ensure hotel profits. Each revenue component is multiplied by its corresponding weight and then summed to obtain historical operating revenue.

[0044] Step S303: Using historical discount amounts and historical influencing factors as input data, operational strategies and predicted operational revenue as output data, and using historical operational strategies corresponding to the maximum historical operational revenue as label data, train a large model to obtain a planning large model.

[0045] Understandably, by using labeled supervised learning algorithms, such as random forests, gradient boosting trees, or deep neural networks, models learn how to predict the optimal operating strategy and the corresponding operating revenue based on historical discount amounts and influencing factors. In this application, the labeled data uses the specific strategy corresponding to the maximum historical operating revenue to guide the model in learning how to find the point of maximum revenue under similar conditions.

[0046] In practical implementation, based on business characteristics and data scale, appropriate machine learning models are selected for training, such as decision trees, LSTM (Long Short-Term Memory) networks, or Transformers, which can handle time-series data or complex relational data. The historical dataset is split into training and test sets. Supervised learning is employed, using historical discount amounts and influencing factors as input, and the optimal operating strategy and its corresponding operating returns as output, for model training. During training, hyperparameters need to be adjusted multiple times to optimize the model's predictive ability. The model's performance is evaluated on the test set, using metrics such as accuracy, recall, and F1 score. If the model's performance is unsatisfactory, the model parameters are adjusted again or different algorithms are tried until satisfactory predictive results are achieved.

[0047] The above embodiments have achieved full automation of the process from data collection and profit calculation to model training. The introduction of large-scale planning models not only solves the limitations and subjectivity of traditional manual decision-making, but also enables timely and accurate strategy adjustments to adapt to rapid changes in the market environment.

[0048] To ensure the accuracy of the product listing and operation strategy, in one optional implementation, step S202 includes:

[0049] Step S2021: Obtain historical transaction data, user behavior data, and historical operation strategies from the logs of the OTA platform, and determine the corresponding values ​​of preset influencing factors in the logs to obtain historical influencing factors;

[0050] Understandably, mining historical data can reveal the intrinsic connections between past market dynamics, user behavior patterns, and operational strategies. By cleaning and organizing log data, transforming it into a structured and analyzable format, and then using Natural Language Processing (NLP) and data mining techniques, numerical representations of predefined influencing factors can be identified, providing a basis for subsequent model training and strategy evaluation.

[0051] In practice, files containing historical data such as transaction details, user activity records, and operational strategy changes are extracted from the platform's log system. Unstructured or semi-structured log data is transformed into structured data, and outliers and irrelevant information are removed through data cleaning. Simultaneously, feature engineering is performed to convert preset influencing factors (such as weather changes, holidays, and news events) into numerical features that can be directly input into the model. Mapping rules between preset influencing factors and log data are determined; for example, the number of times a news event is mentioned is converted into a numerical value representing its degree of influence, or holiday information is converted into specific date stamps.

[0052] Step S2022: Train the first simulation model based on historical transaction data, historical influencing factors and historical operating strategies. The first simulation model is used to simulate the market transaction process under the influence of different combinations of influencing factors and operating strategies.

[0053] Understandably, the first simulation model aims to simulate the complex processes of market trading, particularly market responses under the interaction of various influencing factors and operational strategies. By integrating historical trading data, influencing factors, and operational strategies, the model is trained to predict overall market reactions, such as order volume, transaction value, and refund rates.

[0054] In practice, suitable machine learning algorithms for predicting market transactions are selected, which may include, but are not limited to, ensemble learning and deep learning networks, to adapt to time series analysis and high-dimensional feature processing. A training set is constructed using a combination of historical transaction data, influencing factors, and operational strategies to train a first simulation model to predict changes in the market transaction process. Similarly, methods such as cross-validation are used to evaluate the model's ability to predict the market transaction process, and the model is fine-tuned based on the evaluation results.

[0055] Step S2024: Input the initial operating strategy and preset influencing factors into the first simulation model to obtain the predicted transaction process;

[0056] In practice, the initial operational strategy to be evaluated, including factors such as discount amounts and activity durations, is provided as input to the model. The first simulation model is then used to predict the market transaction process under the new strategy. The model output is interpreted to identify which strategies might lead to user churn, reduced order volume, or decreased transaction value, thus forming a preliminary risk assessment.

[0057] Step S2025: Based on the predicted user behavior and the predicted transaction process, determine the risk points of the initial operation strategy and obtain the strategy risk group. The risk points include user churn, order loss and revenue decline.

[0058] Understandably, based on model predictions, key risk indicators such as user churn, order loss, and declining revenue are identified through analysis of user behavior and market transaction processes in order to assess potential risks associated with strategy implementation.

