AI content production platform for offline store marketing

Abstract

The present invention provides an AI content production platform for offline store marketing, and relates to the technical field of store marketing strategies. The present invention uses an environment acquisition module and a preprocessing module, so that the platform can efficiently collect and process future weather data, generate average environment data and temperature change rate, so that the platform can provide reliable weather impact analysis for each marketing cycle. Through a model building module, the platform uses deep neural network technology to train an accurate thermal resistance value prediction model based on the material composition, thickness and fabric density of clothing, so that the recommended clothing is more in line with current environmental conditions. Through an inventory analysis module combined with an object retrieval module, clothing in offline stores can be quickly retrieved, and the environmental matching index of each type of clothing is calculated according to real-time environmental data, so as to screen out the most suitable clothing for current weather conditions for marketing.

Description

Technical Field

[0001] The present invention relates to the technical field of store marketing strategies, and specifically to an AI content production platform for offline store marketing. Background Art

[0002] In modern retail, as competition in the offline clothing retail industry becomes increasingly fierce, merchants need more flexible and precise marketing strategies to attract consumers. The marketing strategies of offline stores need to respond quickly to changing environmental conditions, especially the impact of weather conditions on consumer shopping behavior. Reasonable adjustment of clothing marketing strategies to match environmental conditions can significantly improve sales performance. Therefore, it is of great significance to provide data-driven marketing strategies based on weather changes and clothing characteristics.

[0003] In the existing technology, most offline stores' marketing strategies rely on human experience and static sales data. These traditional methods usually lack real-time and precision, and it is difficult to dynamically adapt to weather changes. They also cannot make full use of the specific characteristics of clothing such as material composition, thickness, and fabric density to predict consumer demand for different types of clothing. In addition, some offline stores may rely on basic weather forecasts to adjust inventory and promotional activities, but these adjustments are usually not quantified and optimized through a systematic approach.

[0004] There are several shortcomings in the existing technology: First, the traditional experience-driven method cannot effectively integrate the relationship between environmental factors and clothing characteristics, resulting in low accuracy of marketing strategies. Second, the lack of quantitative analysis of clothing thermal resistance properties limits the ability of stores to accurately recommend clothing under different weather conditions.

[0005] The above information disclosed in the Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may include information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide an AI content production platform for offline store marketing to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] An AI content production platform for offline store marketing, including:

[0009] Environmental collection module, used to collect Days after continuous The weather data predicted for the day and the actual weather data for the day, the weather data including temperature, humidity, rainfall and wind speed data;

[0010] Preprocessing module, used to generate The average environmental data after the day, the average environmental data includes average temperature, average humidity, average rainfall and average wind speed data, and combined with Days after continuous The rate of change of the temperature generated by the predicted minimum temperature of the day and the current actual maximum temperature data;

[0011] A model building module is used to collect sample data of clothing and Clo values ​​of corresponding clothing, wherein the sample data includes material composition, thickness and fabric density data of clothing, and to build a thermal resistance value prediction model of clothing based on a deep neural network, and to train the thermal resistance value prediction model using the sample data of clothing as a training set and the Clo values ​​of corresponding clothing as labels;

[0012] The inventory analysis module is used to input the material composition, thickness and fabric density data of each type of clothing in the offline store into the trained thermal resistance value prediction model, obtain the predicted Clo value of each type of clothing, and obtain the fabric density and fabric thickness data of the outermost fabric of each type of clothing, and form a set of clothing data groups with the predicted Clo value of each type of clothing, the fabric density and fabric thickness data of the outermost fabric;

[0013] Object retrieval module, used to The average environmental data and temperature change rate data after the day are used to retrieve the clothing data group, generate the environmental matching index of each clothing data group, and select the clothing that needs to be marketed according to the environmental matching index;

[0014] The content production module is used to obtain promotional pictures of corresponding clothing according to the selected clothing that needs to be marketed, and mark the parameters of the clothing on the promotional pictures.

[0015] Furthermore, collecting Days after continuous The geographical location of offline stores is obtained when the weather data is predicted for the day, and it is obtained through the database of the weather API platform. The API is called every day to obtain the actual weather data of the day. The weather data is obtained once a day at a fixed time.

