Dynamic packaging method and device for aviation tourism products, electronic equipment and medium

Abstract

The invention discloses a dynamic packaging method and device for aviation tourism products, electronic equipment and a medium, and the dynamic packaging method for the aviation tourism products comprises the steps: obtaining the product attribute information of a plurality of aviation tourism products; determining a plurality of target association rules according to the product attribute information; obtaining a target product attribute set based on the plurality of target association rules; and dynamically packaging the plurality of aviation tourism products based on the target product attribute set. The technical problem of low dynamic packaging recommendation efficiency and accuracy of multi-category aviation tourism products is solved.

Description

Technical Field

[0001] The present invention relates to the field of tourism information technology, and in particular to a dynamic packaging method, device, electronic equipment and medium for aviation tourism products. Background Art

[0002] The rapid development of Internet technology has led to the emergence of massive amounts of information and resources, which is both an opportunity and a challenge for the aviation and travel retail industry. When planning a trip, users often find it difficult to make the best choice when faced with a wide range of products such as flights, hotels, insurance, and catering. Especially when seeking personalized and comprehensive travel plans, the complexity of the choices has increased exponentially. In order to optimize the user experience and provide more accurate and personalized services, the recommendation system came into being and became an efficient information filtering and recommendation tool in the aviation and travel retail field.

[0003] However, although the existing recommendation system performs well in recommending single products, it has obvious shortcomings in dynamically packaging and recommending multi-category products. The dynamic packaging process not only requires identifying and mining the correlation between products, but also considers multiple dimensions such as product attributes, user preferences, and market dynamics to generate personalized product packages that meet user needs. In the field of aviation and tourism retail, multi-category dynamic packaging recommendations face two major challenges: one is the efficiency of data processing, and the other is the accuracy and personalization of packaging strategies.

[0004] In response to the above challenges, traditional product packaging methods often rely on predefined fixed bundle datasets. This method cannot generate new bundles when the dataset is missing or incomplete, thus limiting the flexibility and applicability of the recommendation system. Although some existing methods for generating new bundles attempt to automatically identify product associations through algorithms, they rely too much on domain knowledge and are inefficient when processing large-scale data, making it difficult to meet real-time requirements. In addition, existing methods often find it difficult to capture the associations at the product attribute level, resulting in a deviation between the generated packaging solutions and the actual needs of users and a lack of sufficient accuracy.

[0005] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0006] The embodiments of the present invention provide a method, device, electronic device and medium for dynamically packaging aviation travel products, so as to at least solve the technical problem of low efficiency and accuracy in dynamic packaging recommendation of multiple categories of aviation travel products.

[0007] According to one aspect of an embodiment of the present invention, a method for dynamically packaging aviation tourism products is provided, comprising: obtaining product attribute information of a plurality of aviation tourism products; determining a plurality of target association rules according to the product attribute information; obtaining a target product attribute set based on the plurality of target association rules; and dynamically packaging the plurality of aviation tourism products based on the target product attribute set.

[0008] Optionally, obtaining product attribute information of a plurality of aviation travel products includes: determining a plurality of product features of the plurality of aviation travel products; and obtaining a plurality of product attributes corresponding to the plurality of product features to obtain product attribute information.

[0009] Optionally, determining multiple target association rules based on product attribute information includes: constructing a frequent item set based on the product attribute information; constructing a frequent pattern tree and header table information based on the frequent item set; mining target frequent patterns based on the frequent pattern tree and header table information; and determining multiple target association rules based on the target frequent patterns.

[0010] Optionally, constructing a frequent item set based on product attribute information includes: performing a scanning operation on the product attribute information to obtain multiple product support values, wherein the product attribute information includes: multiple candidate product attributes, and there is a one-to-one correspondence between the multiple candidate product attributes and the multiple product support values; comparing a predetermined target support value with the multiple product support values ​​to determine a frequent item set, wherein the frequent item set includes multiple target product attributes, and multiple target product support values ​​corresponding to the multiple target product attributes are all greater than the target support value.

[0011] Optionally, constructing a frequent pattern tree and header table information based on a frequent item set includes: determining that a root node of the frequent pattern tree is an empty node; performing a scanning operation on the product attribute information to obtain a plurality of frequent pattern tree branches; constructing a frequent pattern tree based on the root node and the plurality of frequent pattern tree branches; and determining header table information based on the frequent pattern tree.

[0012] Optionally, mining the target frequent pattern based on the frequent pattern tree and the header table information includes: performing a traversal operation on multiple frequent pattern tree branches based on the header table information to obtain the target frequent pattern.

[0013] Optionally, determining a plurality of target association rules according to the target frequent pattern includes: determining a plurality of initial association rules according to the target frequent pattern; obtaining a plurality of confidence values ​​of the plurality of initial association rules; and determining a plurality of target association rules according to the plurality of confidence values.

[0014] According to another aspect of an embodiment of the present invention, a dynamic packaging device for aviation tourism products is also provided, including: a first acquisition module, used to obtain product attribute information of multiple aviation tourism products; a determination module, used to determine multiple target association rules according to the product attribute information; a second acquisition module, used to obtain a target product attribute set based on the multiple target association rules; and a packaging module, used to dynamically package multiple aviation tourism products based on the target product attribute set.

[0015] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute any of the above-mentioned methods for dynamically packaging aviation travel products.

[0016] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute any of the above-mentioned methods for dynamically packaging aviation travel products when running on a computer or a processor.