[0059] Through the above embodiments, a data-driven risk prediction platform is constructed. It can not only automatically assess the potential risks of operational strategies, but also provide refined risk analysis, such as the reasons for user churn, the specific scenarios of order loss, and possible factors contributing to declining revenue. This allows for more flexible and precise strategy adjustments during market fluctuations, avoiding unnecessary financial losses while improving user satisfaction and market competitiveness.

[0060] To ensure the accuracy of the operational strategy, in one optional implementation, after the target operational strategy is deployed and executed, the above method further includes:

[0061] Step S401: Monitor preset influencing factors and preset discount amounts in real time;

[0062] Understandably, real-time monitoring based on data analytics and stream processing technologies is used to capture any sudden changes in the market environment or promotional strategies, with the aim of continuously tracking key variables affecting the hotel OTA business.

[0063] In practical implementation, a real-time data stream processing system, such as Apache Kafka or Kinesis, is built on the platform to collect and transmit log data, weather API data, news events, and other information in real time. Upon arrival, the data is parsed using a stream processing framework (such as Apache Flink or Spark Streaming), transforming it into an analytically usable format and storing it in a high-speed data warehouse (such as Elasticsearch or Amazon Redshift) for later use. Real-time monitoring scripts or applications are developed to periodically check the latest values ​​of preset influencing factors and discount amounts, comparing them with established benchmark values. A monitoring threshold (second threshold) is set for each monitored item to measure whether its changes exceed the normal fluctuation range. Once a fluctuation exceeding the second threshold is detected, the subsequent correction process is immediately initiated.

[0064] Step S402: If the fluctuation of the preset influencing factors and / or the preset discount amount is greater than the second threshold, determine the amount of fluctuation of the preset influencing factors and / or the preset discount amount to obtain the first deviation amount;

[0065] In practice, historical averages or recent stable period data for preset influencing factors and discount amounts are retrieved from the data warehouse. The current real-time monitored value is subtracted from the historical value to obtain the first deviation. If multi-dimensional influencing factors are involved, weighted summation or compound deviation calculation should also be considered. The magnitude, direction, and persistence of the deviation are analyzed to provide input data for revising the large model, ensuring the rationality and effectiveness of the revision strategy.

[0066] Step S403: Input the first deviation into the pre-trained correction model to obtain the correction operation strategy. The correction model is used to generate correction rules for the target operation strategy based on changes in influencing factors and / or preset discount amounts.

[0067] Understandably, large-scale correction models are based on deep learning or reinforcement learning techniques. By learning the relationship between fluctuations in historical data and correction strategies, they can predict the optimal strategy adjustment under the current deviation.

[0068] In practice, the first deviation is used as the input feature of the model (time stamps, market trends and other auxiliary information can also be introduced to enhance the prediction accuracy of the model). The large model is corrected based on the first deviation of the input, and the corrected operation strategy is output, including but not limited to adjusting the discount amount, repositioning the target user group, and changing the promotional activities.

[0069] In addition, to ensure the rationality and controllability of strategy adjustments, the model output should include an explanation of the logic behind the adjustments and the expected effects, so that operations personnel can understand and accept them.

[0070] Step S404 involves revising the operational strategy, implementing it, and updating the version number of the target operational strategy.

[0071] In practice, the revised operational strategy is uploaded to the execution environment via API or the management backend to ensure that all relevant systems can receive and execute the new strategy. Information about this strategy adjustment is recorded in the strategy management system, including the adjustment time, content, and reason, while the strategy version number is updated, for example, from v1.0 to v1.1. After the strategy is implemented, market response and business revenue are continuously monitored, and data is collected to evaluate the effectiveness of the strategy adjustment, providing a basis for subsequent strategy optimization.

[0072] The above embodiments introduce a dynamic strategy adjustment mechanism, which can capture market changes in real time and make rapid and effective strategy corrections accordingly, greatly improving operational efficiency and revenue stability. The combination of the real-time monitoring system and the large-scale correction model automates and intelligently formulates strategies, reducing reliance on manual intervention and making strategy adjustments more accurate and timely.

[0073] To further ensure the accuracy of the operational strategy, in one optional implementation, after the revised operational strategy is deployed and implemented, the above method further includes:

[0074] Step S501: Monitor the operational revenue within the first preset time period after the revised operational strategy is implemented, and obtain the monitored operational revenue;

[0075] Understandably, real-time monitoring systems typically utilize big data processing frameworks, such as Apache Kafka, Apache Flink, or Kafka Streams, to process and analyze continuously flowing data streams in order to enable real-time monitoring and data analysis of operational revenue, with the aim of evaluating the actual performance after strategy adjustments.

