Personalized recommendation method and system for cultural and creative products

Abstract

The present invention relates to the field of data processing technology, and in particular to a personalized recommendation method and system for cultural and creative products, including: obtaining a prototype reference value of each best-selling product in each cultural and creative product category according to the similarity of the main view picture between each best-selling product and other best-selling products in its cultural and creative product category; screening the best-selling products in each cultural and creative product category according to the prototype reference value to obtain the prototype products in each cultural and creative product category; obtaining the adjusted main push priority of each non-best-selling product by analyzing the similarity of the main view picture between each non-best-selling product and the prototype product, as well as the accurate weight of the decision in the sales cycle; and making personalized recommendations for cultural and creative products based on the adjusted main push priority. The present invention can improve user experience and increase sales of various cultural and creative products for different user groups.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a personalized recommendation method and system for cultural and creative products. Background Art

[0002] Cultural and creative products refer to products that are based on traditional culture and are produced by combining creative design and modern technology, aiming to transform cultural value into commercial value. Such products cover a wide range, such as handicrafts, stationery, fashion accessories, household items and digital content, and usually have unique cultural connotations and artistic expressions, which can meet consumers' pursuit of personalization and quality. With the diversification of market demand, personalized group recommendations have gradually become an important strategy for the marketing of cultural and creative products. Through data analysis and consumer behavior research, companies can identify different consumer groups, such as young people, cultural enthusiasts, tourists and high-end markets, and make personalized recommendations based on the characteristics of these groups. The existing personalized recommendations for cultural and creative products are mainly based on the sales of different products for different groups, and there is no in-depth exploration of the category characteristics of personalized recommended products, which may result in the possibility of a single recommended product, and similar categories may be affected by products with good sales, resulting in them being difficult to sell, which in turn makes it possible for one cultural and creative product to dominate, resulting in poor recommendation effects for different groups, and hot-selling cultural and creative products being out of stock, while other cultural and creative products are unsalable. Summary of the invention

[0003] In order to solve the above problems, the present invention provides a personalized recommendation method and system for cultural and creative products.

[0004] An embodiment of the present invention provides a personalized recommendation method for cultural and creative products, the method comprising the following steps:

[0005] Acquire sales data of a number of cultural and creative works within a sales cycle; the sales data of the cultural and creative works includes sales volume, main promotion priority and main visual image of the cultural and creative works;

[0006] Classify all cultural and creative works into several categories of cultural and creative works; obtain the best-selling index of each cultural and creative work in each category of cultural and creative works according to the sales volume of each cultural and creative work; filter all cultural and creative works according to the best-selling index to obtain all best-selling products and non-best-selling products in each category of cultural and creative works;

[0007] The main push priority is adjusted by the best-selling index of the best-selling product to obtain the adjusted main push priority of each best-selling product; the accurate decision weight within the sales cycle is obtained according to the adjusted main push priority of the best-selling product; the prototype reference value of each best-selling product in each cultural and creative work category is obtained according to the similarity of the main view picture between each best-selling product and other best-selling products in its cultural and creative work category; the best-selling products in each cultural and creative work category are screened according to the prototype reference value to obtain the prototype product in each cultural and creative work category; the adjusted main push priority of each non-best-selling product is obtained by analyzing the similarity of the main view picture between each non-best-selling product and the prototype product, as well as the accurate decision weight within the sales cycle;

[0008] Provide personalized recommendations for cultural and creative products based on the adjusted promotion priorities of best-selling and non-best-selling products.

[0009] Preferably, the method of obtaining the best-selling index of each cultural and creative work in each category of cultural and creative works according to the sales volume of each cultural and creative work includes:

[0010] The first In the category of cultural and creative works The sales volume of the first cultural and creative work is The difference between the mean sales of all cultural and creative works in the cultural and creative work category is recorded as the first difference; In the category of cultural and creative works The difference between the sales volume of a cultural and creative work and the mean sales volume of all cultural and creative works in all cultural and creative work categories is recorded as the second difference; the normalized value of the maximum value between the first difference and the second difference is recorded as the first difference. In the category of cultural and creative works The best-selling index of a cultural and creative work.