[0017] In an embodiment of the present invention, product attribute information of multiple aviation tourism products is obtained; multiple target association rules are determined according to the product attribute information; a target product attribute set is obtained based on the multiple target association rules; and multiple aviation tourism products are dynamically packaged based on the target product attribute set. The present invention achieves the purpose of efficiently and accurately generating packaging plans by combining product attribute information and mining potential associations between different products, thereby achieving the technical effect of effectively improving the efficiency and accuracy of dynamic packaging recommendations for multiple categories of aviation tourism products, thereby solving the technical problem of low efficiency and accuracy of dynamic packaging recommendations for multiple categories of aviation tourism products. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0019] Figure 1 is a flow chart of a method for dynamically packaging aviation travel products according to one embodiment of the present invention;

[0020] Figure 2 is a schematic diagram of the result of the FP-Tree generated in the first step according to one embodiment of the present invention;

[0021] Figure 3 is a schematic diagram of the result of the FP-Tree generated in the second step according to one embodiment of the present invention;

[0022] Figure 4is a schematic diagram of the result of the FP-Tree generated in the third step according to one embodiment of the present invention;

[0023] Figure 5 is a schematic diagram of a complete FP-Tree generated according to one embodiment of the present invention;

[0024] Figure 6 is a schematic diagram of the result of constructing header table information according to one embodiment of the present invention;

[0025] Figure 7 is a schematic diagram of the result of constructing a conditional FP-Tree according to one embodiment of the present invention;

[0026] Figure 8 It is a structural block diagram of a dynamic packaging device for aviation travel products according to one embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0028] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] According to an embodiment of the present invention, an embodiment of a dynamic packaging method for air travel products is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system containing at least one set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0030] The method embodiment may also be performed in an electronic device, a similar control device or an electronic device including a memory and a processor. Taking an electronic device as an example, the electronic device may include one or more processors and a memory for storing data. Optionally, the electronic device may also include a communication device and a display device for communication functions. It will be understood by those skilled in the art that the above structural description is only illustrative and does not limit the structure of the above electronic device. For example, the electronic device may also include more or fewer components than the above structural description, or have a configuration different from the above structural description.

[0031] The processor may include one or more processing units. For example, the processor may include a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processing (DSP) chip, a microcontroller unit (MCU), a field-programmable gate array (FPGA), a neural-network processing unit (NPU), a tensor processing unit (TPU), an artificial intelligence (AI) type processor, and other processing devices. Among them, different processing units may be independent components or integrated into one or more processors. In some instances, the electronic device may also include one or more processors.

[0032] The memory can be used to store computer programs, such as storing the computer programs corresponding to the dynamic packaging method of the aviation tourism products in the embodiment of the present invention, and the processor implements the dynamic packaging method of the aviation tourism products mentioned above by running the computer programs stored in the memory. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the electronic device via a grid. Examples of the above-mentioned grid include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0033] Figure 1 is a flow chart of a method for dynamically packaging aviation travel products according to one embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0034] Step S101, obtaining product attribute information of a plurality of aviation travel products.

[0035] The product attribute information of aviation travel products covers multiple aspects such as flights, hotels, insurance, and catering.

[0036] Specifically, flight attribute information includes but is not limited to the origin, destination, departure time, arrival time, airline, aircraft type, cabin class (such as economy class, business class, first class) and price, etc. These attributes are crucial to building a complete picture of user travel and can help understand the user's basic travel needs and preferences.

[0037] Specifically, hotel attribute information includes but is not limited to hotel grade (such as three-star, four-star, five-star), accommodation type (business, resort, fast chain), geographical location (city, region), price, facilities and services, etc. These attributes can reflect the user's considerations on accommodation comfort, convenience and budget.

[0038] Specifically, insurance attribute information includes but is not limited to insurance type (such as aviation accident insurance, health insurance, baggage insurance), coverage, insured amount, premium, applicable area, etc. These attributes help identify users' specific needs for security protection.

[0039] Specifically, the catering attribute information involves catering types (such as vegetarian, non-vegetarian), dish styles (Chinese, Western), catering services included in the flight, etc., which reflects the user's eating habits and preferences.

[0040] Step S102: determining a plurality of target association rules according to product attribute information.

[0041] Specifically, first, the attribute information of aviation tourism products is scanned, the frequently occurring attribute combinations are screened out, and a frequent item set is constructed; then, the association rule learning algorithm is used to perform data mining on the constructed frequent item set to quickly extract frequent patterns in the data; then, preliminary association rules are generated based on the extracted frequent patterns; finally, the generated preliminary association rules are further screened and optimized to obtain multiple target association rules.

[0042] Among them, the association rule learning algorithm is a machine learning method that discovers interesting associations or relationships between items in a data set, and is mainly used in the fields of user behavior analysis, recommendation systems, etc. The main goal of this type of algorithm is to discover the statistical laws of the co-occurrence of certain items (or events) in a data set. These laws are usually presented in the form of "if...then..." rules, such as "if a user buys item A, then they are likely to buy item B."

[0043] Exemplarily, the FP-growth (Frequent Pattern Growth) association rule learning algorithm can be used to mine frequent item sets. FP-growth reduces the number of database scans by constructing an FP tree (a compressed data structure), thereby greatly improving the speed of mining frequent item sets. Especially when processing large-scale data sets, its advantages are more obvious.

[0044] Step S103, obtaining a target product attribute set based on multiple target association rules.

[0045] Specifically, in the target association rule mining stage, a series of rules reflecting the correlation between the attributes of aviation tourism products were obtained, such as [Business Hotel => Domestic Comprehensive Accident Insurance]. Among them, the latter term (right side) of the rule represents the additional product attributes that users are likely to demand under certain conditions. Therefore, it is necessary to focus on the latter terms of these rules and analyze which product attributes users are most likely to be interested in in the context of selecting a certain type of product.

[0046] Specifically, by analyzing the latter terms of multiple target association rules, a target product attribute set is constructed. This set contains all frequently mentioned product attributes, which are regarded as key constraints in dynamic packaging services. For example, the latter terms of the rules [five-star luxury hotel=>Beijing-Tokyo air ticket] and [business hotel=>domestic comprehensive accident insurance] are "Beijing-Tokyo air ticket" and "domestic comprehensive accident insurance" respectively. These attributes will be added to the target product attribute set for subsequent product packaging screening.

[0047] Furthermore, in order to ensure the accuracy and practicality of the target product attribute set, the initial target product attribute set needs to be optimized. This includes further screening attributes using indicators such as lift, eliminating attributes that appear frequently but are not highly correlated with user needs, and merging attributes that are highly relevant to user travel scenarios. For example, the latter items of multiple rules all point to "Beijing-Tokyo air tickets", which shows that in different travel contexts, users have a high demand for air tickets from Beijing to Tokyo, so this attribute will be strengthened in the set.

