CN104732419A - Application of positive and negative sequence mode screening method in customer purchasing behavior analysis - Google Patents

Abstract

The invention relates to the application of positive and negative sequence pattern screening methods in the analysis of customer purchase behavior, and proposes an efficient algorithm named SAP to select feasible positive and negative sequence patterns. The main idea of the algorithm is to mine through the e-NSP method All positive sequence patterns and negative sequence patterns are extracted, and then each sequence pattern is screened to select feasible positive and negative sequence patterns. The present invention is applied in the process of analyzing the customer's purchase behavior, not only using the minimum support to screen out the products that the customer purchases more in a certain period of time, but also using the correlation coefficient to screen out the goods that have a relatively high correlation with the customer's purchase during this period of time In this way, when a customer purchases a product, the present invention can be used to recommend to him some products that other customers will buy and have a relatively high correlation with this product, thereby increasing the transaction opportunities of the customer and converting the website browser into a buyer. Improve cross-selling capabilities, increase customer loyalty, and improve website profitability.

Description

Application of Positive and Negative Sequential Pattern Screening Method in Analysis of Customer Purchasing Behavior

technical field

The invention relates to the application of a positive and negative sequence pattern screening method in the analysis of customer purchase behavior, and belongs to the technical field of sequence pattern mining applications.

Background technique

Along with the popularization of Internet and the development of e-commerce business, online shopping has also become one of people's main shopping methods. We only need to operate online without going out of the house to get more satisfactory products. However, with the surge in the amount of information, online shopping has become complicated and time-consuming. At the same time, many large-scale e-commerce websites, such as Amazon, Alibaba’s Taobao and Tmall Mall, and JD.com, have accumulated a large amount of customer transaction data. . How to make full use of these data, obtain customers' shopping patterns, and recommend personalized products to customers, so as to better improve the service quality and economic benefits of the website, is an urgent problem to be solved by e-commerce.

Compared with traditional business methods, e-commerce merchants cannot intuitively understand customers and obtain limited relevant data (such as customer registration information, purchase records, etc.). Through the analysis and mining of a large number of customer purchase records, the frequent access sequence pattern of customers is discovered, and different product recommendation forms are adopted for different customer attributes and online shopping steps, so as to recommend appropriate products to customers in a timely manner and optimize e-commerce The placement of products on the website can effectively increase the transaction opportunities of customers, turn website visitors into buyers, improve cross-selling capabilities, increase customer loyalty, and improve the service quality and economic benefits of the website.

Personalized product recommendation for e-commerce does not require a lot of cost. It only needs to adjust the content of the website according to the characteristics of each customer, and make more personalized product recommendation pages according to the consumption preferences of each customer. Provide customers with more choices. This is equivalent to establishing an online store for each customer on the network, recommending targeted products to each customer, and helping customers choose products that really suit their needs from a huge catalog.

At present, most people use association rule analysis for personalized product recommendation, and few people use sequential pattern analysis to solve this problem. The problem to be solved by association rule analysis is to find out which commodities customers like to buy together, and which commodities customers will buy after purchasing certain commodities in one transaction. It is the process of discovering the internal laws of transactions, so that analysts can Arrange the placement of products according to the purchasing interests of customers to increase the transaction volume. The problem to be solved by the sequential pattern analysis is that after a customer completes a transaction, what products will he buy in a certain period of time in the future? To predict the future trading situation of commodities, so as to better arrange the placement of commodities. Its main purpose is to study the sequence of commodity purchases and find out the rules, that is, it is not only necessary to know whether the commodity is purchased, but also to determine the order in which the commodity is purchased with other commodities. For example, customers who have ordered commodity A online, 40% of people will order item B within 2 months. Sequence mode can find a frequent sequence in a certain period of time in the database, that is, which products will be purchased by customers more during this period of time, and the standard of more or less is determined by the minimum support. Each sequence is a set of collections arranged according to the time of the transaction, and the minimum support can be set to mine sequences that meet different frequency levels. However, when applying sequential patterns to analyze customer purchase behavior and solve the problem of personalized product recommendation, they only consider the events that have occurred, which is also called positive sequential pattern (Positive Sequential Pattern, PSP) mining.