[0016] Furthermore, generate The average environmental data for days is based on the formula: in, Respectively Days after continuous The average temperature, average humidity, average rainfall and average wind speed data of the day, Respectively Queen of Heaven The lowest temperature, highest humidity, maximum rainfall and maximum wind speed of the day;

[0017] The rate of change of the resulting temperature is based on the formula: in, represents the rate of change of temperature, Indicates the highest temperature in the actual weather data for the day.

[0018] Furthermore, the thermal resistance value prediction model includes an input layer, a hidden layer and an output layer. The input layer receives the material composition, thickness and fabric density data of the clothing in the sample data, and transmits it to the hidden layer in a fully connected manner. In the hidden layer, the data is linearly transformed and processed by an activation function, and then transmitted to the output layer. The output layer calculates and generates a predicted Clo value. The hidden layer adopts a ReLU activation function and uses the mean square error to calculate the error between the predicted value and the actual Clo value.

[0019] Furthermore, the specific method for generating the environment matching index of each clothing data group is as follows:

[0020] According to the rate of change of temperature and Days after continuous The average temperature and average humidity of the weather data of the day are analyzed and processed to obtain the recommended Clo value, and the recommended Clo value is compared with the predicted Clo value of each type of clothing to generate the comfort fit index of each type of clothing;

[0021] Based on the fabric density and thickness data of the garment, and Days after continuous The average rainfall and average wind speed data of the daily weather data are combined to generate the environmental matching index of various types of clothing;

[0022] A mathematical analysis model is established based on the comfort fit index and the environment fit index to obtain the environment matching index of each type of clothing.

[0023] Furthermore, the specific formula for generating the comfort fit index of clothing is: in, Indicates the recommended Clo value, Respectively Days after continuous The average temperature, average humidity, and temperature change rate of the weather data for each day. Respectively represent Comfort fit index and predicted Clo value of this type of clothing.

[0024] Furthermore, the formula for generating the environmental matching index of each type of clothing is:

[0025] in, Indicates Environmental compatibility index of this type of clothing, Respectively Days after continuous Average rainfall and average wind speed data for the day, Respectively represent Fabric density and fabric thickness data for this type of clothing.

[0026] Furthermore, the formula for obtaining the environmental matching index of each type of clothing is:

[0027] in, Indicates When selecting clothing for marketing based on the environmental matching index, select the clothing with the largest environmental matching index. Clothing for display.

[0028] Furthermore, when calibrating the parameters of clothing on the promotional picture, the predicted Clo value, material composition, thickness and fabric density data of the clothing are calibrated on the promotional line, and Days after continuous The average temperature, average humidity, average rainfall and average wind speed data of the day's weather data are uniformly named as the recommended wearing environment and are marked below the recommended wearing environment in the order of average temperature, average humidity, average rainfall and average wind speed.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] Through the environment acquisition module and the preprocessing module, the platform of the present invention can efficiently collect and process future weather data, generate average environment data and temperature change rate, so that the platform can provide reliable weather impact analysis for each marketing cycle. Through the model building module, the platform uses deep neural network technology to train an accurate thermal resistance value prediction model based on the material composition, thickness and fabric density of the clothing, overcoming the problem of insufficient analysis of thermal properties of clothing in traditional methods, making the recommended clothing more in line with current environmental conditions, thereby improving sales conversion rate;

[0031] The present invention can also quickly search for clothing in offline stores through the inventory analysis module combined with the object retrieval module, calculate the environmental matching index of each type of clothing based on real-time environmental data, and thus screen out the clothing that is most suitable for current weather conditions for marketing. The content production module automatically generates promotional pictures containing detailed clothing parameters based on the screening results, providing consumers with intuitive and personalized shopping information. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of the system structure of the production platform in the present invention. DETAILED DESCRIPTION

[0033] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0034] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0035] Example:

[0036] See also Figure 1 , the present invention provides a technical solution:

[0037] An AI content production platform for offline store marketing includes an environment acquisition module, a preprocessing module, a model building module, an inventory analysis module, an object retrieval module and a content production module, wherein:

[0038] Environmental collection module collection Days after continuous The weather data includes the temperature, humidity, rainfall and wind speed data.