[0011] Preferably, the method of screening all cultural and creative works according to the best-selling index to obtain all best-selling products and non-best-selling products in each category of cultural and creative works includes:

[0012] Preset a best-selling threshold parameter , In the category of cultural and creative works The best-selling index of the cultural and creative works is greater than the best-selling threshold parameter , will The literary and creative works as the best-selling products in the cultural and creative works category; In the category of cultural and creative works The best-selling index of the cultural and creative works is less than or equal to the best-selling threshold parameter , will The literary and creative works as the Non-best-selling products in the cultural and creative product category.

[0013] Preferably, the method of adjusting the main push priority by the best-selling index of the best-selling product to obtain the adjusted main push priority of each best-selling product includes:

[0014] For any best-selling product in any category of cultural and creative works, the ratio of the best-selling index of the best-selling product to the main push priority of the best-selling product is used as the adjusted main push priority of the best-selling product.

[0015] Preferably, the method of obtaining the accurate decision weight within the sales cycle according to the adjustment priority of the best-selling product includes the following specific methods:

[0016] Preset a weight parameter , the average and weight parameters of the adjusted priority of all best-selling products in all cultural and creative product categories The sum of the above is used as the accurate weight for decision making within the sales cycle.

[0017] Preferably, the method of obtaining the prototype reference value of each best-selling product in each category of cultural and creative works according to the similarity between the main view pictures of each best-selling product and other best-selling products in its category of cultural and creative works includes the following specific methods:

[0018] For The jth best-selling product in the category of cultural and creative works, the main image of the jth best-selling product and the Input the main view images of all other best-selling products in the category of cultural and creative works into the image structure similarity algorithm to obtain the relationship between the jth best-selling product and the jth best-selling product. The local structural similarity values ​​between the jth best-selling product and all other best-selling products in the category of cultural and creative works; the Otsu threshold algorithm is used to compare the jth best-selling product with the All other best-selling products in the category of cultural and creative works are classified into two categories based on the local structural similarity value to obtain the classification threshold. The best-selling products with a local structural similarity value greater than or equal to the classification threshold are regarded as the same category as the j-th best-selling product; the best-selling products with a local structural similarity value less than the classification threshold are regarded as the different category of the j-th best-selling product;

[0019] In the In each category of cultural and creative works, the mean of the local structural similarity values ​​between the j-th best-selling product and all its similar products is recorded as the first mean; the mean of the local structural similarity values ​​between the j-th best-selling product and all its different products is recorded as the second mean; the normalized value of the absolute value of the difference between the first mean and the second mean is used as the prototype reference value of the j-th best-selling product.

[0020] Preferably, the method of screening the best-selling products in each category of cultural and creative works according to the prototype reference value to obtain the prototype products in each category of cultural and creative works includes the following specific methods:

[0021] Preset a reference threshold parameter , if the prototype reference value of the jth best-selling product is greater than the reference threshold parameter , the jth best-selling product is recorded as A prototype product in the category of cultural and creative works.

[0022] Preferably, the adjusting main push priority of each non-best-selling product is obtained by analyzing the similarity of the main view pictures between each non-best-selling product and the prototype product, as well as the accurate weight of the decision in the sales cycle, including the specific method of:

[0023] For In the category of cultural and creative works A non-selling product, Among the categories of cultural and creative works, The similarity between the main view pictures of the non-best-selling products and the prototype products is obtained. Initial adjustment of the main promotion priority of non-best-selling products;

[0024] Preset a weighting parameter , will The inverse proportional value of the product of the priority of the main push of the non-best-selling product and the accurate weight of the decision in the sales cycle is recorded as the first inverse proportional value; the weighted parameter The sum of the first inverse proportional value is recorded as the adjusted weighted value; the adjusted weighted value and the first The product of the initial adjustment priority of the non-best-selling products is used as the first Adjust the promotion priority of non-best-selling products.

[0025] Preferably, the obtaining The initial adjustment of the main promotion priority of non-best-selling products includes the following specific methods:

[0026] The first The main image of the second non-best-selling product The main view images of all prototype products in the cultural and creative product category are input into the image structure similarity algorithm to obtain the The non-selling products and The local structural similarity values ​​between all prototype products in the cultural and creative work category; the prototype product corresponding to the maximum local structural similarity value is taken as the main promotion priority. Initial adjustment of the main promotion priority of non-best-selling products.

[0027] The present invention also proposes a personalized recommendation system for serving cultural and creative products, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any step of the personalized recommendation method for serving cultural and creative products when executing the computer program.

