E-commerce transaction data cleaning method and system

Abstract

The invention discloses an electronic commerce transaction data cleaning method and system, and belongs to the technical field of privacy protection, and the method comprises the steps: obtaining an original transaction database; mining a rare item set of the original transaction database, extracting a sensitive rare item set in the rare item set according to a preset transaction demand, and dividing the original transaction database and the sensitive rare item set into a sensitive database and a non-sensitive database; and cleaning the sensitive database to obtain cleaned data, and obtaining a cleaned database according to the non-sensitive database and the cleaned data. According to the method, the sensitive rare item set is hidden by identifying and selecting proper victim transactions and victim items, so that a complete privacy protection effect on electronic commerce transaction data is realized, and when the same rare mining threshold is applied, the sensitive rare item set of transactions cannot be mined.

Description

Technical Field

[0001] The present invention relates to the field of privacy protection technology, and in particular to a method and system for cleaning e-commerce transaction data. Background Art

[0002] With the rapid development of information technology, data mining technology has been widely used in the field of e-commerce, helping platforms to extract valuable patterns and information from huge transaction data. As an important task in data mining, rare item set mining has gradually attracted the attention of researchers in recent years. Rare item set mining aims to discover those item sets that appear less frequently in the data set but have higher potential commercial value. These rare item sets often reflect special consumption behaviors or personalized preferences and have important practical application significance. For example, by mining rare item sets, e-commerce platforms can discover consumption patterns such as users' occasional large purchases and unpopular product combination purchases, so as to recommend products more accurately and improve marketing effects. However, with the increasing demand for privacy protection from users, how to protect data privacy when conducting rare item set mining has become a key issue that needs to be solved urgently.

[0003] E-commerce platforms process a large amount of user data every day, and analyze this data through data mining algorithms to provide support for business optimization. However, the data mining process involves a large amount of personal sensitive information, such as user identity, consumption records, and payment preferences. This sensitive information is vulnerable to attack and abuse during the processing process, bringing the risk of privacy leakage. In the current information age, the commercial value of data is becoming more and more prominent, and sensitive data leakage incidents are increasing year by year, further exacerbating privacy risks. As an important means to deal with this problem, Privacy-Preserving Data Mining (PPDM) technology has gradually received attention and application in e-commerce data analysis in recent years.

[0004] Currently, in the field of e-commerce, there is little research on privacy protection of rare itemsets, which results in a large loss of data utility. On some data sets, the cleaning process will produce more serious side effects, leading to the destruction of important non-sensitive consumer behavior patterns. Summary of the invention

[0005] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a method and system for cleaning e-commerce transaction data, which can achieve higher privacy protection accuracy and avoid leakage of sensitive data caused by incomplete cleaning.

[0006] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0007] In one aspect, the present invention provides a method for cleaning e-commerce transaction data, comprising:

[0008] Obtain the original transaction database;

[0009] Mining the rare item sets of the original transaction database, extracting sensitive rare item sets from the rare item sets according to preset transaction requirements, and dividing the original transaction database and the sensitive rare item sets into a sensitive database and a non-sensitive database;

[0010] Cleaning the sensitive database to obtain cleaned data, and obtaining a cleaned database based on the non-sensitive database and the cleaned data;

[0011] Wherein, the cleaning of the sensitive database includes:

[0012] Calculate the item count of each rare 1-itemset in the rare itemset, the item count of each rare 1-itemset in the insensitive rare itemset, and the rare count of each transaction;

[0013] selecting a victim transaction according to a rare count, and selecting a victim item in the victim transaction according to an item count;

[0014] The victim item is deleted in the victim transaction to obtain cleansed data.

[0015] Optionally, the cleaning of the sensitive database further includes:

[0016] Construct transaction retrieval table and support index table for sensitive rare itemsets.

[0017] Optionally, selecting a victim transaction according to a rare count, and selecting a victim item in the victim transaction according to an item count, comprises:

[0018] Select the transaction with the smallest rare count as the victim transaction;

[0019] In the victim transaction, an item with the smallest item count in the rare 1-item set in the rare item set is selected as a first victim item, and in the victim transaction, an item with the largest item count in the rare 1-item set in the non-sensitive rare item set is selected as a second victim item;

[0020] The first victim item and the second victim item are deleted from the victim transaction to obtain cleaned data, and other sensitive rare item sets, transaction retrieval tables, and support index tables are updated.

[0021] Optionally, also include:

[0022] Calculate the difference between the support of the sensitive rare itemset and the preset threshold;

[0023] When the difference between the support of the sensitive rare itemset and a preset threshold is less than or equal to zero, the cleaning of the sensitive database is stopped.