[0039] In this embodiment, the collection Days after continuous The geographical location of offline stores is obtained when the weather data is predicted for the day, and it is obtained through the database of the weather API platform. The API is called every day to obtain the actual weather data of the day. The weather data is obtained once a day at a fixed time.

[0040] The environment collection module first obtains the geographic location information of offline stores, including longitude and latitude data, and obtains future weather forecast data through the weather API platform. The environment collection module calls the API and passes the geographic location parameters to obtain the weather forecast for a specific location. At a fixed time every day, the environment collection module will call the same weather API to obtain the actual weather data of the day. In order to ensure the accuracy and consistency of the data, the system will perform this operation at the same time every day. The weather APIs used in this embodiment are Amap Weather API and Xinzhi Weather API.

[0041] The preprocessing module generates The average environmental data after the day, the average environmental data includes average temperature, average humidity, average rainfall and average wind speed data, and combined with Days after continuous The predicted minimum temperature for the day and the current actual maximum temperature data generate the rate of change of temperature.

[0042] In this embodiment, the average environmental data generated by the preprocessing module includes average temperature, average humidity, average rainfall and average wind speed, and is combined with the temperature change rate calculation to calculate the future The average processing of the weather forecast data for the past 24 hours can smooth out the fluctuations caused by short-term weather changes, thereby providing more stable environmental data. This processing method reduces the impact of extreme weather on decision-making, can better adapt to complex weather changes, especially under changeable climate conditions, and provides stronger support for marketing.

[0043] In this embodiment, the generated The average environmental data for days is based on the formula:

[0044] in, Respectively Days after continuous The average temperature, average humidity, average rainfall and average wind speed data of the day, Respectively Queen of Heaven The lowest temperature, highest humidity, maximum rainfall and maximum wind speed of the day.

[0045] The choice of calculating the average of minimum temperature, maximum humidity, maximum rainfall and maximum wind speed is mainly to ensure the suitability of clothing in different weather conditions, especially in extreme weather, to provide consumers with adequate comfort and protection. The average of the minimum temperature helps to predict the coldest weather conditions, thereby guiding the selection of appropriate clothing thickness to ensure that consumers can keep warm in cold environments. Similarly, considering the maximum humidity can help determine the breathability requirements of clothing in humid weather conditions, so as to select appropriate materials to improve wearing comfort.

[0046] Average data for maximum rainfall and maximum wind speed can help plan protective clothing for extreme weather. By knowing the maximum rainfall, you can consider the need for waterproofing, while the prediction of maximum wind speed can help choose a more windproof design and provide an extra layer of warmth to prevent the cold from settling in on windy days.

[0047] The rate of change of the resulting temperature is based on the formula: in, represents the rate of change of temperature, Indicates the highest temperature in the actual weather data for the day.

[0048] In the process of generating the temperature change rate, the change between the current maximum temperature and the predicted minimum temperature can be quantified. It can provide a clear indicator of the temperature difference between the current weather and the future weather, which is of great significance for seasonal product adjustments in the apparel industry. Especially when dealing with extreme weather changes, the temperature change rate calculation is adopted. This is to obtain the minimum temperature during the forecast period, which can provide a most conservative estimate of the future The lowest temperature scenario that may be encountered during the day.

[0049] The model building module collects sample data of clothing and the Clo value of the corresponding clothing, wherein the sample data includes the material composition, thickness and fabric density data of the clothing, and constructs a thermal resistance value prediction model for clothing based on a deep neural network. The sample data of clothing is used as a training set, and the Clo value of the corresponding clothing is used as a label to train the thermal resistance value prediction model.

[0050] When building a clothing thermal resistance prediction model based on a deep neural network, this method has significant advantages by using features such as material composition, thickness, and fabric density in sample data, as well as the corresponding Clo values ​​as training data and labels. First, this deep learning-based model can capture complex nonlinear relationships that traditional statistical methods may not be able to identify, making the prediction model more accurate and reliable. By automatically learning feature patterns from a large amount of sample data, the model can adapt to a variety of different types and styles of clothing, thereby improving the accuracy of Clo value prediction.