[0028] The beneficial effects of the technical solution of the present invention are as follows: the present invention obtains the prototype reference value of each best-selling product in each category of cultural and creative works according to the similarity of the main view pictures between each best-selling product and other best-selling products in its category of cultural and creative works; screens the best-selling products in each category of cultural and creative works according to the prototype reference value to obtain the prototype products in each category of cultural and creative works; obtains the adjustment of the main push priority of each non-best-selling product by analyzing the similarity of the main view pictures between each non-best-selling product and the prototype product, as well as the accurate weight of the decision in the sales cycle; and makes personalized recommendations for cultural and creative products based on the adjustment of the main push priority. The present invention can improve user experience and increase the sales of various cultural and creative products for different user groups. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0030] Figure 1 A flowchart of the steps of the personalized recommendation method for cultural and creative products provided by the present invention;

[0031] Figure 2 A feature relationship flow chart of the personalized recommendation of cultural and creative products serving the present invention. DETAILED DESCRIPTION

[0032] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the personalized recommendation method and system for serving cultural and creative products proposed by the present invention, its specific implementation method, structure, characteristics and effects are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0033] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0034] The specific scheme of the personalized recommendation method and system for cultural and creative products provided by the present invention is described in detail below with reference to the accompanying drawings.

[0035] See also Figure 1 , which shows a flowchart of a personalized recommendation method for cultural and creative products provided by an embodiment of the present invention, the method comprising the following steps:

[0036] Step S001: Acquire sales data of a number of cultural and creative works within a sales cycle; the sales data of the cultural and creative works include the sales volume, main promotion priority and main view picture of the cultural and creative works.

[0037] It should be noted that when making personalized recommendations for cultural and creative works, the current technology is mainly based on analyzing the sales data of various cultural and creative products. For different recommendation channels, the consumption tendencies shown by different consumer groups are different, so it is necessary to collect sales data of several cultural and creative works.

[0038] Specifically, we first need to collect sales data of several types of cultural and creative works within the sales cycle. The specific process is as follows:

[0039] Since the sales cycle of various cultural and creative products generally ranges from half a month to half a year, this embodiment collects the sales cycle in the last three months and collects the sales volume, number of advertisements, and main-view pictures of several cultural and creative products within the sales cycle;

[0040] It should be noted that since the publicity subject has publicity themes, such as traditional culture, Q version, cute version, etc., there are main products and accessories for different cultural and creative products. The difference between the two lies in the number of advertisements placed for each cultural and creative product during the sales cycle, which indicates the current publicity subject's main promotion priority for each cultural and creative product.

[0041] Specifically, for any cultural and creative work, the number of times the advertisements of the cultural and creative work are placed during the sales cycle is recorded as the advertising volume; the advertising volumes of all cultural and creative works during the sales cycle are linearly normalized, and the normalized value is used as the main promotion priority of each cultural and creative work;

[0042] The sales volume, main promotion priority and main visual image of each cultural and creative work are used as the sales data of each cultural and creative work.

[0043] So far, the sales data of several cultural and creative works within the sales cycle have been obtained through the above method.

[0044] Step S002: Classify all cultural and creative works into several categories of cultural and creative works; obtain the best-selling index of each cultural and creative work in each category of cultural and creative works according to the sales volume of each cultural and creative work; filter all cultural and creative works according to the best-selling index to obtain all best-selling products and non-best-selling products in each category of cultural and creative works.

[0045] Cultural and creative products are products that are the result of a close integration of the innovative expression of traditional culture embodied by the propaganda subject and modern life. Therefore, for a propaganda subject, all of its cultural and creative products have obvious thematic characteristics. Specifically, they are all designed to reflect its characteristic theme, and therefore have obvious homogeneity.

[0046] Preferably, in some implementations of the embodiments of the present invention, the main view image of each cultural and creative work in the sales cycle is input into a trained neural network classification model, and all cultural and creative works are classified into several categories of cultural and creative works; the classification model based on the CNN neural network used in this embodiment, the specific method of training the classification model is:

[0047] A large number of main view pictures of cultural and creative works are collected as a data set, and a classification model based on a CNN neural network is constructed. The CNN neural network can classify all samples in the data set into several categories based on the similarity of their main view pictures. The training loss function is a cross-entropy loss function. All the main view pictures of cultural and creative products are randomly classified in a ratio of 8:2, of which 8 are used as a training set, and 2 are used as a verification set through manual labeling to obtain a neural network classification model for cultural and creative products; the specific training process is a well-known content of neural networks, and the specific training process will not be repeated in this embodiment; each cultural and creative work category contains several cultural and creative works with the same theme characteristics.