[0048] The final set of target product attributes will be used as constraints for product packaging. In the subsequent recommendation system, these constraints will be used to filter and combine aviation travel products that meet user needs. For example, if "five-star luxury hotel" and "Beijing-Tokyo air ticket" appear in the latter items of multiple target association rules at the same time, the system will give priority to product combinations containing these attributes to generate a packaging solution that better meets user expectations.

[0049] Step S104: dynamically package multiple aviation travel products based on the target product attribute set.

[0050] Specifically, the target product attribute set obtained in step S103 is deeply analyzed. This set includes all the attributes of aviation travel products that are identified as being of interest to the user in a specific scenario, such as "five-star luxury hotel", "economy class air ticket", "overseas travel insurance", etc. Each attribute represents an aspect that the user may consider when planning a trip, such as accommodation type, air ticket class, and insurance options.

[0051] Specifically, based on the target product attribute set, the system quickly queries the product database and selects products that meet these attributes. For example, if the attribute set includes "five-star luxury hotel", the system will retrieve all hotel information classified as five-star from the database and further select hotels that match the travel destination. This process utilizes the accuracy and generalization of attribute information and can effectively process massive product data in the field of aviation tourism.

[0052] Furthermore, after selecting products that meet the target attribute set, the system will dynamically combine and package these products according to the user's current travel needs and preferences. This process is based on the user's input of departure, destination, travel time and other information, as well as previously mined association rules, such as "users who buy economy class tickets are also likely to choose overseas travel insurance." The system will automatically match and package these products to provide users with a one-stop, personalized travel solution.

[0053] Optionally, in order to ensure a high-quality experience of the packaged service, additional algorithms or strategies can be used to optimize the product packages. For example, by comparing indicators such as price, schedule, and user satisfaction of different combinations, the system can recommend the product package with the best cost-effectiveness or the one that best meets the user's preferences. In addition, the system will also take into account real-time inventory and promotion information to ensure that the recommended product package not only meets user needs but is also competitive in the market.

[0054] In an embodiment of the present invention, product attribute information of multiple aviation tourism products is obtained; multiple target association rules are determined according to the product attribute information; a target product attribute set is obtained based on the multiple target association rules; and multiple aviation tourism products are dynamically packaged based on the target product attribute set. The present invention achieves the purpose of efficiently and accurately generating packaging plans by combining product attribute information and mining potential associations between different products, thereby achieving the technical effect of effectively improving the efficiency and accuracy of dynamic packaging recommendations for multiple categories of aviation tourism products, thereby solving the technical problem of low efficiency and accuracy of dynamic packaging recommendations for multiple categories of aviation tourism products.

[0055] Optionally, in step S101, obtaining product attribute information of a plurality of aviation travel products includes: determining a plurality of product features of the plurality of aviation travel products; and obtaining a plurality of product attributes corresponding to the plurality of product features to obtain product attribute information.

[0056] Specifically, a detailed analysis is conducted on multiple air travel products to identify the important features of each product. These features include the cabin type of the flight, the star rating and type of the hotel, the coverage of insurance, and additional services such as meal selection and baggage allowance. For example, for flight products, the features include "economy class", "business class" or "first class"; for hotel products, the features involve categories such as "five-star luxury", "business type" or "fast chain".

[0057] Specifically, after identifying the product features, these features are mapped to specific product attributes. Product attributes are more detailed and standardized information used to describe product features. For example, "five-star luxury hotel" can be further subdivided into two attributes: "hotel star" and "hotel type", and "economy class air tickets" can be subdivided into attributes such as "air ticket class" and "airline". Through this attribute mapping, product information can be converted into a structured data format for easy algorithm processing.

[0058] Furthermore, for each aviation tourism product, the corresponding product attribute information is collected and organized. This process includes the establishment of an attribute database, which stores all possible attributes and their descriptions, such as the "hotel star" attribute includes "five stars", "four stars", "three stars" and other options; the "ticket class" attribute includes "economy class", "business class" and "first class" and other information. At the same time, it is also necessary to record the corresponding relationship between each product and its attributes for subsequent processing.

[0059] Optionally, in step S102, determining multiple target association rules according to product attribute information includes: constructing a frequent item set according to the product attribute information; constructing a frequent pattern tree and header table information according to the frequent item set; mining target frequent patterns based on the frequent pattern tree and header table information; and determining multiple target association rules according to the target frequent patterns.

[0060] Optionally, constructing a frequent item set based on product attribute information includes: performing a scanning operation on the product attribute information to obtain multiple product support values, wherein the product attribute information includes: multiple candidate product attributes, and there is a one-to-one correspondence between the multiple candidate product attributes and the multiple product support values; comparing a predetermined target support value with the multiple product support values ​​to determine a frequent item set, wherein the frequent item set includes multiple target product attributes, and multiple target product support values ​​corresponding to the multiple target product attributes are all greater than the target support value.

[0061] Specifically, the entire transaction data set is traversed, and the product attribute information in each transaction record is scanned. The "product attribute information" here includes multiple candidate product attributes, such as "five-star luxury hotel", "business hotel", "economy class air ticket", "domestic comprehensive accident insurance", etc. The purpose of the scan is to count the number of occurrences of each attribute in all transaction records to calculate its support value.

[0062] Specifically, during the scanning process, the frequency of occurrence of each product attribute is recorded, and the support of each product attribute is calculated based on the total size of the transaction data set. Support is defined as the frequency of an attribute appearing in all transactions, that is, the number of times the attribute appears in the transaction record divided by the total number of transactions. For example, if "economy class ticket" appears 800 times in 1,000 transactions, then its support is 80%.

[0063] For a certain product set I, its support is defined as follows:

[0064]

[0065] Where T represents the set of all transaction orders, |T| represents the number of elements in the transaction order set, I represents the set of several commodities, and T I represents the set consisting of all transaction orders in set I in set T, |T I | represents the set T I The number of elements in .

[0066] Specifically, after the calculation of the support of the product attributes is completed, these values ​​are compared with the predetermined "target support value". The target support value is an important threshold for mining frequent item sets, representing the minimum frequency of occurrence of the attribute that the user is interested in. If the support value of an attribute is higher than the target support value, it will be considered to be frequent and included in the frequent item set. For example, if the target support value is set to 60%, then only attributes with a support value of more than 60% will be selected.