[0024] Optionally, according to preset parameters, rare itemsets of the original transaction database are mined, and the rare itemsets satisfy:

[0025] ;

[0026] ;

[0027] In the formula, represents the first parameter of the preset; Indicates the preset second parameter; Representing Item Sets Frequency of occurrence in the original transaction database; Representing Item Sets Support in the original transaction database; represents the original transaction database; Represents the total number of transactions in the raw transaction database.

[0028] Optionally, the rare count of each transaction is expressed as:

[0029] ;

[0030] In the formula, Indicates Rare count of transactions; Indicates The number of times rare itemsets appear in a transaction; Indicates transactions; Indicates The length of a transaction.

[0031] In a second aspect, the present invention provides a system for cleaning e-commerce transaction data, comprising:

[0032] Data acquisition module, to obtain the original transaction database;

[0033] A data partitioning module, mining rare item sets of the original transaction database, extracting sensitive rare item sets from the rare item sets according to preset transaction requirements, and dividing the original transaction database and the sensitive rare item sets into a sensitive database and a non-sensitive database;

[0034] A data cleaning module cleans the sensitive database to obtain cleansed data, and obtains a cleansed database based on the non-sensitive database and the cleansed data;

[0035] Cleaning of sensitive databases includes:

[0036] Calculate the item count of each rare 1-itemset in the rare itemset, the item count of each rare 1-itemset in the insensitive rare itemset, and the rare count of each transaction;

[0037] selecting a victim transaction according to a rare count, and selecting a victim item in the victim transaction according to an item count;

[0038] The victim item is deleted in the victim transaction to obtain cleansed data.

[0039] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for cleaning e-commerce transaction data described in the first aspect.

[0040] In a fourth aspect, the present invention provides a computer system, comprising:

[0041] Memory, for storing computer instructions;

[0042] The processor is used to execute the computer instructions to implement the steps of the e-commerce transaction data cleaning method described in the first aspect.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] The present invention hides sensitive rare item sets by identifying and selecting appropriate victim transactions and victim items, thereby achieving complete privacy protection for e-commerce transaction data. When the same rare mining threshold is applied, sensitive transaction rare item sets will not be mined. Compared with common encryption and perturbation methods, higher privacy protection accuracy is achieved, and sensitive data leakage caused by incomplete cleaning is avoided. No false costs are generated, and the data loss caused in intensive data sets is relatively small, ensuring the overall integrity of the database structure and the high availability of data. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 FIG. 1 is a flow chart of a method for cleaning e-commerce transaction data in an embodiment of the present invention;

[0046] Figure 2 Shown is a schematic diagram of the evaluation relationship of the e-commerce transaction data after cleaning in one embodiment of the present invention. DETAILED DESCRIPTION

[0047] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. The embodiments of the present invention and the technical features in the embodiments may be combined with each other unless there is a conflict.

[0048] The term "and / or" is only a description of the association relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " generally indicates that the related objects are in an "or" relationship.

[0049] Example 1

[0050] like Figure 1 As shown, this embodiment introduces a method for cleaning e-commerce transaction data, designs two algorithm strategies, namely the Maximum Ratio Combining-MIN (MRC-MIN) algorithm and the Maximum Ratio Combining-MAX (MRC-MAX) algorithm, and optimizes the cleaning operation through different selection rules. First, the e-commerce transaction database is preprocessed to calculate the frequency of occurrence (rare correlation number) and support information of each rare consumer item set in different transactions; next, by gradually adjusting the support of the rare item set, the support of the sensitive rare item set is reduced to below a given threshold to ensure that the sensitive item set cannot be mined again under the same mining conditions. This method maximizes the integrity of the database structure while minimizing the impact on non-sensitive rare item sets, and specifically includes the following steps:

[0051] Step 1: Obtain the original transaction database.

[0052] Step 2: Mining the rare item sets of the original transaction database, extracting the sensitive rare item sets in the rare item sets according to the preset transaction requirements, and dividing the original transaction database and the sensitive rare item sets into sensitive databases and non-sensitive databases. The preset transaction requirements can be the requirements of the task, or the needs of the government or enterprises, which are subjective to a certain extent.

[0053] According to preset parameters, rare item sets of the original transaction database are mined, and the rare item sets satisfy:

[0054] ;

[0055] ;

[0056] In the formula, represents the first parameter of the preset; Indicates the preset second parameter; ( ); Representing Item Sets Frequency of occurrence in the original transaction database; Representing Item Sets Support in the original transaction database; represents the original transaction database; Represents the total number of transactions in the raw transaction database.