[0048] It should be noted that, since different themes are popular in the market to different degrees, in each category of cultural and creative works, there may be one or more cultural and creative works whose sales volume is much higher than that of other cultural and creative works, or the entire category of cultural and creative works may be higher than other cultural and creative works, or there may be no best-sellers at all. In this case, when analyzing the recommendation priorities, it is necessary to consider cultural and creative products that are best-sellers during the sales cycle. The characteristics contained in these cultural and creative works are mainly divided into: either the sales volume is much higher than that of other cultural and creative works in the current category of cultural and creative works, or the sales volume is much higher than that of all cultural and creative works.

[0049] Preferably, in some implementations of the embodiments of the present invention, the specific method for obtaining the best-selling index of each cultural and creative work in each category of cultural and creative works according to the sales volume of each cultural and creative work is:

[0050] The first In the category of cultural and creative works The sales volume of the first cultural and creative work is The difference between the mean sales of all cultural and creative works in the cultural and creative work category is recorded as the first difference; In the category of cultural and creative works The difference between the sales volume of a cultural and creative work and the mean sales volume of all cultural and creative works in all cultural and creative work categories is recorded as the second difference; the normalized value of the maximum value between the first difference and the second difference is recorded as the first difference. In the category of cultural and creative works The best-selling index of individual cultural and creative works;

[0051] The specific formula is:

[0052]

[0053] In the formula, Indicates In the category of cultural and creative works The best-selling index of individual cultural and creative works; Indicates In the category of cultural and creative works The sales volume of individual cultural and creative works; Indicates The average sales volume of all cultural and creative works in a cultural and creative work category; It represents the mean sales volume of all cultural and creative works in all cultural and creative work categories; represents the maximum value function; represents the linear normalization function.

[0054] It should be noted that The value is greater than 0 and the larger it is, the The literary works are in Under the theme characteristics of each category of cultural and creative works, the market popularity is higher than that of other cultural and creative works; The value is greater than 0 and the larger it is, the The market popularity of a cultural and creative work is higher than that of all other cultural and creative works; To avoid the All cultural and creative works in the cultural and creative works category are popular, resulting in The literary works are in There is no significant difference in sales volume between this category of cultural and creative works and other cultural and creative works.

[0055] Preferably, in some implementations of the embodiments of the present invention, all cultural and creative works are screened according to the best-selling index, and the specific method for obtaining all best-selling products in each category of cultural and creative works is:

[0056] Preset a best-selling threshold parameter , wherein this embodiment is based on This example is described as an example, and this embodiment is not specifically limited. Depends on the specific implementation situation;

[0057] Jordi In the category of cultural and creative works The best-selling index of the cultural and creative works is greater than the best-selling threshold parameter , will The literary and creative works as the best-selling products in the cultural and creative works category; In the category of cultural and creative works The best-selling index of the cultural and creative works is less than or equal to the best-selling threshold parameter , will The literary and creative works as the Non-best-selling products in the cultural and creative product category.

[0058] So far, all the best-selling products and non-best-selling products in each category of cultural and creative works have been obtained through the above method.

[0059] Step S003: Adjust the main push priority by the best-selling index of the best-selling product to obtain the adjusted main push priority of each best-selling product; obtain the accurate decision-making weight within the sales cycle according to the adjusted main push priority of the best-selling product; obtain the prototype reference value of each best-selling product in each cultural and creative work category according to the similarity of the main view picture between each best-selling product and other best-selling products in its cultural and creative work category; screen the best-selling products in each cultural and creative work category according to the prototype reference value to obtain the prototype product in each cultural and creative work category; obtain the adjusted main push priority of each non-best-selling product by analyzing the similarity of the main view picture between each non-best-selling product and the prototype product, as well as the accurate decision-making weight within the sales cycle.

[0060] 1. Get the adjusted main promotion priority of each best-selling product.