[0076] In practice, the aforementioned first preset duration is typically a short time window after strategy execution, such as a few hours or a day, determined based on business needs and market response time. A real-time data stream processing pipeline is established to collect transaction data, user behavior data, and market environment data after the implementation of the revised operational strategy, ensuring data real-time performance and completeness. A calculation module is designed to process the transaction data in the real-time data stream, calculating the operational revenue after the revised strategy is implemented, including platform revenue, hotel revenue, and the combined revenue from user discounts. At the end of the monitoring window, the operational revenue after the implementation of the revised strategy is calculated (monitoring operational revenue), and compared with the predicted operational revenue of the target operational strategy to evaluate the effectiveness of the strategy adjustment.

[0077] Step S502: If the deviation between the monitored operating revenue and the predicted operating revenue of the target operating strategy is greater than or equal to the third threshold, the target operating strategy is rolled back according to the current version number of the target operating strategy.

[0078] In practice, a reasonable third threshold is set based on business risk tolerance and historical data fluctuation range to determine whether the strategy execution effect is lower than expected. When the deviation between monitored and predicted operating revenue exceeds the third threshold, the system automatically initiates a rollback mechanism to restore the previous stable version of the target operating strategy, thus mitigating losses in a timely manner. The version number of each strategy change is recorded in the strategy management system to ensure that the rollback operation can accurately restore to the specified version. After the rollback operation is executed, the system should send a notification to the relevant operations personnel, informing them of the reason and result of the rollback, and record the details of the rollback operation in the log for later analysis and problem localization.

[0079] The above embodiments achieve automation of strategy adjustments and real-time feedback on effectiveness evaluation. By monitoring the operational benefits after the implementation of corrective strategies in real time, the actual effectiveness of new strategies can be quickly verified, problems can be identified early, and rollback measures can be taken, ensuring the stability of business operations and maximizing profits.

[0080] To dynamically update the operational strategy, in one optional implementation, after determining the initial operational strategy that maximizes the corresponding operational revenue as the target operational strategy, the method further includes:

[0081] Step S601: Determine the monitoring indicators for the user dimension, hotel dimension, and market dimension respectively to obtain the monitoring dictionary;

[0082] In practical implementation, clearly define user-level metrics, such as user activity and user satisfaction ratings; hotel-level metrics, including room sales and average daily revenue per minute (ADR); and market-level metrics, covering market share and competition index. Organize the above-defined metrics into a monitoring dictionary, with each metric including its calculation method, data source, and monitoring frequency to ensure the operability of the metrics and the effectiveness of monitoring.

[0083] Step S602: Determine the quality inspection rules for each monitoring indicator in the monitoring dictionary to obtain the target rule set;

[0084] In practice, the normal fluctuation range (or other verification method) of each monitoring indicator is determined from the monitoring dictionary to obtain the above target rule set.

[0085] Step S603: Determine the benchmark values ​​of the quality inspection rules in the target rule set based on the preset discount amount and preset influencing factors through multiple baseline prediction models, and obtain multiple benchmark value groups, which correspond one-to-one with the baseline prediction models.

[0086] Understandably, the baseline prediction model is used to build a predictive model of business performance under normal circumstances. By learning from historical data, it can predict the expected business performance under the current discount amount and market environment, providing a quantitative standard for the execution of quality inspection rules.

[0087] In practice, based on historical data, multiple prediction models, such as linear regression, random forest, and LSTM (Long Short-Term Memory network), are selected and trained to predict baseline values ​​for indicators such as user activity, hotel sales, and market competitiveness. Using the trained models, and inputting the current discount amount and market environment factors, predicted values ​​for each monitoring indicator are obtained, forming a baseline value set.

[0088] Step S604: If the target operation strategy exceeds its effective time limit, determine the historical transaction data under the effect of the target operation strategy;

[0089] In practice, historical transaction data, including order volume, transaction amount, and user behavior, is retrieved from the database or log system based on the strategy's execution time period. The retrieved data is then organized to ensure its format and content match the calculation logic of the monitoring indicators, facilitating subsequent deviation rate calculations.

[0090] Step S605: Calculate the first deviation rate between the historical transaction data and each benchmark value group respectively. The first deviation rate is the sum of the second deviation rates between the historical transaction data and each benchmark value in the benchmark value group.

[0091] In practice, for each monitoring indicator, the deviation between its actual historical transaction data and the benchmark value is calculated, expressed as a percentage deviation: (First Deviation Rate = (Actual Value - Benchmark Value) / Benchmark Value) (100%), sum the first deviation rates of all monitoring indicators to obtain the overall deviation rate of strategy execution.

[0092] Step S606: Determine the abnormal impact level of the target operation strategy based on all first deviation rates. If the abnormal impact level is greater than or equal to the fourth threshold, generate alarm information.