Compared with positive sequential pattern mining, negative sequential pattern (Negative Sequential Pattern, NSP) mining also considers non-occurring events, which provides a new perspective for data analysis and enables deeper analysis and understanding of potential meanings in data. For example: a stands for bread, b stands for coffee, c stands for tea, d stands for sugar, Represents a customer purchase sequence pattern, which indicates that within a certain period of time, the customer purchased product d after purchasing product a and b without purchasing product c. Such negative sequence patterns help to more comprehensively obtain the hidden rules and patterns in the data, and play an irreplaceable role in many application fields such as customer behavior analysis, medical insurance fraud detection, and the relationship between missing genes and diseases. However, after mining the negative sequence patterns, we found it difficult to select feasible patterns that can be used for decision-making. Not all sequence patterns can be used in the analysis of customer purchase behavior, and after mining the negative sequence patterns, we found the previously mined positive sequences Patterns can mislead decisions, and decision makers don't know it yet. For example, when only positive sequence patterns are mined, if <abc> is obtained, decision makers can use it to make decisions. But now based on <abc>, the negative sequence patterns we can get are and Obviously, not all patterns can be used for decision making. If the selected decision-making mode does not contain <abc>, it means that the original decision-making using <abc> is misleading, and the decision-maker does not know it. Therefore, how to choose a feasible positive and negative sequence pattern that can be used for decision-making is the most urgent problem that needs to be solved after mining the negative sequence pattern.

Although there are many papers discussing feasible knowledge discovery and selecting feasible patterns/rules or methods of interest in association rule mining, so far no papers have been found on how to select feasible positive and negative sequential patterns. This may be because it is difficult to find this problem before mining negative sequence patterns, and there are very few papers on negative sequence pattern mining, most of which focus on how to design a mining algorithm and how to improve the efficiency of the algorithm.

The data source for mining is the purchase order data of website users in the e-commerce platform.

Taking the transactions of 5 customers within 3 months as an example, Table 1 is the customer purchase sequence database that has been sorted out. The letters represent the product ID.

Table 1 is the customer purchase sequence database that has been sorted out

customer ID customer purchase sequence 10 <a(abc)(ac)d(cf)> 20 <(ad)c(bc)(ae)> 30 <(ef)(ab)(df)cb> 40 <eg(af)cbc> 50 <(de)>

All transaction records of a customer within a certain period of time form an ordered sequence, and the sequence is represented by <>. In the sequence, the items/items are in order, each item represents a product of the transaction, and the element refers to all the products purchased by the customer at a specific point in time, using {} or ( ) indicates that the customer may purchase the same item in different time periods, that is, an item may occur in different elements of a sequence. For example, the purchase sequence of the customer whose ID is 20 in Table 1 is <(ad)c(bc)(ae)>, the customer purchased commodity a in the first and fourth shopping, where (ad), c, The four item sets (bc), (ae) can be called the elements of the sequence, and the commodities a, b, c, d, and e are called items. If there is only one item in an element, the brackets can be omitted, such as the sequence Element c in .

Contents of the invention

Aiming at the deficiencies of the prior art, the present invention provides the application of the positive and negative sequential pattern screening method in the analysis of customer purchase behavior. In the present invention, an efficient algorithm named SAP is proposed to select feasible positive and negative sequence patterns. The main idea of the algorithm is to dig out all positive sequence patterns and negative sequence patterns through the e-NSP method, and then for each sequence Modes are screened to select feasible positive and negative sequence modes. The sequence pattern obtained after screening by this algorithm is used to analyze the customer's purchase behavior, so that the seller can predict the future commodity sales situation according to the current commodity sales situation, so as to better arrange the placement of commodities and improve commodity sales quantity.