[0051] In this embodiment, the thermal resistance value prediction model includes an input layer, a hidden layer and an output layer. The input layer receives the material composition, thickness and fabric density data of the clothing in the sample data, and transmits it to the hidden layer in a fully connected manner. In the hidden layer, the data is linearly transformed and processed by an activation function, and then transmitted to the output layer. The output layer calculates and generates a predicted Clo value. The hidden layer adopts a ReLU activation function and uses the mean square error to calculate the error between the predicted value and the actual Clo value.

[0052] In this embodiment, the structural design of the thermal resistance value prediction model follows the traditional deep neural network architecture and includes an input layer, a hidden layer, and an output layer. First, the input layer is specifically used to receive and process different features from sample data, including information such as the material composition, thickness, and fabric density of the clothing. These feature data are passed to the hidden layer in a fully connected manner to ensure that each input feature can interact with each neuron in the hidden layer. This connection method can maximize the use of all the information of the input data and provide a basis for subsequent data processing.

[0053] In the hidden layer, the input data is first linearly transformed, which usually includes applying weights to the input features and adding bias terms. Next, the linearly transformed data is processed by the ReLU activation function. ReLU is a nonlinear activation function whose output is the positive part of the input value. It can effectively solve the gradient vanishing problem and improve the training speed and convergence efficiency of the model. The role of the hidden layer is to extract and combine high-level representations of the input features, so that the model can learn complex patterns and nonlinear relationships related to Clo values.

[0054] Finally, the data processed by the hidden layer is passed to the output layer, which is responsible for calculating and generating the final Clo value prediction. In order to measure the difference between the model prediction value and the actual Clo value, the mean square error is used as the loss function. The mean square error provides a measure of the accuracy of the model prediction by calculating the squared difference between the predicted value and the true value and taking the average. By continuously adjusting the model parameters to minimize the loss function, the model can gradually improve the prediction accuracy and finally generate an accurate Clo value prediction. This structured model design and training process can effectively capture the thermal resistance characteristics of clothing and provide valuable prediction support for clothing design and production.

[0055] The inventory analysis module inputs the material composition, thickness and fabric density data of each type of clothing in the offline store into the trained thermal resistance prediction model, obtains the predicted Clo value of each type of clothing, and obtains the fabric density and fabric thickness data of the outermost fabric of each type of clothing, and forms a clothing data group with the predicted Clo value of each type of clothing, the fabric density and fabric thickness data of the outermost fabric.

[0056] In the inventory analysis module, by inputting the material composition, thickness, fabric density and other data of each type of clothing in offline stores into the pre-trained thermal resistance value prediction model, the corresponding predicted Clo value can be obtained. Subsequently, the system will collect the fabric density and fabric thickness data of the outermost fabric of each type of clothing, and integrate this information with the predicted Clo value to form a complete clothing data set. This data set not only includes the thermal resistance performance of the clothing, but also covers the key physical properties that affect its comfort and functionality, providing comprehensive data support for further analysis and decision-making.

[0057] In this embodiment, each type of clothing in the offline store specifically refers to each style and each type of clothing. For each specific clothing style, there will be corresponding material composition, thickness and fabric density data, thereby providing consumers with detailed product information. By providing these detailed attribute data for each clothing style, offline stores can help consumers better understand the characteristics of each clothing style, so as to make more appropriate purchase decisions. The fabric density and fabric thickness data of the outermost fabric of each type of clothing is measured based on the fabric with the largest proportion of the outermost fabric of the product.

[0058] Using the predictive power of deep learning models makes inventory analysis more accurate and efficient. Through automated model prediction, human errors can be reduced and the accuracy of the evaluation of the functionality of various types of clothing can be improved. At the same time, this method can process and analyze a large amount of data in real time, so as to quickly adapt to market changes and consumer needs. In this embodiment, the material composition, thickness, fabric density and other data of each type of clothing are first input into the trained thermal resistance value prediction model to obtain the corresponding Clo value. Next, the density and fabric thickness data of the outermost fabric of each type of clothing are collected. Finally, this information is integrated into a complete data set, which contains the predicted Clo value, the fabric density of the outermost fabric, and the fabric thickness, thereby forming a comprehensive data set that can fully reflect the thermal performance and physical properties of clothing.