[0093] In practice, the anomaly impact level is a quantitative assessment of the degree of impact on the strategy. A fourth threshold is set to identify significant anomalies during strategy execution. Once the anomaly impact level reaches or exceeds the set threshold, the system will trigger an alarm mechanism, alerting operations personnel to intervene promptly and prevent potential business losses. Specifically, the anomaly impact level can be calculated using rules, such as categorizing it into minor, moderate, and severe levels based on the magnitude of the first deviation rate.

[0094] Through the above embodiments, a closed-loop management system for operational strategies, encompassing monitoring, evaluation, and feedback, was achieved, ensuring accurate monitoring and timely adjustments to strategy effectiveness. The construction of a monitoring dictionary and quality rule set guarantees data accuracy and monitoring effectiveness; the application of a baseline prediction model provides objective quantitative standards for strategy effectiveness evaluation; and the setting of anomaly impact levels and alarm mechanisms can identify and respond promptly to significant anomalies in strategy execution, preventing potential business losses.

[0095] In order to generate the above alarm information, in one optional implementation, step S606 includes:

[0096] Step S6061: Calculate the first impact and first contribution of each monitoring indicator in the monitoring dictionary to operating revenue.

[0097] Understandably, the first impact measure aims to quantify the degree to which a change in a single monitored indicator affects operating revenue. This can be achieved using methods such as regression analysis, causal inference models, or difference analysis, revealing the quantitative relationship between each indicator and revenue. The first contribution measure further analyzes the relative importance of the monitored indicator to changes in revenue. The contribution measure is typically calculated based on the correlation between the magnitude of the indicator's change and the magnitude of the revenue change, using statistical methods such as principal component analysis (PCA) and eigenvalue analysis.

[0098] Step S6062: If the first influence and / or the first contribution are greater than or equal to the fifth threshold, the corresponding monitoring indicator is determined as an abnormal indicator.

[0099] Understandably, based on business characteristics and risk appetite, a fifth threshold is set to define which metric changes have an abnormal impact on operating revenue. The threshold setting needs to balance revenue sensitivity and strategy stability. By comparing the first impact or first contribution of a metric with the fifth threshold, those monitoring metrics that have an unusually strong impact on operating revenue—these are the abnormal metrics.

[0100] Step S6063: Randomly combine the abnormal indicators and their associated indicators to obtain multiple indicator combinations;

[0101] Understandably, by randomly combining outlier indicators with their potential related indicators, it is possible to explore the impact of different indicator combinations on operating revenue, helping to reveal complex interactions and supporting in-depth problem analysis. Using combinatorial algorithms in statistics, such as Monte Carlo simulation, different combinations of outlier indicators and other potential related indicators are generated. For each generated indicator combination, its overall impact on operating revenue is assessed, identifying which combinations may produce synergistic or contradictory effects, affecting revenue performance.

[0102] Step S6064: Calculate the second impact and second contribution of each indicator combination on operating revenue.

[0103] Understandably, calculating the second influence and second contribution of a combination of indicators aims to quantify the effect of multiple indicators working together. This requires more advanced statistical models, such as generalized linear models (GLM), multiple nonlinear regression, or deep learning models, to handle the complex correlations between multiple variables.

[0104] Step S6065: If the second influence and / or the second contribution are greater than or equal to the sixth threshold, the corresponding index combination is determined to be an abnormal combination.

[0105] Understandably, based on an understanding of tolerance for business fluctuations, a sixth threshold is set to define which combinations of indicators have an unusual impact on operating revenue. By comparing the second impact or second contribution of indicator combinations with the sixth threshold, abnormal combinations—those that have a significant impact on revenue changes—are screened out, providing trigger conditions for subsequent alert mechanisms.

[0106] Step S6066: Generate alarm information based on abnormal indicators and abnormal combinations.

[0107] Understandably, the generation of alarm information is based on the anomaly identification results of monitoring indicators and indicator combinations, and aims to quickly communicate information on high-risk or abnormal profit changes to the operations team so that timely countermeasures can be taken.

[0108] The above embodiments enable comprehensive monitoring and analysis of factors affecting operating revenue, ensuring timely response and accurate identification of abnormal revenue fluctuations. This mechanism not only improves the transparency of revenue management but also promotes the scientific and flexible nature of strategy optimization, helping to reduce ineffective or harmful strategy adjustments and enhance overall profitability and market competitiveness.

[0109] To enable those skilled in the art to better understand the technical solution of this application, the implementation process of the operational strategy processing method for hotel OTA business of this application will be described in detail below with reference to specific embodiments.

[0110] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0111] This application also provides an operation strategy processing apparatus for hotel OTA business. It should be noted that this apparatus can be used to execute the operation strategy processing method for hotel OTA business provided in this application. This apparatus is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0112] The following describes the operation strategy processing device for hotel OTA business provided in the embodiments of this application.