Technical scheme of the present invention is as follows:

The application of the positive and negative sequential pattern screening method in the analysis of customer purchase behavior includes the following steps:

(1) Use the correlation coefficient function to measure the relationship between commodities and commodities:

A product purchased by a customer each time is a separate item, and all the products purchased by the customer at one time at a specific point in time are an element, and the sequence includes all items corresponding to a customer's purchase of products within a certain period of time. element; ρ represents the correlation coefficient between any two elements in a sequence:

If ρ>0, then the above two elements are positively correlated; the more goods purchased at one time, the more goods purchased at the other time;

If ρ=0, then the above two elements have no correlation; the purchase behavior of the commodity purchased once and the commodity purchased in another purchase are independent of each other, and the two purchase behaviors do not affect each other;

If ρ<0, then the above two elements are negatively correlated; the more goods purchased at one time, the less goods purchased at the other time;

The range of ρ is between -1 and 1, the smaller the absolute value of ρ, the smaller the correlation between these two elements; set the threshold ρ _min , that is, only select the sequence corresponding to ρ≥ρ _min , that is,

When the correlation coefficients of any two elements in a sequence are ≥ ρ _min , the sequence is selected; otherwise, the sequence is excluded;

(2) Select a feasible sequence mode:

The way to judge whether a sequence is feasible is to test whether any subsequence of size 2 is feasible, and these subsequences of size 2 are composed of any adjacent 2 elements in a sequence, that is, if a size is k, where k>1, the positive sequence pattern or negative sequence pattern P=<e ₁ e ₂ …e _k > is feasible, then requires <e ₁ e ₂ >,<e ₂ e ₃ >,…,< e _k-1 e _k > is also feasible, defined as follows:

Definition 1. Feasible sequential patterns

A positive sequence pattern or negative sequence pattern P=<e ₁ e ₂ …e _k > with size k (k>1) is feasible if $\&ForAll;$ $i\&Element;\{2...k\},$

asp(e _i-1 ,e _i )=s(<e _i-1 e _i >)≥ms∧

(ii)

(f(e _i-1 ,e _i ,ms,ρ _min )=1),

in

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; ms</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; ms</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; ms</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; iii</span>}<span\; class="notranslate">\; )</span>$

$\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span>}{\sqrt{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}}<span\; class="notranslate">\; .</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; iv</span>}<span\; class="notranslate">\; )</span>$

f() is a constraint function related to support and correlation; ms is the minimum support threshold set by the user, which is used to cut off infrequent sequences; where s() represents the support of the sequence, s()≠0 or 1;

According to definition 1, a positive sequence pattern or negative sequence pattern P=<e ₁ e ₂ …e _k > with a size of k (k>1), if <e _i-1 e _i > is not a feasible sequence pattern, then the sequence pattern P is not feasible;

From definition 1, if P=<e ₁ e ₂ …e _k > is feasible, then <e _i-1 e _i > must be a feasible sequence pattern, otherwise P=<e ₁ e ₂ …e _k > will not be a feasible sequence pattern;

(3) The steps of using the SAP algorithm are as follows:

In the first step, all positive sequence patterns and negative sequence patterns are mined with the e-NSP method;

The second step, for each sequence pattern with a size of 2, use the above definition 1 to test the sequence pattern with a size of 2, if the sequence pattern with a size of 2 is not a feasible sequence pattern, delete this pattern and all the sequence pattern containing the pattern;

In the third step, for a sequence pattern whose size exceeds 2, combine it into a sequence according to every adjacent 2 elements, split it into multiple sequences with a size of 2, and use the method described in the second step for each size of 2 sequences, if each sequence with a size of 2 is a feasible sequence, then the sequence pattern with a size greater than 2 is a feasible sequence pattern;

The fourth step is to obtain all feasible sequence patterns according to the third step; analyze the customer's purchase behavior through these feasible sequence patterns;

(4) The pseudo code of the SAP algorithm is as follows:

Algorithm SAP

Input: D: customer purchase sequence database; ms: minimum support; ρ _min : correlation threshold;

Output: ASP: A collection of sequential patterns that can be used to analyze customer purchase behavior;

(1) Let ASP=Φ;

(2) Use the e-NSP method to mine all the positive sequence patterns PSPs and negative sequence patterns NSPs, and then store them in the set {PNSP};

(3) for k from 2 to the maximum length of PSP in{PNSP}do{

(4) for each pattern P=<e ₁ e ₂ ... e _k >in{PNSP}do{

(5) Test pattern P with definition 1;

(6) If P is a feasible sequence pattern, then

(7) adding the sequence pattern P to the set ASP;

(8) otherwise

(9) Delete pattern P and all patterns containing P from {PNSP};

(10)}

(11) k++;

(12)}

(13) Return to ASP; analyze the customer's purchase behavior through the returned result.

The advantages of the present invention are:

The present invention is applied in the process of analyzing the customer's purchase behavior, not only using the minimum support to screen out the products that the customer purchases more in a certain period of time, but also using the correlation coefficient to screen out the goods that have a relatively high correlation with the customer's purchase during this period of time In this way, when a customer purchases a product, the present invention can be used to recommend to him some products that other customers will buy and have a relatively high correlation with this product, thereby increasing the transaction opportunities of the customer and converting the website browser into a buyer. Improve cross-selling capabilities, increase customer loyalty, and improve website profitability.

specific embodiment

The present invention will be described in detail below in conjunction with the examples, but not limited thereto.

Embodiment 1,

The application of the positive and negative sequential pattern screening method in the analysis of customer purchase behavior includes the following steps:

(1) Use the correlation coefficient function to measure the relationship between commodities and commodities:

A product purchased by a customer each time is a separate item, and all the products purchased by the customer at one time at a specific point in time are an element, and the sequence includes all items corresponding to a customer's purchase of products within a certain period of time. element; ρ represents the correlation coefficient between any two elements in a sequence:

If ρ>0, then the above two elements are positively correlated; the more goods purchased at one time, the more goods purchased at the other time;

If ρ=0, then the above two elements have no correlation; the purchase behavior of the commodity purchased once and the commodity purchased in another purchase are independent of each other, and the two purchase behaviors do not affect each other;

If ρ<0, then the above two elements are negatively correlated; the more goods purchased at one time, the less goods purchased at the other time;

The range of ρ is between -1 and 1, the smaller the absolute value of ρ, the smaller the correlation between these two elements; set the threshold ρ _min , that is, only select the sequence corresponding to ρ≥ρ _min , that is,

When the correlation coefficients of any two elements in a sequence are ≥ ρ _min , the sequence is selected; otherwise, the sequence is excluded;

(2) Select a feasible sequence mode:

The way to judge whether a sequence is feasible is to test whether any subsequence of size 2 is feasible, and these subsequences of size 2 are composed of any adjacent 2 elements in a sequence, that is, if a size is k, where k>1, the positive sequence pattern or negative sequence pattern P=<e ₁ e ₂ …e _k > is feasible, then requires <e ₁ e ₂ >,<e ₂ e ₃ >,…,< e _k-1 e _k > is also feasible, defined as follows:

Definition 1. Feasible sequential patterns

A positive sequence pattern or negative sequence pattern P=<e ₁ e ₂ …e _k > with size k (k>1) is feasible if $\&ForAll;$ $i\&Element;\{2...k\},$

asp(e _i-1 ,e _i )=s(<e _i-1 e _i >)≥ms∧

(ii)

(f(e _i-1 ,e _i ,ms,ρ _min )=1),

in

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; ms</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; ms</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; ms</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; iii</span>}<span\; class="notranslate">\; )</span>$

$\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span>}{\sqrt{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}}<span\; class="notranslate">\; .</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; iv</span>}<span\; class="notranslate">\; )</span>$

f() is a constraint function related to support and correlation; ms is the minimum support threshold set by the user, which is used to cut off infrequent sequences; where s() represents the support of the sequence, s()≠0 or 1;