[0057] Step 3: Clean the sensitive database to obtain cleaned data, and obtain a cleaned database based on the non-sensitive database and the cleaned data, wherein the cleaning of the sensitive database includes:

[0058] Calculate the item count of each rare 1-item set in the rare item set, the item count of each rare 1-item set in the non-sensitive rare item set, and the rare count of each transaction, and build a transaction retrieval table and support index table for sensitive rare item sets. Establish an index table to store the relationship information between rare item sets, so as to avoid repeated scanning of the database when processing information, reduce operation time and optimize efficiency. When the victim item is removed, the support of the related item set can be quickly updated to improve the cleaning efficiency. The rare count of each transaction is expressed as:

[0059] ;

[0060] In the formula, Indicates Rare count of transactions; Indicates The number of times rare itemsets appear in a transaction; Indicates transactions; Indicates The length of a transaction.

[0061] Use the MRC-MIN algorithm strategy to clean the sensitive database, including:

[0062] Select the transaction with the smallest rare count as the victim transaction to minimize the impact on non-sensitive data;

[0063] In the victim transaction, the item with the smallest item count in the rare 1-item set is selected as the first victim item to minimize the impact range of each cleaning;

[0064] The first victim item is deleted from the victim transaction to obtain the first cleaned data, and other sensitive rare item sets, transaction retrieval tables and support index tables are updated to reduce the support of sensitive rare item sets and ensure the accuracy of the next round of cleaning.

[0065] Use the MRC-MAX algorithm strategy to clean the sensitive database, including:

[0066] Select the transaction with the smallest rare count as the victim transaction to minimize the impact on non-sensitive data;

[0067] In the victim transaction, an item with the largest item count in the rare 1-item set in the non-sensitive rare item set is selected as the second victim item to maximize the cleaning effect each time;

[0068] The second victim item is deleted from the victim transaction to obtain the second cleaned data, and other sensitive rare item sets, transaction retrieval tables and support index tables are updated to reduce the support of sensitive rare item sets and ensure the accuracy of the next round of cleaning.

[0069] The first cleaned data and the second cleaned data constitute cleaned data.

[0070] Calculate the difference between the support of the sensitive rare itemset and a preset threshold; when the difference between the support of the sensitive rare itemset and the preset threshold is greater than zero, it indicates that the support has not dropped below the preset threshold, and the cleaning cycle will continue; when the difference between the support of the sensitive rare itemset and the preset threshold is less than or equal to zero, stop cleaning the sensitive database. In this embodiment, the preset threshold represents the minimum threshold for rare mining.

[0071] The evaluation of the cleaning results of this embodiment is as follows Figure 2 As shown, the evaluation is carried out through three basic rating criteria and database structure integrity criteria, including:

[0072] Hiding failure describes the situation where the privacy item set fails to be hidden under the same rare parameter mining condition, that is, the rare item set that should be hidden is still mined. Figure 2 As shown, Sensitive rare item sets mined from the cleaned database, generally, Represents the sensitive rare itemsets that are not hidden during the hiding process The total number, expressed as:

[0073] .

[0074] The loss cost describes the situation where the non-sensitive private itemsets that should have been mined are not mined after the cleaning process. Figure 2 As shown, To clean the database, we need to mine non-sensitive rare itemsets. Generally, represents the set of non-sensitive rare items lost in the hiding process ( ) total, expressed as:

[0075] .

[0076] The false cost describes that after cleaning, itemsets that should not be rare itemsets are mined. This is partly due to false information generated by modifying the database, such as Figure 2As shown, generally, The total number of non-rare itemsets mined incorrectly is expressed as:

[0077] .

[0078] In order to understand the relationship between them more clearly, we can use the following figure to better observe the inclusion relationship. Assume that there is an original transaction database And its cleaning database , then the rare item sets mined by them are and . It can also be divided into sensitive rare itemsets ( ) and non-sensitive rare itemsets ( ). As shown in the figure, It represents the mining of the privacy rare itemsets that should be hidden in the cleansing database. Represents the non-sensitive rare itemsets that should be mined but are not mined in the cleaned database due to the cleaning operation. It represents the false rare itemset that does not exist in the original rare itemset.

[0079] The cleaning method in this embodiment achieves privacy protection for sensitive rare itemsets in e-commerce transaction data by efficiently selecting and optimizing cleaning operations, ensuring that the platform hides users' unique consumption behavior patterns in data analysis while maintaining the validity and accuracy of other non-sensitive data. This privacy protection scheme can improve the data utilization efficiency and user privacy security of e-commerce platforms.