[0067] Through the comparison and screening process, a frequent item set is finally generated. Each element in the set is a target product attribute, and corresponding to it is a "target product support value" associated with the attribute that is higher than the target support value. The frequent item set not only reflects which product attributes are more popular in the field of aviation tourism retail, but also provides basic data for identifying the association between different attributes. It is a key intermediate result in the entire dynamic packaging method.

[0068] For example, Table 1 shows the transaction records of purchasing goods.

[0069] Table 1 Commodity transaction record

[0070]

[0071]

[0072] Set the target support value to 55%, scan all transactions, filter out items with support greater than the target support value, and generate a frequent item set, each element of which contains an item and its support (the number of times the product attribute appears is used to represent the support), and arrange them in the order in which the products appear in the transaction.

[0073] Based on the commodity transaction records shown in Table 1, the frequent item set A is generated as follows:

[0074] A=<(Five-star luxury hotels: 4), (Business hotels: 4), (Beijing-Tokyo air tickets: 3), (Air China air tickets: 3), (Vegetarian airplane meals: 3), (Domestic comprehensive accident insurance: 3)>

[0075] Optionally, constructing a frequent pattern tree and header table information based on a frequent item set includes: determining that a root node of the frequent pattern tree is an empty node; performing a scanning operation on the product attribute information to obtain a plurality of frequent pattern tree branches; constructing a frequent pattern tree based on the root node and the plurality of frequent pattern tree branches; and determining header table information based on the frequent pattern tree.

[0076] Specifically, the first step in constructing a frequent pattern tree (FP-tree) is to define its root node, which is usually marked as a null node (NULL). It does not correspond to any specific product attribute, but serves as the starting point of the entire tree. The root node is the entrance of the FP-tree, and all frequent patterns will branch from here.

[0077] Specifically, after the root node is determined, a scan operation is performed on the product attribute information. Here, "product attribute information" refers to all product attributes whose support exceeds the target support value contained in the frequent item set obtained after preprocessing. During the scanning process, the branches of the FP-tree are constructed according to the product attribute sequence in the transaction record.

[0078] Specifically, when building branches, the following principles are followed: for each transaction record, the product attributes extracted from the frequent item set are arranged in descending order according to their support, and then the branches are built from the root node downward step by step. If an existing branch (i.e., the same attribute sequence) is encountered during the construction process, only the count value of the branch is updated, indicating that the frequency of this product attribute sequence in the transaction has increased. If a new attribute sequence is encountered, a new branch is created and inserted into the FP-tree.

[0079] Through the above multiple scans and branch construction, the entire FP-tree is gradually constructed. The final FP-tree consists of a root node and a series of frequent pattern tree branches. Each branch represents a set of frequently occurring product attribute sequences and their frequency of occurrence in transaction records. The data structure design of the FP-tree is intended to reduce data redundancy and repeated scanning and improve the efficiency of data mining.

[0080] Furthermore, a header table is constructed based on the constructed frequent pattern tree. The header table is a data table containing frequent items and their location information in the FP-tree. Each item in the header table contains frequent attributes (such as "five-star luxury hotel", "economy class air ticket") and a pointer, which points to the first node in the FP-tree where the attribute appears, and links to all nodes where the attribute appears. Through the header table, all occurrences of a specific frequent attribute in the FP-tree can be quickly located, thereby improving the efficiency of subsequent mining of association rules.

[0081] It should be noted that the process of constructing the header table involves traversing the FP-tree, finding the node where each frequent attribute first appears, and recording the location information of the node in the FP-tree. In this way, the header table not only stores the frequent attributes themselves, but also saves the path information for quickly accessing the distribution of these attributes in the FP-tree.

[0082] Exemplarily, based on the commodity transaction records shown in Table 1 and the established frequent item set A, the specific process of constructing the frequent pattern tree and header table information is as follows:

[0083] Set the root node of FP-Tree to NULL and rescan the transaction record database. Each time a new transaction record is processed, find the frequent item set of the transaction record that is identical to the existing tree branch, and then build a new branch of FP-Tree starting from the next different item, as follows.

[0084] Step 1: Find the common items in transaction number 100 with those in set A, and generate the first branch in the FP-Tree <(five-star luxury hotel: 1), (business hotel: 1), (Beijing-Tokyo air ticket: 1), (vegetarian airplane meal: 1), (domestic comprehensive accident insurance: 1)>.

[0085] Specifically, when transaction number 100 is scanned, there are 5 products with support greater than the target value, namely {Beijing-Tokyo air ticket, business hotel, five-star luxury hotel, vegetarian airplane meal, domestic comprehensive accident insurance}. Then the products are sorted in the order in which they appear in product set A. The result is recorded as <(five-star luxury hotel: 1), (business hotel: 1), (Beijing-Tokyo air ticket: 1), (vegetarian airplane meal: 1), (domestic comprehensive accident insurance: 1)>. Each node contains product attributes and corresponding support value.

[0086] Figure 2 is a schematic diagram of the result of the FP-Tree generated in the first step according to one embodiment of the present invention, Figure 2 The FP-Tree generated in the first step is shown.

[0087] Step 2: Scan the transaction with transaction number 200 and obtain the attribute set {Beijing-Tokyo air tickets, Air China air tickets, business hotels, five-star luxury hotels, vegetarian airplane meals} that is greater than the target support. Continue to arrange the elements in the order of set A to generate a sequence <five-star luxury hotel, business hotel, Beijing-Tokyo air tickets, Air China air tickets, vegetarian airplane meals>.

[0088] Since the second sequence <five-star luxury hotel, business hotel, Beijing-Tokyo air ticket, Air China air ticket, vegetarian airplane meal> has the same prefix as the first sequence <five-star luxury hotel, business hotel, Beijing-Tokyo air ticket>, the support of <five-star luxury hotel, business hotel, Beijing-Tokyo air ticket> is increased by one, and the nodes (Air China air ticket: 1) and (vegetarian airplane meal: 1) are added under the node (Beijing-Tokyo air ticket: 2). Therefore, the second branch in the FP-Tree is <(five-star luxury hotel: 2), (business hotel: 2), (Beijing-Tokyo air ticket: 2), (Air China air ticket: 1), (vegetarian airplane meal: 1)>, and the generated FP-Tree is as follows: Figure 3 shown.

