Commodity attention tracking system and method based on user behavior analysis

Abstract

The invention discloses a commodity attention tracking system and method based on user behavior analysis, and relates to the technical field of computer data analysis. Taking a time ratio of browsing different commodity attribute introduction contents of the same commodity by the network account as a browsing feature of the network account, calculating average residence time of a sharing link in the browsing record sequence, and obtaining a historical record of sharing the commodity by the target user, the method comprises the following steps: collecting network accounts of a sharing link sent by a target user in a historical record to introduction content, obtaining all browsing record sequences of a target contact, and evaluating various browsing characteristics of the target contact according to the difference of the browsing characteristics and the average residence time; and matching the contact persons according to the proportion of the introduction content of each commodity attribute in the introduction content of the to-be-shared commodity, and sorting the matched contact persons through the average residence time to obtain a recommendation sharing sequence.

Description

Technical Field

[0001] The present invention relates to the technical field of computer data analysis, and in particular to a commodity attention tracking system and method based on user behavior analysis. Background Art

[0002] With the development of Internet technology, more and more product recommendation services are being conducted online, and related business tasks promote products by sending product links. Different network users receive product link information by viewing notification prompts in application software, and product links are also easier to share among network users, that is, product links can be spread among other network users without being managed by the initial sharer. Such technology greatly improves the promotion efficiency of product information.

[0003] However, at the same time, since the initial sharer is unaware of the link sharing, it is difficult for the initial sharer to evaluate the link sharing process, which is not conducive to the initial sharer's evaluation of the browser's acceptance of the product. It is also difficult for the initial sharer to judge the pertinence of information sharing. Summary of the invention

[0004] The purpose of the present invention is to provide a commodity attention tracking system and method based on user behavior analysis to solve the problems raised in the prior art.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for tracking commodity attention based on user behavior analysis, the method comprising:

[0006] Step S100: Classify and collect the introduction contents of the products according to the product attributes, collect the introduction contents corresponding to several products and merchants to establish a product database, and set the sharing link corresponding to each product in the product database;

[0007] Step S200: a user in the product database is taken as a target user. The target user sends a sharing link to a certain network account through the network. The certain network account sends the sharing link to another network account. The propagation order of the same sharing link in several network accounts is obtained. The account that places an order for the product is taken as the last network account. The network accounts are arranged into a browsing record sequence according to the propagation order.

[0008] Step S300: Obtain browsing records of each network account in the browsing record sequence, record the time taken by the network account to browse the introduction content of different product attributes, take the time proportion of the network account browsing the introduction content of different product attributes of the same product as the browsing feature of the network account, and compare the difference in browsing features between the last network account in the target sequence and the target sequence;

[0009] Step S400: collecting the propagation time of the shared links in the browsing record sequence, obtaining the length of the browsing record sequence, and calculating the average retention time of the shared links in each browsing record sequence;

[0010] Step S500: Obtain the target user's history of sharing products, collect the network accounts of the target user who sent the sharing link in the history into the introduction content, take a contact in the target user's sharing contact set as the target contact, obtain all browsing record sequences of the target contact, classify the browsing record sequences according to browsing features, and evaluate various browsing features of the target contact through differences in browsing features and average retention time;

[0011] Step S600: The target user selects a product from the product database as the product to be shared, matches contacts according to the proportion of each product attribute introduction content in the introduction content of the product to be shared, and sorts the matched contacts by average residence time to obtain a sequence of recommended sharing.

[0012] Further, step S200 includes:

[0013] Step S201: taking a product in the product database as a target product and obtaining the introduction content of the target product;

[0014] Step S202: taking a certain sharing link shared by the target user as the target link, and sharing the target link from one account to another account is recorded as one transmission;

[0015] Step S203: Arrange the network accounts that view the target product according to the propagation order of the target link, record the account that purchases the target product as the last network account, and form a target sequence with all network accounts from the first account to the last network account shared by the target user.