According to definition 1, a positive sequence pattern or negative sequence pattern P=<e ₁ e ₂ …e _k > with a size of k (k>1), if <e _i-1 e _i > is not a feasible sequence pattern, then the sequence pattern P is not feasible;

From definition 1, if P=<e ₁ e ₂ …e _k > is feasible, then <e _i-1 e _i > must be a feasible sequence pattern, otherwise P=<e ₁ e ₂ …e _k > will not be a feasible sequence pattern;

(3) The steps of using the SAP algorithm are as follows:

In the first step, all positive sequence patterns and negative sequence patterns are mined with the e-NSP method;

The second step, for each sequence pattern with a size of 2, use the above definition 1 to test the sequence pattern with a size of 2, if the sequence pattern with a size of 2 is not a feasible sequence pattern, delete this pattern and all the sequence pattern containing the pattern;

In the third step, for a sequence pattern whose size exceeds 2, combine it into a sequence according to every adjacent 2 elements, split it into multiple sequences with a size of 2, and use the method described in the second step for each size of 2 sequences, if each sequence with a size of 2 is a feasible sequence, then the sequence pattern with a size greater than 2 is a feasible sequence pattern;

The fourth step is to obtain all feasible sequence patterns according to the third step; analyze the customer's purchase behavior through these feasible sequence patterns;

(4) The pseudo code of the SAP algorithm is as follows:

Algorithm SAP

Input: D: customer purchase sequence database; ms: minimum support; ρ _min : correlation threshold;

Output: ASP: A collection of sequential patterns that can be used to analyze customer purchase behavior;

(1) Let ASP=Φ;

(2) Use the e-NSP method to mine all the positive sequence patterns PSPs and negative sequence patterns NSPs, and then store them in the set {PNSP};

(3) for k from 2 to the maximum length of PSP in{PNSP}do{

(4) for each pattern P=<e ₁ e ₂ ... e _k >in{PNSP}do{

(5) Test pattern P with definition 1;

(6) If P is a feasible sequence pattern, then

(7) adding the sequence pattern P to the set ASP;

(8) otherwise

(9) Delete pattern P and all patterns containing P from {PNSP};

(10)}

(11) k++;

(12)}

(13) Return to ASP; analyze the customer's purchase behavior through the returned result.

In the analysis of customer purchase behavior, the focus of sequence pattern analysis is to analyze the before and after or causal relationship between data. It is to find those internal transaction patterns of "some items follow other items" in the time-ordered transaction set. Example: A customer who bought a Pentium PC 9 months ago is likely to order a new CPU chip within a month. For another example, a customer who bought a PC may then buy a memory chip and then a CD-ROM.

By digging out a lot of products that customers will buy within a period of time and have a relatively high correlation from customer purchase records, it is convenient for e-commerce organizers to predict customer behavior and provide personalized services to customers, and find out what products will be in other products Purchase after purchase, so that you can recommend to customers some products that other customers will buy and are relatively related to this product, and these products can be placed in the most prominent position. For example, when a customer buys a personal computer online, the system may suggest that he consider buying other things at the same time based on the previously mined sequence patterns, such as "people who buy this kind of personal computer are likely to buy it again within three months." Buy a particular printer or CD-ROM" can give users a short-term coupon to boost product sales. And the negative items in the negative sequence pattern, that is, the goods that customers do not buy, we do not need to recommend to customers, for example, <smartphone, Camera, memory card > this sequence mode, when customers buy a smart phone, the system will recommend to the customer to buy a memory card instead of a camera, because people who buy a smart phone are likely to buy a memory card within three months instead of a camera. Buy a camera.

By analyzing customer purchase behavior and discovering the relationship between transactions, it can not only predict the future commodity sales situation according to the current commodity sales situation, but also better arrange the placement of commodities, thereby increasing commodity sales.