[0059] Object retrieval module based on The average environmental data and temperature change rate data after the day are used to search for clothing data groups, generate environmental matching indexes for each clothing data group, and screen clothing that needs to be marketed based on the environmental matching indexes.

[0060] In this embodiment, the specific method of generating the environment matching index of each clothing data group is:

[0061] According to the rate of change of temperature and Days after continuous The average temperature and average humidity of the weather data of the day are analyzed and processed to obtain the recommended Clo value, and the recommended Clo value is compared with the predicted Clo value of each type of clothing to generate the comfort fit index of each type of clothing;

[0062] Based on the fabric density and thickness data of the garment, and Days after continuous The average rainfall and average wind speed data of the daily weather data are combined to generate the environmental matching index of various types of clothing;

[0063] A mathematical analysis model is established based on the comfort fit index and the environment fit index to obtain the environment matching index of each type of clothing.

[0064] This embodiment first obtains and analyzes the future Days after continuous The weather data of the day is used to calculate the average temperature and humidity of the period, and combined with the temperature change rate, a recommended Clo value is obtained. Then, this recommended Clo value is compared with the predicted Clo value of each type of clothing to generate the comfort fit index of each type of clothing, reflecting the thermal comfort of clothing under specific climatic conditions. Next, the fabric density and fabric thickness data of the clothing are combined with the future average rainfall and wind speed data to generate the environmental fit index of each type of clothing to evaluate the adaptability of clothing to specific environmental factors (such as rainfall and wind speed). Finally, by establishing a mathematical analysis model, the comfort fit index and the environmental fit index are combined to calculate the environmental fit index of each type of clothing as a comprehensive adaptability index of clothing under predicted climatic conditions.

[0065] The advantage of this embodiment is that it closely combines climate prediction with the physical properties of clothing, providing a scientific data-driven fit evaluation system for each type of clothing. It not only takes into account thermal comfort, but also comprehensively considers the interaction between the physical properties of fabrics and environmental factors, providing a more comprehensive and accurate clothing performance evaluation. Accurate calculation of the environmental matching index can help offline stores position products more effectively in marketing, improve consumer satisfaction and brand loyalty.

[0066] Furthermore, the specific formula for generating the comfort fit index of clothing is: in, Indicates the recommended Clo value, Respectively Days after continuous The average temperature, average humidity, and temperature change rate of the weather data for each day. Respectively represent Comfort fit index and predicted Clo value of this type of clothing.

[0067] In this embodiment, the calculation of the recommended Clo value combines multiple factors of future climate conditions. First, the sigmoid function is used. The average temperature is mapped. This nonlinear function can smoothly convert the temperature value into a value between 0 and 1, reflecting the nonlinear changes under extreme temperature conditions. Then, multiply it by To take into account the effect of humidity, the higher the humidity, the lower the thermal resistance effect generally felt by the human body; therefore, humidity reduces the effect of the Clo value in a linear manner. Finally, the hyperbolic tangent function is used To reflect the impact of temperature change rate, the function provides a smooth transition at a larger change rate, thereby reflecting the impact of sudden temperature changes on the thermal adaptability of clothing.

[0068] The calculation of is based on the difference between the recommended Clo value and the predicted Clo value for each type of garment, The calculation of is used to measure the relative error between the predicted Clo value and the recommended Clo value, and is converted into a measure of the degree of comfort fit. The closer it is to 1, the more consistent the predicted Clo value is with the recommended Clo value, that is, the better the comfort of the clothing under a specific predicted climate. This embodiment comprehensively considers temperature, humidity, and the rate of temperature change, reflecting the adaptability of clothing under different environmental conditions. Through such a construction, the formula is not just linear weighting, but a reasonable modeling of these factors through nonlinear functions, reflecting the complex impact of the actual environment on human thermal induction. The use of nonlinear functions such as sigmoid and tanh can handle irregular impact patterns in climate change. Especially under extreme conditions, these functions can better simulate the human body's response to temperature changes, reflecting the diversity of clothing adaptability.