[0016] Furthermore, step S300 includes:

[0017] Step S301: obtain the i-th account Ui in the target sequence, obtain the total time ta that the account Ui browses the introduction content of the target product in the target sequence, obtain the k-th introduction content in the introduction content of the target product, record the k-th product attribute as qk, obtain the time that the user Ui browses the k-th introduction content as tk, and calculate the characteristic value αik of the introduction content corresponding to the account Ui browsing the k-th product attribute, αik=tk / ta;

[0018] Step S302: Combining the commodity attributes with the corresponding feature values, obtaining the feature value corresponding to the k-th value to obtain the evaluation group (qk, αik), respectively obtaining the feature values ​​of all commodity attributes read by account Ui, and collecting the evaluation groups corresponding to each feature value to obtain the user reading record list of account Ui;

[0019] Step S303: collect the feature values ​​corresponding to the same product attribute of the target product, calculate the average value of the feature value of the same product attribute, import each product attribute and the corresponding average value into a reference state list, obtain the product attributes in the reference state list and import them into the reference attribute set R1;

[0020] Step S304: Obtain the user reading record list of the last account in the target sequence, collect the product attributes in the user reading record list into the target attribute set R2, obtain the product attributes that are the same as R2 from R1 as the comparison attributes, obtain the average values ​​corresponding to the comparison attributes from the reference status list, collect the comparison attributes and the corresponding average values ​​into the comparison set, and the arrangement order of the product attributes in the comparison set is the same as the arrangement order of the product attributes in the target attribute set R2;

[0021] Step S305: Calculate the difference value w between the target attribute set R2 and the comparison set. Among them, q 1 p1 represents the average value of the attributes of the product in p1 in the comparison set, q 2 p1 The feature value corresponding to the p1th product attribute in the target attribute set R2, where p2 represents the number of product attributes in the target attribute set R2;

[0022] In the process of the same sharing link spreading among network accounts, the network accounts are sorted. In a sequence, the first network account is the network account that directly receives the sharing link sent by the target user, and the last network account is the network account that purchases the product corresponding to the link.

[0023] The purpose of sharing links is to find suitable buyers. When a transaction is completed, the browsing record sequence is marked according to the attributes of the product introduction information browsed by the buyer. The mark corresponding to the browsing record sequence is the product attribute browsed by the last network account and the corresponding time percentage;

[0024] In the browsing record sequence, each network account browses the product. By comparing the product attributes and corresponding time proportion browsed by the last network account with the average of the product attributes and corresponding time proportion browsed by the network accounts in the sequence, the difference between the demand of the last network account and the average of the demand in the sequence is obtained.

[0025] When the difference is higher, it means that the sharing process is less targeted. When the difference is lower, it means that the sharing process is more targeted. The targeting of a browsing record sequence is measured by product attributes, corresponding time proportions and difference values.

[0026] Furthermore, step S400 includes:

[0027] Step S401: the length of time from when a certain account receives a target link to when the target link is shared is recorded as retention time, and the total retention time in the target sequence is calculated, and the total retention time is recorded as T;

[0028] Step S402: Obtain user reading record lists of all accounts in the target sequence, count the number of user reading record lists that are the same as the user reading record list of account Ui, and record the number as M;

[0029] Step S403: Calculate the retention index γ of the target sequence, γ=N×lg(M×T), where N represents the number of target link propagations in the target sequence, and lg represents a logarithmic function with a base of 10;

[0030] The browsing characteristics of the last online account in the browsing sequence are used as the reference standard. The number of times such characteristics appear in the browsing record sequence is retrieved. The more times such characteristics appear, the more mediocre the attribute characteristics of the product are. Only after online accounts with the same needs browse many times will someone be willing to place an order. When the number of times such characteristics appear is small, it means that the product is attractive to customers with certain needs.

[0031] N and T measure the distance of the browsing record sequence from two different perspectives. N represents the level of the browsing record sequence, and T represents the time span. Therefore, in terms of product sharing and transaction efficiency, the smaller the retention index γ value, the stronger the sharing link propagation efficiency.

[0032] Furthermore, step S500 includes:

[0033] Step S501: Acquire all transaction record sequences of the target contact, take the user reading record list of the last account in the target sequence as the target feature, acquire the transaction record sequences whose user reading record lists of the last account in all browsing record sequences are the same as the target feature, and aggregate the transaction record sequences into a feature record set;

[0034] Step S502: obtaining the difference value and retention index of each browsing record sequence in the feature record set, recording the average value of the difference value as the first evaluation coefficient, and recording the average value of the retention index as the second evaluation coefficient;

[0035] Step S503: taking the user reading record list of the last account of the target sequence as the feature label of the target sequence, and collecting the feature label, the first evaluation coefficient and the second evaluation coefficient to form an evaluation group;

[0036] Step S504: Gather the evaluation groups corresponding to all feature tags of the target contact to obtain an evaluation database of the target contact.