[0089] Step 3: Scan the transaction with transaction number 300 to obtain the third sequence <five-star luxury hotel, Air China air ticket>.

[0090] Compared with the first two records, the third sequence <Five-star luxury hotel, Air China ticket> has only one common prefix <Five-star luxury hotel>. Therefore, we only need to add a node (Air China ticket: 1) under (Five-star luxury hotel: 3). The generated FP-Tree is as follows: Figure 4 shown.

[0091] According to the above method, the remaining transactions are scanned and processed. After the transaction number 500 is scanned, the complete FP-Tree is constructed as follows Figure 5 shown.

[0092] Step 4: Traverse the entire tree and create a header table of the FP-Tree. The first column of this table is the frequent items in descending order, and the second column is the pointer to the node position in the FP-Tree. Each node in the FP-Tree also has a pointer to the same name, such as Figure 6 shown.

[0093] Optionally, mining the target frequent pattern based on the frequent pattern tree and the header table information includes: performing a traversal operation on multiple frequent pattern tree branches based on the header table information to obtain the target frequent pattern.

[0094] Specifically, when traversing the FP-tree, we focus on those branch paths that end with specific frequent attributes. These paths are called conditional pattern bases. For example, when focusing on the frequent attribute "five-star luxury hotel", all branch paths ending with "five-star luxury hotel" will be traversed. Then, these conditional pattern bases are used as new "conditional" transaction data sets to construct a conditional FP-tree. The conditional FP-tree is an optimized version of the FP-tree for specific frequent attributes. It only contains patterns related to the target frequent attributes, thereby further compressing the data and reducing the amount of computation for subsequent mining.

[0095] Specifically, the conditional FP-tree is used to mine all frequent patterns ending with the target frequent attribute. This process is usually performed recursively, starting from the root node of the conditional FP-tree and traversing down along each branch path until a complete path ending with the target frequent attribute is found, i.e., a frequent pattern. For each frequent pattern found, its support is checked to see if it meets the preset threshold. Only patterns that meet the threshold will be retained as the final target frequent pattern.

[0096] It should be noted that after mining all frequent patterns ending with a specific frequent attribute, the pattern growth method can be used to try to merge these patterns to generate longer frequent patterns. For example, if the patterns "Beijing-Tokyo air tickets -> vegetarian airplane meals" and "Beijing-Tokyo air tickets -> domestic comprehensive accident insurance" have been found, the algorithm will try to check whether there is also a pattern "Beijing-Tokyo air tickets -> vegetarian airplane meals -> domestic comprehensive accident insurance", that is, whether multiple frequent attributes can be combined into a more complex frequent pattern. This process is achieved by checking the branch paths in the FP-tree. As long as a path combination with high enough support is found, a longer frequent pattern can be generated.

[0097] For each frequent attribute in the header table, the conditional pattern base is generated, the conditional FP-tree is constructed, and the frequent pattern is mined. When the entire header table is traversed, a complete list of all target frequent patterns is obtained. These target frequent patterns not only reflect which product attribute combinations frequently appear in transaction data, but also provide a basis for the subsequent association rule generation.

[0098] Exemplarily, based on the frequent pattern tree and header table information generated in the above embodiment, the specific process of further determining the target frequent pattern is as follows:

[0099] from Figure 6 The bottom of the header table shown starts mining frequent patterns in the FP-Tree.

[0100] In the FP-Tree, there are two node chains ending with domestic comprehensive accident insurance (starting from the NULL root node), namely <(five-star luxury hotel: 4), (business hotel: 3), (Beijing-Tokyo air tickets: 3), (vegetarian airplane meals: 2), (domestic comprehensive accident insurance: 2)> and <(business hotel: 1), (Air China air tickets: 1), (domestic comprehensive accident insurance: 1)>.

[0101] Among them, the first node chain list indicates that the list of items purchased by the customer <five-star luxury hotel, business hotel, Beijing-Tokyo air ticket, vegetarian airplane meal, domestic comprehensive accident insurance> appears twice in the database. Although <five-star luxury hotel, business hotel, Beijing-Tokyo air ticket> appears 3 times in the first node chain and the single item <five-star luxury hotel> appears 4 times, they only appear together with domestic comprehensive accident insurance 2 times, so in the conditional FP-Tree, <(five-star luxury hotel: 4), (business hotel: 3), (Beijing-Tokyo air ticket: 3), (vegetarian airplane meal: 2), (domestic comprehensive accident insurance: 2)> is recorded as <(five-star luxury hotel: 2), (business hotel: 2), (Beijing-Tokyo air ticket: 2), (vegetarian airplane meal: 2), (domestic comprehensive accident insurance: 2)>.

[0102] Similarly, the second node chain indicates that the list of items purchased by the customer <Business hotel, Air China ticket, domestic comprehensive accident insurance> only appears once in the database. The prefix node chain of domestic comprehensive accident insurance <(five-star luxury hotel: 2), (business hotel: 2), (Beijing-Tokyo ticket: 2), (vegetarian airplane meal: 2)> and <(business hotel: 1), (Air China ticket: 1)> are called the conditional pattern base of domestic comprehensive accident insurance. Now we have obtained the transaction records and support values ​​of all conditional pattern bases about the domestic comprehensive accident insurance node. Take the conditional pattern base of domestic comprehensive accident insurance as a new transaction database, repeat the above steps to construct FP-Tree, and obtain the conditional FP-Tree of domestic comprehensive accident insurance, recorded as mine <(business hotel: 3)|(domestic comprehensive accident insurance: 3)>, and the NULL node directly points to the (business hotel: 3) node.