[0113] Figure 3 This is a structural block diagram of an operation strategy processing device for hotel OTA business according to an embodiment of this application. Figure 3 As shown, the device includes:

[0114] The first acquisition unit 10 is used to acquire the preset discount amount and multiple preset influencing factors, and generate at least one initial operation strategy based on the preset discount amount and preset influencing factors through a pre-trained planning big model;

[0115] The first calculation unit 20 is used to replay and evaluate each initial operation strategy through a pre-built risk pre-verification platform to obtain multiple strategy risk groups;

[0116] The first determining unit 30 is used to determine the initial operating strategy with the maximum corresponding operating revenue as the target operating strategy when the operating revenue loss corresponding to the strategy risk group is less than the first threshold, and to put the target operating strategy on the platform for execution. The operating revenue is determined based on platform revenue, hotel revenue and user discounts.

[0117] In this embodiment, the first acquisition unit acquires a preset discount amount and multiple preset influencing factors, and generates at least one initial operation strategy based on the preset discount amount and preset influencing factors using a pre-trained planning model. The first calculation unit replays and evaluates each initial operation strategy through a pre-built risk prediction platform to obtain multiple strategy risk groups. The first determination unit determines the initial operation strategy with the maximum corresponding operational revenue as the target operation strategy when the operational revenue loss corresponding to the strategy risk group is less than a first threshold, and puts the target operation strategy on the platform for execution. The operational revenue is determined based on platform revenue, hotel revenue, and user discounts. This application incorporates multiple preset influencing factors and preset discount amounts into operational considerations through a large model to generate operation strategies, avoiding the subjective influence of human intervention. At the same time, the operation strategy to be launched is verified based on the risk prediction platform to ensure the accuracy of the strategy expectation, thereby solving the problem in the prior art that the lack of access to the output of dimensions such as market, weather, and discounts when formulating operation strategies leads to the actual launch effect not meeting expectations.

[0118] To train the aforementioned large-scale planning model, in one optional implementation, the apparatus further includes:

[0119] The second acquisition unit is used to acquire historical transaction data, historical discount amounts, and historical operating strategies from the logs of the OTA platform before generating at least one initial operating strategy based on the preset discount amount and preset influencing factors through the pre-trained planning model. It also determines the corresponding values ​​of the preset influencing factors in the logs and obtains the historical influencing factors. The preset influencing factors include regional traffic, weather changes, news events, and holiday data.

[0120] The second calculation unit is used to determine historical platform revenue, historical hotel revenue, and historical user discounts based on historical transaction data, and to perform weighted calculations on historical platform revenue, historical hotel revenue, and historical user discounts to obtain historical operating revenue.

[0121] The first training unit is used to train a large model with historical discount amounts and historical influencing factors as input data, operational strategies as output data and predicted operational revenue as output data, and historical operational strategies corresponding to the maximum historical operational revenue as label data, thus obtaining a large planning model.

[0122] To ensure the accuracy of the product launch and operation strategy, in one optional implementation, the first calculation unit includes:

[0123] The first acquisition module is used to acquire historical transaction data, user behavior data and historical operation strategies from the logs of the OTA platform, and determine the corresponding values ​​of preset influencing factors in the logs to obtain historical influencing factors.

[0124] The first training module is used to train the first simulation model based on historical transaction data, historical influencing factors and historical operating strategies. The first simulation model is used to simulate the market transaction process under the influence of different combinations of influencing factors and operating strategies.

[0125] The first input module is used to input the initial operating strategy and preset influencing factors into the first simulation model to obtain the predicted transaction process;

[0126] The first determination module is used to identify the risk points of the initial operation strategy based on the predicted user behavior and the predicted transaction process, and obtain the strategy risk group. The risk points include user churn, order loss and revenue decline.

[0127] To ensure the accuracy of the operational strategy, in one optional implementation, the above-mentioned apparatus further includes:

[0128] The third acquisition unit is used to monitor preset influencing factors and preset discount amounts in real time after the target operation strategy is put on the platform and executed.

[0129] The second determining unit is used to determine the amount of fluctuation of the preset influencing factors and / or the preset discount amount when the fluctuation of the preset influencing factors and / or the preset discount amount is greater than the second threshold, and obtain the first deviation amount.

[0130] The third calculation unit is used to input the first deviation into the pre-trained correction model to obtain the correction operation strategy. The correction model is used to generate correction rules for the target operation strategy based on changes in influencing factors and / or preset discount amounts.

[0131] The first processing unit is used to implement the revised operational strategy and update the version number of the target operational strategy.

[0132] To further ensure the accuracy of the operational strategy, in one optional implementation, the above-mentioned device further includes:

[0133] The fourth acquisition unit is used to monitor the operational revenue within the first preset time period after the modified operational strategy is put on the platform and executed, and to obtain the monitored operational revenue.