[0037] Furthermore, step S600 includes:

[0038] Step S601: Obtain the product introduction content of the product to be shared, group all the feature words in the product introduction content into a set D0, collect the feature words according to the product attributes, and record the set composed of the feature words corresponding to the e-th product attribute of the product to be shared as De;

[0039] Step S602: Calculate the attribute ratio αe of the e-th commodity attribute, αe=num(De) / num(D0), where num is a function for obtaining the number of set feature words;

[0040] Step S603: Collect the attributes of each product to be shared and the corresponding attribute ratios to form an attribute label;

[0041] Step S604: obtaining evaluation groups corresponding to the feature tags that are the same as the attribute tags, and merging all evaluation groups into a first set to be shared;

[0042] Step S605: setting a judgment threshold F, merging the evaluation groups whose first evaluation coefficients are less than F in the first set to be shared into the second set to be shared, and sorting the second set to be shared in ascending order of the second evaluation coefficients to obtain a third set to be shared;

[0043] Step S606: sequentially obtain the accounts of each evaluation group in the third sharing set to form a contact sequence, and push the contact sequence to the target user.

[0044] In order to better implement the above method, a product attention tracking system based on user behavior analysis is also proposed, and the system includes:

[0045] A product management module, a browsing record sequence management module, a difference comparison module, a retention time management module, a feature evaluation module and a recommendation sorting module, wherein the product management module is used to manage product information and the corresponding relationship between the sharing link of the product and the product, the browsing record sequence management module is used to obtain the propagation record of the sharing link, and organize the network accounts into a browsing record sequence according to the propagation order of the sharing link, the difference comparison module is used to compare the difference between the browsing features of the last network account in the target sequence and the target sequence, the retention time management module is used to calculate the average retention time of the sharing link in each browsing record sequence, the feature evaluation module is used to evaluate the browsing record sequence corresponding to the target contact, and the recommendation sorting module is used to analyze the introduction content of the product, sort the matched contacts, and obtain a recommended sharing sequence;

[0046] Furthermore, the browsing record sequence management module includes: an introduction content management unit, a propagation record management unit and a browsing record sequence management unit, wherein the introduction content management unit is used to manage the introduction content of the product, the propagation record management unit is used to manage the propagation record of the sharing link, and the browsing record sequence management unit is used to capture the transaction record of the product to obtain the browsing record sequence;

[0047] Furthermore, the difference comparison module includes: a characteristic value calculation unit, a user reading record management unit, a comparison management unit and a difference value calculation unit, wherein the characteristic value calculation unit is used to obtain the browsing time of the introduction content corresponding to each commodity attribute, and calculate the characteristic value of the introduction content of the corresponding category, the user reading record management unit is used to collect the evaluation groups corresponding to each characteristic value to obtain a user reading record list, the comparison management unit is used to select the average browsing ratio of the corresponding category from the target sequence according to the category of the introduction content browsed by the last account in the target sequence, and the difference value calculation unit is used to calculate the difference value between the target attribute set and the comparison set;

[0048] Furthermore, the retention time management module includes: a time collection unit, a repetition detection unit and a retention index calculation unit, wherein the time collection unit is used to collect the retention time in any two network accounts, the repetition detection unit is used to detect the number of repetitions of the user reading record list corresponding to the last network account in the target sequence, and the retention index calculation unit is used to calculate the retention index of the target sequence;

[0049] Further, the feature evaluation module includes: an evaluation coefficient management unit, an evaluation group management unit and an evaluation database management unit, wherein the evaluation coefficient management unit is used to collect the first evaluation coefficient and the second evaluation coefficient, the evaluation group management unit is used to collect the feature label, the first evaluation coefficient and the second evaluation coefficient to form an evaluation group, and the evaluation database management unit is used to manage the evaluation database of the target contact;

[0050] Furthermore, the recommendation sorting module includes an attribute analysis unit, a feature matching unit, a threshold judgment unit and a sorting push unit, wherein the attribute analysis unit is used to analyze the attribute ratio of various product attributes in the product introduction content, the feature matching unit is used to manage the first set to be shared, the threshold judgment unit is used to manage the second set to be shared, and the sorting push unit is used to manage the third set to be shared and push contacts in order to the target user.

[0051] Compared with the prior art, the beneficial effects of the present invention are: capturing the dissemination status of the shared link of the product, and measuring the dissemination effect of the shared link by measuring the reading of the product introduction content by the network account after receiving the shared link. Under the condition of avoiding the acquisition of personal privacy information of the network account in the corresponding dissemination process as much as possible, the dissemination effect of the product link is evaluated and analyzed, enriching the product promotion business data, and effectively feeding back the decision on product promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a structural schematic diagram of a commodity attention tracking system based on user behavior analysis of the present invention;

[0053] Figure 2 The figure is a flow chart of a method for tracking product attention based on user behavior analysis according to the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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 are within the scope of protection of the present invention.