[0103] In the FP-Tree, there are two node chains ending with vegetarian airplane meals, namely <(five-star luxury hotel: 4), (business hotel: 3), (Beijing-Tokyo air ticket: 3), (vegetarian airplane meal: 2)> and <(five-star luxury hotel: 4), (business hotel: 3), (Beijing-Tokyo air ticket: 3), (Air China air ticket: 1), (vegetarian airplane meal: 1)>. Therefore, the conditional pattern bases of vegetarian airplane meals are <(five-star luxury hotel: 2), (business hotel: 2), (Beijing-Tokyo air ticket: 2)> and <(five-star luxury hotel: 1), (business hotel: 1), (Beijing-Tokyo air ticket: 1), (Air China air ticket: 1)>. The conditional pattern base of vegetarian airplane meals is used as a new transaction database. Each row stores a prefix node chain of vegetarian airplane meals. The support of various items in each row of records is calculated, and then they are arranged in descending order of support. Only frequent item sets are retained, and those items below the support threshold are eliminated. The conditional FP-Tree of vegetarian airplane meals is established, such as Figure 7 shown.

[0104] There are 3 nodes in the conditional FP-Tree of vegetarian airplane meals, so it is necessary to recursively mine the frequent item set mine(<(five-star luxury hotel: 3), (business hotel: 3), (Beijing-Tokyo air ticket: 3)|(vegetarian airplane meal: 3)>) multiple times. Recursively call mine(<(five-star luxury hotel: 3), (business hotel: 3)|Beijing-Tokyo air ticket, vegetarian airplane meal>), mine(<(five-star luxury hotel: 3)|business hotel, vegetarian airplane meal>), mine(NULL|five-star luxury hotel, vegetarian airplane meal) in the order of <(Beijing-Tokyo air ticket: 3), (business hotel: 3), (five-star luxury hotel: 3)>.

[0105] from Figure 7It can be seen that the conditional FP-Tree of the node (Beijing-Tokyo flight ticket, vegetarian airline meal) has 2 nodes, and it is necessary to further recursively call mine(<(Five-star luxury hotel: 3)|Business hotel, Beijing-Tokyo flight ticket, vegetarian airline meal>) and mine(<NULL|Five-star luxury hotel, Beijing-Tokyo flight ticket, vegetarian airline meal>). Further recursively calling mine(<(Five-star luxury hotel: 3)|Business hotel, Beijing-Tokyo flight ticket, vegetarian airline meal>) generates mine(<NULL|Five-star luxury hotel, Business hotel, Beijing-Tokyo flight ticket, vegetarian airline meal>). Therefore, the frequent patterns ending with (Beijing-Tokyo flight ticket, vegetarian airline meal) are <(Beijing-Tokyo flight ticket vegetarian airline meal: 3), (Five-star luxury hotel Beijing-Tokyo flight ticket vegetarian airline meal: 3), (Business hotel Beijing-Tokyo flight ticket vegetarian airline meal: 3), (Five-star luxury hotel Business hotel Beijing-Tokyo flight ticket vegetarian airline meal: 3)>.

[0106] From Figure 7 It can be seen that the conditional FP-Tree of the node (Business hotel, vegetarian airline meal) has only 1 node, so it is only necessary to recursively call mine(<NULL|Five-star luxury hotel, Business hotel, vegetarian airline meal>). Therefore, the frequent patterns ending with (Business hotel, vegetarian airline meal) are <(Business hotel vegetarian airline meal: 3), (Five-star luxury hotel Business hotel vegetarian airline meal: 3)>. Similarly, the frequent patterns ending with (Five-star luxury hotel, vegetarian airline meal) are <(Five-star luxury hotel vegetarian airline meal: 3)>.

[0107] There are three node chains ending with Air China flight tickets in the FP-Tree, namely <(Five-star luxury hotel: 4), (Business hotel: 3), (Beijing-Tokyo flight ticket: 3), (Air China flight ticket: 1)>, <(Five-star luxury hotel: 4), (Air China flight ticket: 1)> and <(Business hotel: 1), (Air China flight ticket: 1)>. Since the conditional pattern bases of the node Air China flight ticket <(Five-star luxury hotel: 1), (Business hotel: 1), (Beijing-Tokyo flight ticket: 1)>, <(Five-star luxury hotel: 1)> and <(Business hotel: 1)> do not meet the support threshold, there is no need for further recursion. Therefore, the only frequent pattern ending with Air China flight ticket is <(Air China flight ticket: 3)>.

[0108] According to the above analysis method, by the same token, the frequent pattern ending with Beijing-Tokyo air ticket is <(five-star luxury hotel, Beijing-Tokyo air ticket: 3), (business hotel, Beijing-Tokyo air ticket: 3), (five-star luxury hotel, business hotel, Beijing-Tokyo air ticket: 3), Beijing-Tokyo air ticket: 3)>, the frequent pattern ending with business hotel is <(five-star luxury hotel, business hotel: 3), (business hotel: 4)>, and the frequent pattern ending with five-star luxury hotel is <(five-star luxury hotel: 4)>.

[0109] Optionally, determining a plurality of target association rules according to the target frequent pattern includes: determining a plurality of initial association rules according to the target frequent pattern; obtaining a plurality of confidence values ​​of the plurality of initial association rules; and determining a plurality of target association rules according to the plurality of confidence values.

[0110] Specifically, based on the mined target frequent patterns, a series of initial association rules are generated. These rules are usually expressed in the form of "predecessor -> consequent", where the predecessor and consequent are both sets of product attributes in the frequent pattern. For example, if the frequent pattern is "five-star luxury hotel, business hotel, Beijing-Tokyo air ticket", then the initial association rules that can be generated include "five-star luxury hotel -> business hotel, Beijing-Tokyo air ticket", "business hotel -> five-star luxury hotel, Beijing-Tokyo air ticket", etc. These rules reflect the potential correlation between different product attributes in the data set.

[0111] Furthermore, the reliability of each initial association rule is evaluated by calculating the confidence value of each initial association rule. The confidence value measures the frequency of the occurrence of the subsequent term when the preceding term occurs. The specific calculation formula is as follows:

[0112] For two sets of product attributes P and Q, the association rules The confidence level of is defined as follows:

[0113]

[0114] Specifically, after calculating the confidence value of each initial association rule, a "target confidence value" is set as the threshold of rule validity. Rules with confidence values ​​higher than the target confidence value will be considered reliable and retained as target association rules. For example, if the target confidence value is set to 0.6, then only rules with confidence values ​​greater than or equal to 0.6 will be included in the target association rule set.