[0134] The second processing unit is used to roll back the target operating strategy based on the current version number of the target operating strategy if the deviation between the monitored operating revenue and the predicted operating revenue of the target operating strategy is greater than or equal to a third threshold.

[0135] In order to dynamically update the operational strategy, in one optional implementation, the above-mentioned apparatus further includes:

[0136] The fourth calculation unit is used to determine the monitoring indicators under the user dimension, hotel dimension and market dimension respectively after determining the initial operation strategy with the maximum corresponding operation revenue as the target operation strategy, and obtain the monitoring dictionary.

[0137] The third determining unit is used to determine the quality inspection rules for each monitoring indicator in the monitoring dictionary, thereby obtaining the target rule set;

[0138] The fourth determining unit is used to determine the benchmark values ​​of the quality inspection rules in the target rule set based on the preset discount amount and preset influencing factors through multiple baseline prediction models, thereby obtaining multiple benchmark value groups, which correspond one-to-one with the baseline prediction models.

[0139] The fifth determining unit is used to determine the historical transaction data under the effect of the target operation strategy when the effective time limit of the target operation strategy has been exceeded;

[0140] The fifth calculation unit is used to calculate the first deviation rate between the historical transaction data and each benchmark value group, and the first deviation rate is the sum of the second deviation rates between the historical transaction data and each benchmark value in the benchmark value group;

[0141] The generation unit is used to determine the abnormal impact level of the target operation strategy based on all first deviation rates, and generate alarm information when the abnormal impact level is greater than or equal to the fourth threshold.

[0142] In order to generate the above-mentioned alarm information, in one optional implementation, the generation unit includes:

[0143] The first calculation module is used to calculate the first impact and first contribution of each monitoring indicator in the monitoring dictionary to operating revenue.

[0144] The second determination module is used to determine the corresponding monitoring indicator as an abnormal indicator when the first influence and / or the first contribution is greater than or equal to the fifth threshold.

[0145] The processing module is used to randomly combine abnormal indicators and their associated indicators to obtain multiple indicator combinations.

[0146] The second calculation module is used to calculate the second impact and second contribution of each combination of indicators on operating revenue.

[0147] The third determination module is used to determine the corresponding indicator combination as an abnormal combination when the second influence and / or the second contribution is greater than or equal to the sixth threshold.

[0148] The generation module is used to generate alarm information based on abnormal indicators and abnormal combinations.

[0149] The aforementioned operational strategy processing device for hotel OTA business includes a processor and a memory. The first acquisition unit, the first calculation unit, and the first determination unit are all stored as program units in the memory, and the processor executes the program units stored in the memory to achieve the corresponding functions. All of the above modules are located in the same processor; alternatively, the modules may be located in different processors in any combination.

[0150] The processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and adjusting kernel parameters can improve the accuracy of operational strategies.

[0151] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0152] This invention provides a computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device containing the computer-readable storage medium to perform the aforementioned operational strategy processing method for hotel OTA business.

[0153] Specifically, the operational strategy for hotel OTA business includes:

[0154] Step S201: Obtain the preset discount amount and multiple preset influencing factors, and generate at least one initial operation strategy based on the preset discount amount and preset influencing factors using a pre-trained planning model;

[0155] Step S202: The initial operational strategies are replayed and evaluated through a pre-built risk pre-verification platform to obtain multiple strategy risk groups;

[0156] Step S203: If the loss of operating revenue corresponding to the strategy risk group is less than the first threshold, the initial operating strategy with the maximum corresponding operating revenue is determined as the target operating strategy, and the target operating strategy is put on the platform for execution. The operating revenue is determined based on platform revenue, hotel revenue and user discounts.

[0157] This invention provides a processor for running a program, wherein the program executes the above-mentioned operational strategy processing method for hotel OTA business.

[0158] Specifically, the operational strategy for hotel OTA business includes:

[0159] Step S201: Obtain the preset discount amount and multiple preset influencing factors, and generate at least one initial operation strategy based on the preset discount amount and preset influencing factors using a pre-trained planning model;

[0160] Step S202: The initial operational strategies are replayed and evaluated through a pre-built risk pre-verification platform to obtain multiple strategy risk groups;

[0161] Step S203: If the loss of operating revenue corresponding to the strategy risk group is less than the first threshold, the initial operating strategy with the maximum corresponding operating revenue is determined as the target operating strategy, and the target operating strategy is put on the platform for execution. The operating revenue is determined based on platform revenue, hotel revenue and user discounts.