[0069] The recommended Clo value is an indicator that reflects the thermal resistance of clothing required by the human body to achieve a comfortable state under specific environmental conditions. The larger the Clo value, the heavier or warmer the human body needs to maintain thermal comfort in this environment. The recommended Clo value reflects the thermal insulation performance of clothing required to maintain thermal comfort of the human body under predicted climatic conditions (including temperature, humidity and temperature change rate). It is a quantification of people's demand for warmth in a specific climate. The larger the Clo value, the lower the ambient temperature or the greater the temperature change. The human body needs more heat protection, which means that thicker clothing or multiple layers of clothing are needed to compensate for the heat exchange with the environment to maintain a stable body temperature.

[0070] Generally, the lower the temperature, the higher the recommended Clo value will be, because the human body needs more insulation to resist the cold. Higher humidity will reduce the required Clo value, because increased humidity will make people feel stuffy and uncomfortable, reducing the need for additional heat. The greater the rate of temperature change, the higher the recommended Clo value will be, because drastic temperature fluctuations usually require clothing to provide more flexible thermal adaptation capabilities.

[0071] Furthermore, the formula for generating the environmental matching index of each type of clothing is: in, Indicates Environmental compatibility index of this type of clothing, Respectively Days after continuous Average rainfall and average wind speed data for the day, Respectively represent Fabric density and fabric thickness data for this type of clothing.

[0072] The generated environmental fit index is used to evaluate the adaptability of different garments under specific climatic conditions (mainly rainfall and wind speed). The index provides an indicator to measure the comprehensive adaptability of clothing to the external environment by combining the physical properties of clothing (fabric density and fabric thickness) and environmental factors (rainfall and wind speed). The calculation of the environmental fit index combines the fabric density and fabric thickness of the clothing. These variables reflect the physical properties of clothing, such as windproofness and waterproofness. High fabric density and thickness generally mean better protection and can effectively cope with higher rainfall and wind speed. Environmental factors affect the environmental fit index in an exponential function, indicating that clothing needs to have higher adaptability under extreme weather conditions (such as high rainfall and strong wind).

[0073] Fabric density and fabric thickness are key physical properties of clothing. These parameters directly affect the wind and water resistance of the clothing. Higher density and thickness mean that the clothing can provide better protection. Average rainfall and average wind speed are two important environmental factors that determine the severity of weather conditions and thus affect the level of protection that clothing needs to provide.

[0074] It reflects the multiplicative relationship between the physical properties of clothing and environmental factors. The use of the exponential function shows that as rainfall and wind speed increase, the requirements for clothing increase exponentially. When the average rainfall and average wind speed increase, the fabric density and fabric thickness will significantly increase their contribution to the index, indicating that thick clothing is more popular in bad weather.

[0075] It reflects a correction. When the fabric density and thickness reach a certain level, the corresponding protective effectiveness tends to saturation. This part of the design may be to avoid the actual effect of unlimitedly improving the clothing parameters, so as to more truly reflect the diminishing marginal utility of physical properties. When the rainfall and wind speed are small, this part has a greater impact on the generated environmental matching index, reflecting that thinner clothing is more suitable in mild weather.

[0076] The larger the environmental matching index is, the more adaptable this type of clothing is under the described climate conditions. The formula attempts to reasonably express the relative adaptability and protection ability of each type of clothing under different climate conditions through the comprehensive relationship of these variables. In an environment with high average rainfall and average wind speed, the higher the fabric density and fabric thickness are, the higher the environmental matching index is. In an environment with low average rainfall and average wind speed, the higher the fabric density and fabric thickness are, the lower the environmental matching index is. This embodiment comprehensively considers the impact of different weather conditions on clothing selection by combining the characteristics of the exponential and inverse proportional functions, so that the preference for density and thickness can be dynamically adjusted in different environments.

[0077] Furthermore, the formula for obtaining the environmental matching index of each type of clothing is:

[0078] in, Indicates When selecting clothing for marketing based on the environmental matching index, select the clothing with the largest environmental matching index. Clothing for display.