[0115] It should be noted that in the process of screening association rules, it is necessary to pay attention to those rules that may be misleading, that is, the rules that appear to have high support and confidence meet the conditions, but in fact may not reflect the actual user preferences or behavior patterns. For example, consider the packaging case of 5,000 aviation tourism retail transaction orders. The data shows that 60% of the transaction orders have the product attribute "three-star comfort hotel", and 75% of the transaction orders have the product attribute "comprehensive accident insurance". Among them, 40% of the transaction orders (i.e. 2,000 orders) have not only the product attribute "three-star comfort hotel" but also the product attribute "comprehensive accident insurance". Assuming that the target support is 0.4 and the target confidence is 0.6, A is the set of transaction orders with the product attribute "three-star comfort hotel", and B is the set of transaction orders with the product attribute "comprehensive accident insurance". The following association rule is generated: [three-star comfort hotel=>comprehensive accident insurance], and the confidence c contained in this rule is 0.66, which is greater than the threshold value 0.6. However, the above association rule is misleading because the total proportion of set B is 75%, which is greater than 66%. In other words, the product attributes of "three-star comfort hotel" and "comprehensive accident insurance" are actually negatively correlated, and participating in one of the projects will reduce the possibility of participating in the other project. Therefore, in order to avoid the misleading of such rules, indicators such as promotion degree are used for further screening to ensure the effectiveness and practicality of the rules.

[0116] Specifically, association rules The lift formula is defined as follows:

[0117]

[0118] in, Association rules The lift degree, s(A∪B) is the support value of the set where attributes A and B exist together, s(A) is the support value of attribute A, and s(B) is the support value of attribute B.

[0119] If the lift value is greater than 1, the association rule screening is valid.

[0120] In the above example, the lift test of the association rule fails for the following reasons:

[0121]

[0122] Embodiments of the present invention also provide a kind of dynamic packaging device of aviation tourism products, for realizing the above-mentioned embodiment and preferred implementation mode, what has been described is not repeated here.As used below, term "module" can realize the software and / or hardware combination of predetermined function.Although the device described in the following embodiment is preferably realized by software, the realization of hardware, or the combination of software and hardware is also possible and conceived.

[0123] Figure 8 FIG. 3 is a structural block diagram of a dynamic packaging device 300 for aviation travel products according to one embodiment of the present invention. Figure 8 As shown, the device includes: a first acquisition module 301, used to obtain product attribute information of multiple aviation tourism products; a determination module 302, used to determine multiple target association rules according to the product attribute information; a second acquisition module 303, used to obtain a target product attribute set based on the multiple target association rules; a packaging module 304, used to dynamically package multiple aviation tourism products based on the target product attribute set.

[0124] Optionally, the first acquisition module 301 is further used to: determine multiple product features of multiple aviation travel products; acquire multiple product attributes corresponding to the multiple product features to obtain product attribute information.

[0125] Optionally, the determination module 302 is further used to: construct a frequent item set according to product attribute information; construct a frequent pattern tree and header table information according to the frequent item set; mine target frequent patterns based on the frequent pattern tree and header table information; and determine multiple target association rules according to the target frequent patterns.

[0126] Optionally, the determination module 302 is further used to: perform a scanning operation on the product attribute information to obtain multiple product support values, wherein the product attribute information includes: multiple candidate product attributes, and the multiple candidate product attributes correspond one-to-one to the multiple product support values; compare a predetermined target support value with the multiple product support values ​​to determine a frequent item set, wherein the frequent item set includes multiple target product attributes, and the multiple target product support values ​​corresponding to the multiple target product attributes are all greater than the target support value.

[0127] Optionally, the determination module 302 is further used to: determine that the root node of the frequent pattern tree is an empty node; perform a scanning operation on the product attribute information to obtain multiple frequent pattern tree branches; construct a frequent pattern tree based on the root node and the multiple frequent pattern tree branches; and determine header table information according to the frequent pattern tree.

[0128] Optionally, the determination module 302 is further configured to: perform a traversal operation on multiple frequent pattern tree branches based on the header table information to obtain a target frequent pattern.

[0129] Optionally, the determination module 302 is further configured to: determine a plurality of initial association rules according to the target frequent pattern; obtain a plurality of confidence values ​​of the plurality of initial association rules; and determine a plurality of target association rules according to the plurality of confidence values.

[0130] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the dynamic packaging method for aviation travel products described in any of the above embodiments.

[0131] Optionally, in this embodiment, the processor in the electronic device is configured to run a computer program to perform the following steps:

[0132] Step S101, obtaining product attribute information of a plurality of aviation travel products.

[0133] Step S102: determining a plurality of target association rules according to product attribute information.

[0134] Step S103, obtaining a target product attribute set based on multiple target association rules.

[0135] Step S104: dynamically package multiple aviation travel products based on the target product attribute set.

[0136] Optionally, the processor in the above electronic device is configured to run a computer program to perform the following steps: determining multiple product features of multiple aviation travel products; acquiring multiple product attributes corresponding to the multiple product features to obtain product attribute information.

[0137] Optionally, the processor in the above electronic device is configured to run a computer program to perform the following steps: construct a frequent item set based on product attribute information; construct a frequent pattern tree and header table information based on the frequent item set; mine target frequent patterns based on the frequent pattern tree and header table information; and determine multiple target association rules based on the target frequent patterns.

[0138] Optionally, the processor in the above-mentioned electronic device is configured to run a computer program to perform the following steps: perform a scanning operation on the product attribute information to obtain multiple product support values, wherein the product attribute information includes: multiple candidate product attributes, and there is a one-to-one correspondence between the multiple candidate product attributes and the multiple product support values; compare a predetermined target support value with the multiple product support values ​​to determine a frequent item set, wherein the frequent item set includes multiple target product attributes, and the multiple target product support values ​​corresponding to the multiple target product attributes are all greater than the target support value.

[0139] Optionally, the processor in the above electronic device is configured to run a computer program to perform the following steps: determine that the root node of the frequent pattern tree is an empty node; perform a scanning operation on the product attribute information to obtain multiple frequent pattern tree branches; construct a frequent pattern tree based on the root node and the multiple frequent pattern tree branches; and determine the header table information based on the frequent pattern tree.

[0140] Optionally, the processor in the electronic device is configured to run a computer program to perform the following steps: performing a traversal operation on multiple frequent pattern tree branches based on the header table information to obtain a target frequent pattern.