[0162] This invention provides an electronic device, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it performs at least the following steps:

[0163] Step S201: Obtain the preset discount amount and multiple preset influencing factors, and generate at least one initial operation strategy based on the preset discount amount and preset influencing factors using a pre-trained planning model;

[0164] Step S202: The initial operational strategies are replayed and evaluated through a pre-built risk pre-verification platform to obtain multiple strategy risk groups;

[0165] Step S203: If the loss of operating revenue corresponding to the strategy risk group is less than the first threshold, the initial operating strategy with the maximum corresponding operating revenue is determined as the target operating strategy, and the target operating strategy is put on the platform for execution. The operating revenue is determined based on platform revenue, hotel revenue and user discounts.

[0166] This application also provides a computer program product, which, when executed on a data processing device, is suitable for executing an initialization program having at least the following method steps:

[0167] Step S201: Obtain the preset discount amount and multiple preset influencing factors, and generate at least one initial operation strategy based on the preset discount amount and preset influencing factors using a pre-trained planning model;

[0168] Step S202: The initial operational strategies are replayed and evaluated through a pre-built risk pre-verification platform to obtain multiple strategy risk groups;

[0169] Step S203: If the loss of operating revenue corresponding to the strategy risk group is less than the first threshold, the initial operating strategy with the maximum corresponding operating revenue is determined as the target operating strategy, and the target operating strategy is put on the platform for execution. The operating revenue is determined based on platform revenue, hotel revenue and user discounts.

[0170] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. They can be implemented using computer-executable program code, and thus can be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those described herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0171] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0172] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0173] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0174] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0175] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0176] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0177] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic tape, disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0178] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0179] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0180] As can be seen from the above description, the embodiments of this application achieve the following technical effects:

[0181] 1) The operational strategy processing method for hotel OTA business in this application first obtains preset discount limits and multiple preset influencing factors, and generates at least one initial operational strategy based on the preset discount limits and preset influencing factors through a pre-trained planning model; then, it replays and evaluates each initial operational strategy through a pre-built risk pre-verification platform to obtain multiple strategy risk groups; finally, if the operational revenue loss corresponding to the strategy risk group is less than a first threshold, the initial operational strategy with the maximum corresponding operational revenue is determined as the target operational strategy, and the target operational strategy is launched and executed. The operational revenue is determined based on platform revenue, hotel revenue, and user discounts. This application incorporates multiple preset influencing factors and preset discount limits into operational considerations through a large model to generate operational strategies, avoiding the subjective influence of human intervention. At the same time, it verifies the operational strategy to be launched based on the risk pre-verification platform to ensure the accuracy of the strategy expectations, thereby solving the problem in existing technologies that lack the ability to access outputs from dimensions such as market, weather, and discounts when formulating operational strategies, resulting in actual launch effects that do not meet expectations.

[0182] 2) The operational strategy processing device for hotel OTA business disclosed in this application comprises: a first acquisition unit acquiring preset discount amounts and multiple preset influencing factors, and generating at least one initial operational strategy based on the preset discount amounts and preset influencing factors using a pre-trained planning model; a first calculation unit replaying and evaluating each initial operational strategy through a pre-built risk pre-verification platform to obtain multiple strategy risk groups; and a first determination unit determining the initial operational strategy with the maximum corresponding operational revenue as the target operational strategy when the operational revenue loss corresponding to the strategy risk group is less than a first threshold, and then launching the target operational strategy for execution. The operational revenue is determined based on platform revenue, hotel revenue, and user discounts. This application incorporates multiple preset influencing factors and preset discount amounts into operational considerations through a large model to generate operational strategies, avoiding the subjective influence of human intervention. Simultaneously, it verifies the operational strategies to be launched based on the risk pre-verification platform to ensure the accuracy of the strategy expectations, thus solving the problem in existing technologies where the lack of access to outputs from dimensions such as market, weather, and discounts during operational strategy formulation leads to actual launch results not meeting expectations.

[0183] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for processing operational strategies for hotel OTA business, characterized in that, include: Obtain a preset discount amount and multiple preset influencing factors, and generate at least one initial operating strategy based on the preset discount amount and the preset influencing factors through a pre-trained planning model; By replaying and evaluating each of the initial operational strategies through a pre-built risk assessment platform, multiple strategy risk groups are obtained; If the loss of operating revenue corresponding to the strategy risk group is less than the first threshold, the initial operating strategy with the maximum corresponding operating revenue is determined as the target operating strategy, and the target operating strategy is put on the platform for execution. The operating revenue is determined based on platform revenue, hotel revenue, and user discounts.

2. The method according to claim 1, characterized in that, Before generating at least one initial operational strategy based on the preset discount amount and the preset influencing factors using a pre-trained planning model, the method further includes: Historical transaction data, historical discount amounts, and historical operational strategies are obtained from the logs of the OTA platform, and the values ​​of the preset influencing factors in the logs are determined to obtain the historical influencing factors. The preset influencing factors include regional traffic, weather changes, news events, and holiday data. Based on the historical transaction data, historical platform revenue, historical hotel revenue, and historical user discounts are determined. The historical operating revenue is obtained by weighting the historical platform revenue, historical hotel revenue, and historical user discounts. Using the historical discount amount and historical influencing factors as input data, the operational strategy and predicted operational revenue as output data, and the historical operational strategy corresponding to the maximum historical operational revenue as label data, a large model is trained to obtain the planning large model.