[0061] It should be noted that since recommendations are made based on the different main promotion priorities of cultural and creative works during the sales cycle, what needs to be considered is that if the main promotion priority of a cultural and creative work is greater and the sales volume of the cultural and creative work is greater, it means that the recommendation is effective; on the contrary, if the cultural and creative work is not recognized by the market, then continuing to recommend and sell it based on the main promotion priority of the cultural and creative work will result in a waste of sales channels, and thus adversely affect the subsequent marketing of the publicity subject; for non-best-selling products other than best-selling products, they did not achieve the expected sales performance during the sales cycle, and therefore they are not used as anchor points for updating the main promotion priority; for best-selling products, their sales data in the previous sales cycle is better, then it is necessary to use the best-selling products as anchor points, and then select products from the non-best-selling products that have homogeneous information with the best-selling products for recommendation.

[0062] Preferably, in some implementations of the embodiments of the present invention, the main push priority is adjusted according to the best-selling index of the best-selling product, and the specific method for obtaining the adjusted main push priority of each best-selling product is:

[0063] For any best-selling product in any category of cultural and creative works, the ratio of the best-selling index of the best-selling product to the main push priority of the best-selling product is used as the adjusted main push priority of the best-selling product;

[0064] The specific formula is:

[0065]

[0066] In the formula, Indicates Adjust the priority of the jth best-selling product in a category of cultural and creative works; Indicates The best-selling index of the jth best-selling product in a category of cultural and creative works; Indicates The priority of promoting the jth best-selling product in a category of cultural and creative works; Represents the preset hyperparameters. This implementation presets , used to prevent the denominator from being 0.

[0067] It should be noted that the priority of cultural and creative works is between 0 and 1, and the larger the value, the stronger the recommendation; similarly, the best-selling index of best-selling products is between 0 and 1, and the larger the value, the more popular it is in the market; if a best-selling product can be recognized by the market and the recommendation of the publicity subject is reliable, then its best-selling index and the value of the priority of the main push are closer to the same; therefore, the larger the ratio of the best-selling index of a best-selling product to the priority of the main push, it means that the j-th best-selling product can still be recognized by the market under a smaller priority of the main push, and it needs to be recommended more, thereby increasing the sales of the publicity subject.

[0068] At this point, the adjusted main promotion priority of each best-selling product is obtained.

[0069] 2. Get the accurate weight of decisions within the sales cycle.

[0070] It should be noted that when analyzing whether the recommendation of a best-selling product is a correct decision, it is necessary to analyze the adjustment priority of the best-selling product in the sales cycle. That is, the higher the adjustment priority, the more recognized the product is by the market when selected by the promoter in the sales cycle. Conversely, it means that the product promoted in the sales cycle is not recognized by the market and needs to be replaced.

[0071] Preferably, in some implementations of the embodiments of the present invention, the specific method for obtaining the accurate decision weights within the sales cycle according to the adjustment priority of the best-selling products is:

[0072] Preset a weight parameter , wherein this embodiment is based on This example is described as an example, and this embodiment is not specifically limited. Depends on the specific implementation situation;

[0073] The average and weight parameters of the main promotion priority of all adjusted best-selling products in all categories of cultural and creative works The sum of is used as the accurate weight for decision making within the sales cycle;

[0074] The specific formula is:

[0075]

[0076] In the formula, Indicates the accurate weight of decisions within the sales cycle; Represents the preset weight parameter; Represents the mean of the adjusted main priority of all best-selling products in all cultural and creative product categories.

[0077] It should be noted that the larger the value of the decision accuracy weight is, the more it indicates that the product selected by the promotion subject in the sales cycle is correct; The value is between 0 and 1, so a weight parameter is added to it , so that the value of the decision-making accuracy weight within the sales cycle is distributed between 0.5-1.5; when the decision-making accuracy weight within the sales cycle is greater than 1, the decision is considered correct, otherwise, the decision is considered wrong and needs to be adjusted.

[0078] At this point, the accurate weights of decisions within the sales cycle are obtained.

[0079] 3. Obtain prototype products in each category of cultural and creative works.

[0080] It should be noted that the more similar a non-best-selling product is to a best-selling product, the more worthy it is to be recommended. Therefore, a similarity analysis needs to be performed first. For cultural and creative products in the same category of cultural and creative works, since they are divided based on a neural network classification model, their theme features are the same, which means that their main view images have certain similarities in shape. Therefore, it is necessary to first analyze the similarities between best-selling products in the same category of cultural and creative works.