[0141] Optionally, the processor in the above electronic device is configured to run a computer program to perform the following steps: determine multiple initial association rules according to the target frequent pattern; obtain multiple confidence values ​​of the multiple initial association rules; and determine multiple target association rules according to the multiple confidence values.

[0142] An embodiment of the present invention further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the dynamic packaging method of aviation travel products described in any of the above embodiments when running on a computer or a processor.

[0143] Optionally, in this embodiment, the computer program is configured to store a computer program for performing the following steps:

[0144] Step S101, obtaining product attribute information of a plurality of aviation travel products.

[0145] Step S102: determining a plurality of target association rules according to product attribute information.

[0146] Step S103, obtaining a target product attribute set based on multiple target association rules.

[0147] Step S104: dynamically package multiple aviation travel products based on the target product attribute set.

[0148] Optionally, the computer program is configured to store a computer program for executing the following steps: determining a plurality of product features of a plurality of aviation travel products; acquiring a plurality of product attributes corresponding to the plurality of product features to obtain product attribute information.

[0149] Optionally, the above-mentioned computer program is configured to store a computer program for executing the following steps: constructing a frequent item set based on product attribute information; constructing a frequent pattern tree and header table information based on the frequent item set; mining a target frequent pattern based on the frequent pattern tree and header table information; and determining multiple target association rules based on the target frequent pattern.

[0150] Optionally, the above-mentioned computer program is configured to store a computer program for executing the following steps: performing a scanning operation on product attribute information to obtain multiple product support values, wherein the product attribute information includes: multiple candidate product attributes, and there is a one-to-one correspondence between the multiple candidate product attributes and the multiple product support values; comparing a predetermined target support value with the multiple product support values ​​to determine a frequent item set, wherein the frequent item set includes multiple target product attributes, and the multiple target product support values ​​corresponding to the multiple target product attributes are all greater than the target support value.

[0151] Optionally, the above-mentioned computer program is configured to store a computer program for executing the following steps: determining that the root node of the frequent pattern tree is an empty node; performing a scanning operation on the product attribute information to obtain multiple frequent pattern tree branches; constructing a frequent pattern tree based on the root node and the multiple frequent pattern tree branches; and determining header table information according to the frequent pattern tree.

[0152] Optionally, the computer program is configured to store a computer program for executing the following steps: performing a traversal operation on multiple frequent pattern tree branches based on the header table information to obtain a target frequent pattern.

[0153] Optionally, the computer program is configured to store a computer program for executing the following steps: determining multiple initial association rules according to the target frequent pattern; obtaining multiple confidence values ​​of the multiple initial association rules; and determining multiple target association rules according to the multiple confidence values.

[0154] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be described in detail here.

[0155] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0156] In some embodiments provided in the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0157] 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, that is, they may be located in one place or distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0158] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0159] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a grid device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc., which can store program codes.

[0160] The above are only preferred embodiments of the present invention. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A dynamic packaging method for aviation tourism products, characterized in that: include: Get product attribute information of multiple aviation travel products; Determine multiple target association rules according to the product attribute information; Acquire a target product attribute set based on the multiple target association rules; The plurality of aviation travel products are dynamically packaged based on the target product attribute set.

2. The dynamic packaging method of aviation tourism products according to claim 1, characterized in that: Acquiring the product attribute information of the plurality of aviation travel products includes: determining multiple product features for multiple air travel products; Acquire multiple product attributes corresponding to the multiple product features to obtain the product attribute information.

3. The dynamic packaging method of aviation tourism products according to claim 1, characterized in that: Determining the plurality of target association rules according to the product attribute information includes: Constructing a frequent item set according to the product attribute information; Constructing a frequent pattern tree and header table information according to the frequent item set; Mining a target frequent pattern based on the frequent pattern tree and the header table information; The plurality of target association rules are determined according to the target frequent patterns.

4. The dynamic packaging method of aviation tourism products according to claim 3, characterized in that: Constructing the frequent item set according to the product attribute information includes: Performing a scanning operation on the product attribute information to obtain a plurality of product support values, wherein the product attribute information includes: a plurality of candidate product attributes, and the plurality of candidate product attributes correspond to the plurality of product support values ​​in a one-to-one manner; The predetermined target support value is compared with the multiple product support values ​​to determine the frequent item set, wherein the frequent item set includes multiple target product attributes, and multiple target product support values ​​corresponding to the multiple target product attributes are all greater than the target support value.

5. The dynamic packaging method of aviation tourism products according to claim 3, characterized in that: Constructing the frequent pattern tree and the header table information according to the frequent item set includes: Determine that the root node of the frequent pattern tree is an empty node; Performing a scanning operation on the product attribute information to obtain a plurality of frequent pattern tree branches; Constructing the frequent pattern tree based on the root node and the multiple frequent pattern tree branches; The header table information is determined according to the frequent pattern tree.

6. The dynamic packaging method of aviation tourism products according to claim 5, characterized in that: Mining the target frequent pattern based on the frequent pattern tree and the header table information includes: A traversal operation is performed on the multiple frequent pattern tree branches based on the header table information to obtain the target frequent pattern.

7. The dynamic packaging method of aviation tourism products according to claim 3, characterized in that: Determining the plurality of target association rules according to the target frequent pattern comprises: Determine a plurality of initial association rules according to the target frequent pattern; Obtaining multiple confidence values ​​of the multiple initial association rules; The plurality of target association rules are determined according to the plurality of confidence values.

8. A dynamic packaging device for aviation tourism products, characterized in that: include: A first acquisition module is used to acquire product attribute information of multiple aviation travel products; A determination module, used to determine a plurality of target association rules according to the product attribute information; A second acquisition module, configured to acquire a target product attribute set based on the multiple target association rules; The packaging module is used to dynamically package the multiple aviation travel products based on the target product attribute set.

9. An electronic device, comprising a memory and a processor, characterized in that: The memory stores a computer program, and the processor is configured to run the computer program to execute the dynamic packaging method for aviation travel products as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program is configured to execute the dynamic packaging method for aviation travel products as described in any one of claims 1 to 7 when running on a computer or a processor.

Images