[0079] In this embodiment, the exponential function can significantly amplify the impact of the recommended Clo value. This is because the exponential function grows very quickly and is very suitable for emphasizing the comfort advantages of certain garments. Subtracting 1 from the exponential function makes the overall impact 0 when the recommended Clo value is reached, thereby not affecting the environmental fit index when there is no particular comfort fit advantage.

[0080] The logarithmic function grows slowly for numbers greater than 1, which prevents the results from being too extreme when processing the environmental match index. The logarithmic function can better reflect relative changes when the value is large, and is suitable for emphasizing clothing that has obvious differences in environmental adaptability. Through the combination of index and logarithm, comfort and environmental adaptability information can be integrated in different dimensions. The index part mainly adjusts the comfort characteristics and emphasizes those clothing that have advantages in comfort; the logarithmic part provides smooth processing of the environmental match index, focusing on the basic performance of clothing in environmental adaptability. The final environmental match index provides a comprehensive priority indicator to help offline determine which clothing has the most potential in the upcoming climate conditions and market scenarios.

[0081] The Environmental Match Index reflects the comprehensive adaptability and comfort performance of each type of clothing under future weather conditions. It takes into account the protection and comfort of clothing under different climatic conditions, allowing offline stores to give priority to marketing clothing that will perform well in the upcoming weather.

[0082] The larger the environmental matching index, the more popular this type of clothing may be in the future market and more suitable for current or upcoming weather conditions. Therefore, a higher environmental matching index usually means that this type of clothing should be displayed and promoted more prominently. Offline stores can better combine weather forecasts with consumer preferences to provide customers with the most suitable clothing options and improve sales conversion rates.

[0083] The content production module obtains promotional pictures of the corresponding clothing based on the selected clothing that needs to be marketed, and marks the parameters of the clothing on the promotional pictures.

[0084] In this embodiment, when the parameters of clothing are marked on the promotional picture, the predicted Clo value, material composition, thickness and fabric density data of the clothing are marked on the promotional picture line, and Days after continuous The average temperature, average humidity, average rainfall and average wind speed data of the day's weather data are uniformly named as the recommended wearing environment and are marked below the recommended wearing environment in the order of average temperature, average humidity, average rainfall and average wind speed.

[0085] In the content production module, the step of marking the relevant parameters of the selected clothing to be marketed on the promotional pictures is intended to provide consumers with direct and clear product information to help them make more informed purchasing decisions. This step combines key clothing parameters with environmental data to enable consumers to not only understand the characteristics of the clothing itself, but also understand how these characteristics work under expected weather conditions.

[0086] First, the predicted Clo value, material composition, thickness and fabric density data of the calibrated garment are displayed on the promotional image to help consumers intuitively understand the basic performance and comfort of the garment. As an indicator of the thermal insulation performance of clothing, the Clo value directly affects consumers' understanding of the clothing's ability to keep warm at different temperatures. Material composition, thickness and fabric density provide in-depth information about the durability, breathability and overall quality of the garment.

[0087] Secondly, the weather data for the next few days is summarized as the recommended wearing environment, and the average temperature, humidity, rainfall and wind speed are marked in a specific order below the recommended wearing environment. This way of organizing information allows consumers to see at a glance whether the clothing is suitable for wearing in the expected environment. By connecting weather data with clothing performance, consumers can better evaluate the actual applicability of clothing in different climate conditions.

[0088] This information calibration strategy not only improves the usefulness of the promotional content, but also enhances consumers' purchasing confidence because they can more accurately match their needs with product features. Ultimately, this intuitive and informative display method helps to improve customer satisfaction and may promote sales conversion rates.