[0081] Preferably, in some implementations of the embodiments of the present invention, according to the similarity between the main view pictures of each best-selling product and other best-selling products in its category of cultural and creative works, the specific method for obtaining the prototype reference value of each best-selling product in each category of cultural and creative works is:

[0082] For The jth best-selling product in the category of cultural and creative works, the main image of the jth best-selling product and the Input the main view images of all other best-selling products in the category of cultural and creative works into the image structure similarity algorithm to obtain the relationship between the jth best-selling product and the jth best-selling product. The local structural similarity values ​​between the jth best-selling product and all other best-selling products in the category of cultural and creative works; the Otsu threshold algorithm is used to compare the jth best-selling product with the All other best-selling products in the category of cultural and creative works are classified into two categories based on the local structural similarity value to obtain the classification threshold. The best-selling products with a local structural similarity value greater than or equal to the classification threshold are regarded as the same category as the j-th best-selling product; the best-selling products with a local structural similarity value less than the classification threshold are regarded as the different category of the j-th best-selling product;

[0083] Among them, the image structure similarity algorithm and the Otsu threshold algorithm are existing technologies, and will not be described in detail in this embodiment.

[0084] In the In the category of cultural and creative works, the mean of the local structural similarity values ​​between the jth best-selling product and all its peers is recorded as the first mean; the mean of the local structural similarity values ​​between the jth best-selling product and all its heterogeneous products is recorded as the second mean; the normalized value of the absolute value of the difference between the first mean and the second mean is used as the prototype reference value of the jth best-selling product;

[0085] Preset a reference threshold parameter , wherein this embodiment is based on This example is described as an example, and this embodiment is not specifically limited. Depends on the specific implementation situation;

[0086] If the prototype reference value of the jth best-selling product is greater than the reference threshold parameter , the jth best-selling product is recorded as Prototype products in the categories of cultural and creative works;

[0087] At this point, prototype products in each category of cultural and creative works are obtained.

[0088] 4. Get the adjusted main promotion priority of each non-best-selling product.

[0089] It should be noted that each non-best-selling product corresponds to a prototype product when calculating the initial adjustment of the main promotion priority. These non-best-selling products are regarded as subsidiary products of each prototype product. If the prototype product is not promoted during the sales cycle but is actually a best-selling product, it means that there are certain problems with the current decision. In this case, the subsidiary products under the prototype product need to increase the promotion efforts to make them more recognized by the market.

[0090] Preferably, in some implementations of the embodiments of the present invention, in each category of cultural and creative works, by analyzing the similarity of the main image between each non-best-selling product and the prototype product, as well as the accurate weight of the decision in the sales cycle, the specific method for obtaining the adjustment of the main push priority of each non-best-selling product is:

[0091] For In the category of cultural and creative works non-selling products, The main image of the second non-best-selling product The main view images of all prototype products in the cultural and creative product category are input into the image structure similarity algorithm to obtain the The non-selling products and The local structural similarity values ​​between all prototype products in the cultural and creative work category; the prototype product corresponding to the maximum local structural similarity value is taken as the main promotion priority. Initial adjustment of the main promotion priority of non-best-selling products;

[0092] Preset a weighting parameter , wherein this embodiment is based on This example is described as an example, and this embodiment is not specifically limited. Depends on the specific implementation situation;

[0093] The first The inverse proportional value of the product of the priority of the main push of the non-best-selling product and the accurate weight of the decision in the sales cycle is recorded as the first inverse proportional value; the weighted parameter The sum of the first inverse proportional value is recorded as the adjusted weighted value; the adjusted weighted value and the first The product of the initial adjustment priority of the non-best-selling products is used as the first Adjust the priority of non-best-selling products;

[0094] The specific formula is:

[0095]

[0096] In the formula, Indicates Adjust the priority of non-best-selling products; Indicates the accurate weight of decisions within the sales cycle; Indicates The priority of promoting non-best-selling products; Indicates Initial adjustment of the main promotion priority of non-best-selling products; represents an exponential function with a natural constant as the base, and the embodiment adopts Model to present inverse proportional relationship and normalization processing, As the input of the model, the implementer can choose the inverse proportional function and the normalization function according to the actual situation.