[0080] Example 2 This embodiment introduces a specific experimental example of a method for cleaning e-commerce transaction data, including the following steps: Get the transaction database of e-commerce , Represents the unique transaction number. Representative transactions. Each transaction . Represents the first Item sets. If an item set contains the number of items , then this item set is called k-itemset (i.e. containing A small transaction database example is provided here to help better understand the process. As shown in Table 1, the transaction number represents different transactions, and the letters in the transaction represent different commodities.

[0081] Table 1 Transaction database

[0082] Itemset In the transaction database The support in It represents the item set In the transaction database The number of times it appears in . For example, in Table 1, the item set ( a, c ) occurs 3 times, so .

[0083] Itemset The frequency of occurrence in the transaction database is expressed as Indicates. Represents a transaction database The total number of transactions is Similarly, we can conclude that .

[0084] Given two parameters ,in , when the frequency of the item set in the transaction database meets the condition , then we can call the itemset For rare itemsets .

[0085] For example, in Table 1, we give a sample transaction database consisting of 10 transactions. Assume that the rare itemset mining threshold is (0.3, 0.5). Then according to the rare itemset mining rule, we will get the rare itemsets of this database, and the results are shown in Table 2. All itemsets with support of 3 and 4 are mined, which means that these products show rare connection rules when shopping.

[0086] Table 2 Rare itemsets

[0087] Sensitive rare itemsets ( )∈ rare itemset( ), sensitive rare item sets usually contain sensitive user behavior patterns that need to be protected, such as purchase combinations of specific products, rare consumption habits, etc., which are usually determined according to the privacy protection level or business needs. For example, in the anti-fraud detection of e-commerce platforms, users' high-value single consumption, combination purchases of special products, or abnormal purchase behaviors in a specific period belong to sensitive rare item sets and need to be hidden and protected during the data mining process. In this example, we assume that (d), (a,c), (b,k) are sensitive rare item sets.

[0088] The transaction retrieval table is used to store the transaction number set of each sensitive rare item set to avoid repeated scanning of the database during the cleaning process. By establishing a sensitive transaction index table in advance, the efficiency of the cleaning process can be greatly improved.

[0089] For example, sensitive rare itemsets (d) → , (a, c) → , (b, k) → .

[0090] In the transaction database In, trading The number of rare itemsets that appear in . Indicates transaction The length of the transaction The rare count of can be expressed as: This metric measures the density of rare behaviors in transactions.

[0091] In the transaction database, under the given rare threshold condition, the rare count of each transaction can be obtained based on the mined rare item set. For example, Rcount(T1)=6 / 5, Rcount(T4)=6 / 4, Rcount(T7)=3 / 4, Rcount(T9)=11 / 6, Rcount(T 10 )=8 / 5.

[0092] In the transaction database, item count Indicates a specific product or consumption behavior item The number of times it appears in the sensitive rare itemset. Assume that the sensitive rare itemset is (d), (a, c), (b, k).

[0093] Assume that item sets (d), (a,c), (b,k) are sensitive rare item sets. Then, the corresponding item counts in the rare item sets are Icount(a)=6, Icount(b)=4, Icount(c)=1, Icount(d)=1, Icount(k)=3.

[0094] For example, if the item set (d), (a, c), (b, k) is a sensitive rare item set, first calculate the rare count of each transaction, as shown in Table 3.

[0095] Table 3 Rare count values

[0096] Generate sensitive transaction retrieval table of item set (d): (d) → Since the rare count value of sensitive transactions is the smallest, it is concluded that the transaction Delete the term (d) in =(c, f, g). Update Rcount(T2)=2 / 3. Remove the retrieval table of (d), (d) → . Update the support of item set (d) , as shown in Table 4.

[0097] Table 4 Updated database

[0098] This process is repeated until the itemset support is less than a given threshold.

[0099] When processing multi-item sets, the object selected is the item with the smallest item count, such as Icount(a)=6, Icount(c)=1. Generally, we prefer (c) as the item to be cleared. First generate ( )'s sensitive transaction retrieval table: ( )→(T1, T3, T5). Since the Rcount of transaction T3 is the smallest, we delete (c) from T3 and get T3=( ). After updating the relevant data, repeat this process, and the database after complete cleaning is shown in Table 5.

[0100] Table 5 Database after complete cleaning

[0101] The rare itemsets are mined again on the cleaned database to obtain the rare itemsets obtained by cleaning the database, as shown in Table 6.