[0089] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0090] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0091] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0092] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. An AI content production platform for offline store marketing, characterized in that: include: Environmental collection module, used to collect Days after continuous The weather data predicted for the day and the actual weather data for the day, the weather data including temperature, humidity, rainfall and wind speed data; Preprocessing module, used to generate The average environmental data after the day, the average environmental data includes average temperature, average humidity, average rainfall and average wind speed data, and combined with Days after continuous The rate of change of the temperature generated by the predicted minimum temperature of the day and the current actual maximum temperature data; A model building module is used to collect sample data of clothing and Clo values ​​of corresponding clothing, wherein the sample data includes material composition, thickness and fabric density data of clothing, and to build a thermal resistance value prediction model of clothing based on a deep neural network, and to train the thermal resistance value prediction model using the sample data of clothing as a training set and the Clo values ​​of corresponding clothing as labels; The inventory analysis module is used to input the material composition, thickness and fabric density data of each type of clothing in the offline store into the trained thermal resistance value prediction model, obtain the predicted Clo value of each type of clothing, and obtain the fabric density and fabric thickness data of the outermost fabric of each type of clothing, and form a set of clothing data groups with the predicted Clo value of each type of clothing, the fabric density and fabric thickness data of the outermost fabric; Object retrieval module, used to The average environmental data and temperature change rate data after the day are used to retrieve the clothing data group, generate the environmental matching index of each clothing data group, and select the clothing that needs to be marketed according to the environmental matching index; The content production module is used to obtain promotional pictures of corresponding clothing according to the selected clothing that needs to be marketed, and mark the parameters of the clothing on the promotional pictures; generate The average environmental data for days is based on the formula: in, Respectively Days after continuous The average temperature, average humidity, average rainfall and average wind speed data of the day, Respectively Queen of Heaven The lowest temperature, highest humidity, maximum rainfall and maximum wind speed of the day; The rate of change of the resulting temperature is based on the formula: in, represents the rate of change of temperature, Indicates the highest temperature in the actual weather data of the day; The thermal resistance value prediction model includes an input layer, a hidden layer and an output layer. The input layer receives the material composition, thickness and fabric density data of the clothing in the sample data and transmits it to the hidden layer in a fully connected manner. In the hidden layer, the data is linearly transformed and processed by an activation function and transmitted to the output layer. The output layer calculates and generates a predicted Clo value. The hidden layer uses a ReLU activation function and uses a mean square error to calculate the error between the predicted value and the actual Clo value. The specific method for generating the environmental matching index of each clothing data group is as follows: According to the rate of change of temperature and Days after continuous The average temperature and average humidity of the weather data of the day are analyzed and processed to obtain the recommended Clo value, and the recommended Clo value is compared with the predicted Clo value of each type of clothing to generate the comfort fit index of each type of clothing; Based on the fabric density and thickness data of the garment, and Days after continuous The average rainfall and average wind speed data of the daily weather data are combined to generate the environmental matching index of various types of clothing; A mathematical analysis model is established based on the comfort matching index and the environmental matching index to obtain the environmental matching index of each type of clothing; The specific formula for generating the comfort fit index of clothing is: in, Indicates the recommended Clo value, Respectively Days after continuous The average temperature, average humidity, and temperature change rate of the weather data for each day. Respectively represent Comfort fit index and predicted Clo value of this type of clothing.

2. The AI ​​content production platform for offline store marketing according to claim 1, characterized in that: collection Days after continuous The geographical location of offline stores is obtained when the weather data is predicted for the day, and it is obtained through the database of the weather API platform. The API is called every day to obtain the actual weather data of the day. The weather data is obtained once a day at a fixed time.

3. The AI ​​content production platform for offline store marketing according to claim 1, characterized in that: The formula for generating the environmental fit index for each type of clothing is: in, Indicates Environmental compatibility index of this type of clothing, Respectively Days after continuous Average rainfall and average wind speed data for the day, Respectively represent Fabric density and fabric thickness data for this type of clothing.

4. The AI ​​content production platform for offline store marketing according to claim 3, characterized in that: The formula for obtaining the environmental matching index for each type of clothing is: in, Indicates When selecting clothing for marketing based on the environmental matching index, select the clothing with the largest environmental matching index. Clothing for display.

5. The AI ​​content production platform for offline store marketing according to claim 4, characterized in that: When calibrating the parameters of clothing on the promotional picture, the predicted Clo value, material composition, thickness and fabric density data of the clothing are calibrated on the promotional line and Days after continuous The average temperature, average humidity, average rainfall and average wind speed data of the day's weather data are uniformly named as the recommended wearing environment and are marked below the recommended wearing environment in the order of average temperature, average humidity, average rainfall and average wind speed.

Images