[0097] It should be noted that the value of the decision accuracy weight within the sales cycle represents the decision accuracy of all products in the sales cycle. The larger the value, the more correct the decision. In this case, there is no need to adjust the initial main push priority based on the main push priority within the sales cycle. However, when the value is smaller, it means that the market judgment based on the main push priority in the sales cycle is wrong, which leads to the actual best-selling products being cultural and creative products that are not promoted. Therefore, it is necessary to use the main push priority of the sales cycle for adjustment. When adjusting, the smaller the main push priority, the more it means that the product still maintains high sales without publicity resources. In this case, it is necessary to vigorously promote it in the sales cycle to increase sales.

[0098] At this point, the adjusted main push priority of each non-best-selling product is obtained through the above method.

[0099] Step S004: Personalize recommendations for cultural and creative products based on the adjusted priority of best-selling products and non-best-selling products.

[0100] It should be noted that the adjustment of the main promotion priority is obtained by analyzing the sales volume of a promotional entity in the same sales channel, while the consumer group of the same sales channel remains fixed. For example, the online sales channel is more inclined to young people, while the offline sales channel is mainly sold through surrounding stores near the promotional entity (attractions), and is aimed at everyone.

[0101] Specifically, in the next sales cycle, the adjustment priority of all cultural and creative works is arranged in descending order, and corresponding promotion costs are assigned in the descending order, thereby realizing personalized recommendations for cultural and creative products. That is, personalized recommendations are made for cultural and creative products in descending order of adjustment priority, and the cultural and creative product with the highest adjustment priority is recommended first.

[0102] Through the above steps, you can complete the personalized recommendation of cultural and creative products. Figure 2 , which shows a feature relationship flow chart serving personalized recommendation of cultural and creative products.

[0103] The present invention also proposes a personalized recommendation system for cultural and creative products, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the system implements the steps of the personalized recommendation method for cultural and creative products in steps S001 to S004.

[0104] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A personalized recommendation method for cultural and creative products, characterized in that: The method comprises the following steps: Acquire sales data of a number of cultural and creative works within a sales cycle; the sales data of the cultural and creative works includes sales volume, main promotion priority and main visual image of the cultural and creative works; Classify all cultural and creative works into several categories of cultural and creative works; obtain the best-selling index of each cultural and creative work in each category of cultural and creative works according to the sales volume of each cultural and creative work; filter all cultural and creative works according to the best-selling index to obtain all best-selling products and non-best-selling products in each category of cultural and creative works; The main push priority is adjusted by the best-selling index of the best-selling product to obtain the adjusted main push priority of each best-selling product; the accurate decision weight within the sales cycle is obtained according to the adjusted main push priority of the best-selling product; the prototype reference value of each best-selling product in each cultural and creative work category is obtained according to the similarity of the main view picture between each best-selling product and other best-selling products in its cultural and creative work category; the best-selling products in each cultural and creative work category are screened according to the prototype reference value to obtain the prototype product in each cultural and creative work category; the adjusted main push priority of each non-best-selling product is obtained by analyzing the similarity of the main view picture between each non-best-selling product and the prototype product, as well as the accurate decision weight within the sales cycle; Provide personalized recommendations for cultural and creative products based on the adjusted promotion priorities of best-selling and non-best-selling products.

2. The personalized recommendation method for cultural and creative products according to claim 1, characterized in that: The method of obtaining the best-selling index of each cultural and creative work in each category of cultural and creative works according to the sales volume of each cultural and creative work includes: The first In the category of cultural and creative works The sales volume of the first cultural and creative work is The difference between the mean sales of all cultural and creative works in the cultural and creative work category is recorded as the first difference; In the category of cultural and creative works The difference between the sales volume of a cultural and creative work and the mean sales volume of all cultural and creative works in all cultural and creative work categories is recorded as the second difference; the normalized value of the maximum value between the first difference and the second difference is recorded as the first difference. In the category of cultural and creative works The best-selling index of a cultural and creative work.

3. The personalized recommendation method for cultural and creative products according to claim 1, characterized in that: The specific method of screening all cultural and creative works according to the best-selling index to obtain all best-selling products and non-best-selling products in each category of cultural and creative works is as follows: Preset a best-selling threshold parameter , In the category of cultural and creative works The best-selling index of the cultural and creative works is greater than the best-selling threshold parameter , will The literary and creative works as the best-selling products in the cultural and creative works category; In the category of cultural and creative works The best-selling index of the cultural and creative works is less than or equal to the best-selling threshold parameter , will The literary and creative works as the Non-best-selling products in the cultural and creative product category.