3. The method according to claim 1, characterized in that, By replaying and evaluating each of the initial operational strategies through a pre-built risk pre-verification platform, multiple strategy risk groups are obtained, including: Historical transaction data and historical operation strategies are obtained from the logs of the OTA platform, and the values ​​of the preset influencing factors in the logs are determined to obtain the historical influencing factors. A first simulation model is trained based on the historical transaction data, the historical influencing factors, and the historical operating strategies. The first simulation model is used to simulate the market transaction process under the influence of different combinations of the influencing factors and operating strategies. The initial operating strategy and the preset influencing factors are input into the first simulation model to obtain the predicted transaction process; Based on the predicted transaction process, the risk points of the initial operation strategy are determined, resulting in a strategy risk group. The risk points include user churn, order loss, and revenue decline.

4. The method according to claim 1, characterized in that, After the target operational strategy is deployed and implemented, the method further includes: Real-time monitoring of the preset influencing factors and the preset discount amount; If the fluctuation of the preset influencing factors and / or the preset discount amount is greater than the second threshold, the fluctuation amount of the preset influencing factors and / or the preset discount amount is determined to obtain the first deviation amount; The first deviation is input into the pre-trained correction model to obtain the correction operation strategy. The correction model is used to generate correction rules for the target operation strategy based on the changes in the influencing factors and / or the preset discount amount. The revised operational strategy will be deployed and implemented, and the version number of the target operational strategy will be updated.

5. The method according to claim 4, characterized in that, After implementing the revised operational strategy, the method further includes: The operational revenue is monitored within the first preset time period after the revised operational strategy is implemented, and the monitored operational revenue is obtained. If the deviation between the monitored operational revenue and the predicted operational revenue of the target operational strategy is greater than or equal to a third threshold, the target operational strategy is rolled back according to the current version number of the target operational strategy.

6. The method according to claim 1, characterized in that, After determining the initial operating strategy that maximizes the corresponding operating revenue as the target operating strategy, the method further includes: The monitoring indicators were determined separately for the user dimension, hotel dimension, and market dimension, resulting in a monitoring dictionary; Determine the quality inspection rules for each monitoring indicator in the monitoring dictionary to obtain the target rule set; The baseline values ​​of the quality inspection rules in the target rule set are determined by multiple baseline prediction models based on the preset discount amount and the preset influencing factors, resulting in multiple baseline value groups, each of which corresponds one-to-one with the baseline prediction model. If the target operation strategy exceeds its effective time limit, determine the historical transaction data under the effect of the target operation strategy; Calculate the first deviation rate between the historical transaction data and each of the benchmark value groups, where the first deviation rate is the sum of the second deviation rates between the historical transaction data and each benchmark value in the benchmark value group; The abnormal impact level of the target operation strategy is determined based on all the first deviation rates, and an alarm message is generated if the abnormal impact level is greater than or equal to the fourth threshold.

7. The method according to claim 6, characterized in that, Generate alarm information, including: Calculate the first impact and first contribution of each monitoring indicator in the monitoring dictionary to the operating revenue; If the first influence and / or the first contribution is greater than or equal to the fifth threshold, the corresponding monitoring indicator will be identified as an abnormal indicator. The abnormal indicators and their associated indicators are randomly combined to obtain multiple indicator combinations; Calculate the second impact and second contribution of each of the aforementioned indicator combinations on the operating revenue; If the second influence and / or the second contribution is greater than or equal to the sixth threshold, the corresponding combination of indicators will be identified as an abnormal combination. The alarm information is generated based on the abnormal indicators and the abnormal combinations.

8. An operational strategy processing device for hotel OTA business, characterized in that, The device includes: The first acquisition unit is used to acquire a preset discount amount and multiple preset influencing factors, and generate at least one initial operation strategy based on the preset discount amount and the preset influencing factors through a pre-trained planning model; The first calculation unit is used to replay and evaluate each of the initial operation strategies through a pre-built risk pre-verification platform to obtain multiple strategy risk groups; The first determining unit is configured to, when the operational revenue loss corresponding to the strategy risk group is less than a first threshold, determine the initial operational strategy with the maximum corresponding operational revenue as the target operational strategy, and put the target operational strategy on the platform for execution, wherein the operational revenue is determined based on platform revenue, hotel revenue and user discounts.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device on which the computer-readable storage medium is located to perform the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that, include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs comprising methods for performing any one of claims 1 to 7.

Images