[0102] Table 6 Rare itemsets obtained by cleaning the database

[0103] It can be seen that the privacy rare itemset (d), ( ), (b, k) has been hidden and has not been mined, and other non-sensitive rare item sets can still be mined normally. This means that in the transaction database of e-commerce, after being processed by the cleaning method of this embodiment, the sensitive rare item set is successfully hidden and cannot be identified and extracted in subsequent data mining. The user's specific sensitive behavior patterns (such as high-value single consumption, combination purchase of specific goods, etc.) have been effectively protected to prevent these data from being leaked during the analysis and sharing process. At the same time, other non-sensitive rare item sets can still be mined normally, ensuring that the platform obtains the required data support in analytical applications such as personalized recommendations, precision marketing, and inventory management.

[0104] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0105] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0106] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0107] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0108] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

Claims

1. A method for cleaning e-commerce transaction data, characterized in that: include: Obtain the original transaction database; Mining the rare item sets of the original transaction database, extracting sensitive rare item sets from the rare item sets according to preset transaction requirements, and dividing the original transaction database and the sensitive rare item sets into a sensitive database and a non-sensitive database; Cleaning the sensitive database to obtain cleaned data, and obtaining a cleaned database based on the non-sensitive database and the cleaned data; Wherein, the cleaning of the sensitive database includes: Calculate the item count of each rare 1-itemset in the rare itemset, the item count of each rare 1-itemset in the insensitive rare itemset, and the rare count of each transaction; selecting a victim transaction according to a rare count, and selecting a victim item in the victim transaction according to an item count; The victim item is deleted in the victim transaction to obtain cleansed data.

2. The method for cleaning e-commerce transaction data according to claim 1, characterized in that: The cleaning of the sensitive database also includes: Construct transaction retrieval table and support index table for sensitive rare itemsets.

3. The method for cleaning e-commerce transaction data according to claim 2, characterized in that: Selecting a victim transaction according to a rare count, and selecting a victim item in the victim transaction according to an item count, comprising: Select the transaction with the smallest rare count as the victim transaction; In the victim transaction, an item with the smallest item count in the rare 1-item set in the rare item set is selected as a first victim item, and in the victim transaction, an item with the largest item count in the rare 1-item set in the non-sensitive rare item set is selected as a second victim item; The first victim item and the second victim item are deleted from the victim transaction to obtain cleaned data, and other sensitive rare item sets, transaction retrieval tables, and support index tables are updated.

4. The method for cleaning e-commerce transaction data according to claim 2 or 3, characterized in that: Also includes: Calculate the difference between the support of the sensitive rare itemset and the preset threshold; When the difference between the support of the sensitive rare itemset and a preset threshold is less than or equal to zero, the cleaning of the sensitive database is stopped.

5. The method for cleaning e-commerce transaction data according to claim 1, characterized in that: According to preset parameters, rare item sets of the original transaction database are mined, and the rare item sets satisfy: ； ； In the formula, represents the first parameter of the preset; Indicates the preset second parameter; Representing Item Sets Frequency of occurrence in the original transaction database; Representing Item Sets Support in the original transaction database; represents the original transaction database; Represents the total number of transactions in the raw transaction database.

6. The method for cleaning e-commerce transaction data according to claim 1, characterized in that: The rare count of each transaction is expressed as: ； In the formula, Indicates Rare count of transactions; Indicates The number of times rare itemsets appear in a transaction; Indicates transactions; Indicates The length of a transaction.

7. A cleaning system for e-commerce transaction data, characterized in that: include: Data acquisition module, to obtain the original transaction database; A data partitioning module, mining rare item sets of the original transaction database, extracting sensitive rare item sets from the rare item sets according to preset transaction requirements, and dividing the original transaction database and the sensitive rare item sets into a sensitive database and a non-sensitive database; A data cleaning module cleans the sensitive database to obtain cleansed data, and obtains a cleansed database based on the non-sensitive database and the cleansed data; Cleaning of sensitive databases includes: Calculate the item count of each rare 1-itemset in the rare itemset, the item count of each rare 1-itemset in the insensitive rare itemset, and the rare count of each transaction; selecting a victim transaction according to a rare count, and selecting a victim item in the victim transaction according to an item count; The victim item is deleted in the victim transaction to obtain cleansed data.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for cleaning e-commerce transaction data described in any one of claims 1 to 7 are implemented.

9. A computer system, characterized in that: include: Memory, for storing computer instructions; A processor is used to execute the computer instructions to implement the steps of the method for cleaning e-commerce transaction data according to any one of claims 1 to 7.

Images