4. The personalized recommendation method for cultural and creative products according to claim 1, characterized in that: The method of adjusting the main push priority by the best-selling index of the best-selling product to obtain the adjusted main push priority of each best-selling product includes: For any best-selling product in any category of cultural and creative works, the ratio of the best-selling index of the best-selling product to the main push priority of the best-selling product is used as the adjusted main push priority of the best-selling product.

5. The personalized recommendation method for cultural and creative products according to claim 1, characterized in that: The specific method of obtaining the accurate decision weight within the sales cycle according to the adjustment priority of the best-selling products is as follows: Preset a weight parameter , the mean and weight parameters of the priority of all best-selling products in all categories of cultural and creative works The sum of the above is used as the accurate weight for decision making within the sales cycle.

6. The personalized recommendation method for cultural and creative products according to claim 1, characterized in that: The method of obtaining the prototype reference value of each best-selling product in each category of cultural and creative works according to the similarity between the main view pictures of each best-selling product and other best-selling products in the category of cultural and creative works includes: For The jth best-selling product in the category of cultural and creative works, the main image of the jth best-selling product and the The main view images of all other best-selling products in the category of cultural and creative works are input into the image structure similarity algorithm to obtain the relationship between the jth best-selling product and the jth best-selling product. The local structural similarity values ​​between the jth best-selling product and all other best-selling products in the category of cultural and creative works; the Otsu threshold algorithm is used to compare the jth best-selling product with the All other best-selling products in the category of cultural and creative works are classified into two categories based on the local structural similarity value to obtain the classification threshold. The best-selling products with a local structural similarity value greater than or equal to the classification threshold are regarded as the same category as the j-th best-selling product; the best-selling products with a local structural similarity value less than the classification threshold are regarded as the different category of the j-th best-selling product; In the In each category of cultural and creative works, the mean of the local structural similarity values ​​between the j-th best-selling product and all its similar products is recorded as the first mean; the mean of the local structural similarity values ​​between the j-th best-selling product and all its different products is recorded as the second mean; the normalized value of the absolute value of the difference between the first mean and the second mean is used as the prototype reference value of the j-th best-selling product.

7. The personalized recommendation method for cultural and creative products according to claim 1, characterized in that: The method of screening the best-selling products in each category of cultural and creative works according to the prototype reference value to obtain the prototype products in each category of cultural and creative works includes the following specific methods: Preset a reference threshold parameter , if the prototype reference value of the jth best-selling product is greater than the reference threshold parameter , the jth best-selling product is recorded as A prototype product in the category of cultural and creative works.

8. The personalized recommendation method for cultural and creative products according to claim 1, characterized in that: The adjustment priority of the main push of each non-best-selling product is obtained by analyzing the similarity of the main view pictures between each non-best-selling product and the prototype product, as well as the accurate weight of the decision in the sales cycle, including the specific method of: For In the category of cultural and creative works A non-selling product, Among the categories of cultural and creative works, The similarity between the main view pictures of the non-best-selling products and the prototype products is obtained. Initial adjustment of the main promotion priority of non-best-selling products; Preset a weighting parameter , will The inverse proportional value of the product of the priority of the main push of the non-best-selling product and the accurate weight of the decision in the sales cycle is recorded as the first inverse proportional value; the weighted parameter The sum of the first inverse proportional value is recorded as the adjusted weighted value; the adjusted weighted value and the first The product of the initial adjustment priority of the non-best-selling products is used as the first Adjust the promotion priority of non-best-selling products.

9. The personalized recommendation method for cultural and creative products according to claim 8, characterized in that: The acquisition The initial adjustment of the main promotion priority of non-best-selling products includes the following specific methods: The first The main image of the second non-best-selling product The main view images of all prototype products in the cultural and creative product category are input into the image structure similarity algorithm to obtain the The non-selling products and The local structural similarity values ​​between all prototype products in the cultural and creative work category; the prototype product corresponding to the maximum local structural similarity value is taken as the main promotion priority. Initial adjustment of the main promotion priority of non-best-selling products.

10. A personalized recommendation system for cultural and creative products, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the personalized recommendation method for cultural and creative products as described in any one of claims 1 to 9 are implemented.

Images