Claims (1)

1. The application of the positive and negative sequence pattern screening method in the analysis of customer purchase behavior, is characterized in that, comprising the following steps: (1) Use the correlation coefficient function to measure the relationship between commodities and commodities: A product purchased by a customer each time is a separate item, and all the products purchased by the customer at one time at a specific point in time are an element, and the sequence includes all items corresponding to a customer's purchase of products within a certain period of time. element; ρ represents the correlation coefficient between any two elements in a sequence: If ρ>0, then the above two elements are positively correlated; the more goods purchased at one time, the more goods purchased at the other time; If ρ=0, then the above two elements have no correlation; the purchase behavior of the commodity purchased once and the commodity purchased in another purchase are independent of each other, and the two purchase behaviors do not affect each other; If ρ<0, then the above two elements are negatively correlated; the more goods purchased at one time, the less goods purchased at the other time; The range of ρ is between -1 and 1, the smaller the absolute value of ρ, the smaller the correlation between these two elements; set the threshold ρ _min , that is, only select the sequence corresponding to ρ≥ρ _min ; that is, When the correlation coefficients of any two elements in a sequence are ≥ ρ _min , the sequence is selected; otherwise, the sequence is excluded; (2) Select a feasible sequence mode: The way to judge whether a sequence is feasible is to test whether any subsequence of size 2 is feasible, and these subsequences of size 2 are composed of any adjacent 2 elements in a sequence, that is, if a size is k, where k>1, the positive sequence pattern or negative sequence pattern P=<e ₁ e ₂ …e _k > is feasible, then requires <e ₁ e ₂ >,<e ₂ e ₃ >,…,< e _k-1 e _k > is also feasible, defined as follows: Definition 1. Feasible sequential patterns A positive sequence pattern or negative sequence pattern P=<e ₁ e ₂ …e _k > with size k (k>1) is feasible if asp(e _i-1 ,e _i )=s(<e _i-1 e _i >)≥ms∧ (ii) (f(e _i-1 ,e _i ,ms,ρ _min )=1), in $\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; ms</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; ms</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; ms</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; iii</span>}<span\; class="notranslate">\; )</span>$ $\mathrm{<span\; class="notranslate">\; \&rho;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span>}{\sqrt{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}}<span\; class="notranslate">\; .</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; iv</span>}<span\; class="notranslate">\; )</span>$ f() is a constraint function related to support and correlation; ms is the minimum support threshold set by the user, which is used to cut off infrequent sequences; where s() represents the support of the sequence, s()≠0 or 1; According to definition 1, a positive sequence pattern or negative sequence pattern P=<e ₁ e ₂ …e _k > with a size of k (k>1), if <e _i-1 e _i > is not a feasible sequence pattern, then the sequence pattern P is not feasible; From definition 1, if P=<e ₁ e ₂ …e _k > is feasible, then <e _i-1 e _i > must be a feasible sequence pattern, otherwise P=<e ₁ e ₂ …e _k > will not be a feasible sequence pattern; (3) The steps of using the SAP algorithm are as follows: In the first step, all positive sequence patterns and negative sequence patterns are mined with the e-NSP method; In the second step, for each sequence pattern whose size is 2, use the above definition 1 to test the sequence pattern with a size of 2, if the sequence pattern with a size of 2 is not a feasible sequence pattern, then delete this pattern and all the sequence pattern containing the pattern; In the third step, for a sequence pattern whose size exceeds 2, combine it into a sequence according to every adjacent 2 elements, split it into multiple sequences with a size of 2, and use the method described in the second step for each size of 2 sequences, if each sequence with a size of 2 is a feasible sequence, then the sequence pattern with a size greater than 2 is a feasible sequence pattern; The fourth step is to obtain all feasible sequence patterns according to the third step; (4) The pseudo code of the SAP algorithm is as follows: Algorithm SAP Input: D: customer purchase sequence database; ms: minimum support; ρ _min : correlation threshold; Output: ASP: A collection of sequential patterns that can be used to analyze customer purchase behavior;