[0055] Example: Figure 1 and Figure 2 As shown, the present invention provides a technical solution, a commodity attention tracking method based on user behavior analysis:

[0056] Step S100: Classify and collect the introduction contents of the products according to the product attributes, collect the introduction contents corresponding to several products and merchants to establish a product database, and set the sharing link corresponding to each product in the product database;

[0057] Step S200: a user in the product database is taken as a target user. The target user sends a sharing link to a certain network account through the network. The certain network account sends the sharing link to another network account. The propagation order of the same sharing link in several network accounts is obtained. The account that places an order for the product is taken as the last network account. The network accounts are arranged into a browsing record sequence according to the propagation order.

[0058] In the embodiment, the target user is a direct promoter or seller of the product, and the network account is a network user in the role of a distributor or customer who participates in the sales.

[0059] Wherein, step S200 includes:

[0060] Step S201: taking a product in the product database as a target product and obtaining the introduction content of the target product;

[0061] Step S202: taking a certain sharing link shared by the target user as the target link, and sharing the target link from one account to another account is recorded as one transmission;

[0062] Step S203: Arrange the network accounts that view the target product according to the propagation order of the target link, record the account that purchases the target product as the last network account, and form a target sequence with all network accounts from the first account to the last network account shared by the target user.

[0063] Step S300: Obtain browsing records of each network account in the browsing record sequence, record the time taken by the network account to browse the introduction content of different product attributes, take the time proportion of the network account browsing the introduction content of different product attributes of the same product as the browsing feature of the network account, and compare the difference in browsing features between the last network account in the target sequence and the target sequence;

[0064] Wherein, step S300 includes:

[0065] Step S301: obtain the i-th account Ui in the target sequence, obtain the total time ta that the account Ui browses the introduction content of the target product in the target sequence, obtain the k-th introduction content in the introduction content of the target product, record the k-th product attribute as qk, obtain the time that the user Ui browses the k-th introduction content as tk, and calculate the characteristic value αik of the introduction content corresponding to the account Ui browsing the k-th product attribute, αik=tk / ta;

[0066] Step S302: Combining the commodity attributes with the corresponding feature values, obtaining the feature value corresponding to the k-th value to obtain the evaluation group (qk, αik), respectively obtaining the feature values ​​of all commodity attributes read by account Ui, and collecting the evaluation groups corresponding to each feature value to obtain the user reading record list of account Ui;

[0067] Step S303: collect the feature values ​​corresponding to the same product attribute of the target product, calculate the average value of the feature value of the same product attribute, import each product attribute and the corresponding average value into a reference state list, obtain the product attributes in the reference state list and import them into the reference attribute set R1;

[0068] Step S304: Obtain the user reading record list of the last account in the target sequence, collect the product attributes in the user reading record list into the target attribute set R2, obtain the product attributes that are the same as R2 from R1 as the comparison attributes, obtain the average values ​​corresponding to the comparison attributes from the reference status list, collect the comparison attributes and the corresponding average values ​​into the comparison set, and the arrangement order of the product attributes in the comparison set is the same as the arrangement order of the product attributes in the target attribute set R2;

[0069] Step S305: Calculate the difference value w between the target attribute set R2 and the comparison set. Among them, q 1 p1 represents the average value of the attributes of the product in p1 in the comparison set, q 2 p1 The feature value corresponding to the p1th product attribute in the target attribute set R2, and p2 represents the number of product attributes in the target attribute set R2.

[0070] Step S400: collecting the propagation time of the shared links in the browsing record sequence, obtaining the length of the browsing record sequence, and calculating the average retention time of the shared links in each browsing record sequence;

[0071] Wherein, step S400 includes:

[0072] Step S401: the length of time from when a certain account receives a target link to when the target link is shared is recorded as retention time, and the total retention time in the target sequence is calculated, and the total retention time is recorded as T;

[0073] Step S402: Obtain user reading record lists of all accounts in the target sequence, count the number of user reading record lists that are the same as the user reading record list of account Ui, and record the number as M;

[0074] Step S403: Calculate the retention index γ of the target sequence, γ=N×lg(M×T), where N represents the number of target link propagations in the target sequence, and lg represents a logarithmic function with base 10.

[0075] Step S500: Obtain the target user's history of sharing products, collect the network accounts of the target user who sent the sharing link in the history into the introduction content, take a contact in the target user's sharing contact set as the target contact, obtain all browsing record sequences of the target contact, classify the browsing record sequences according to browsing features, and evaluate various browsing features of the target contact through differences in browsing features and average retention time;

[0076] Wherein, step S500 includes:

[0077] Step S501: Acquire all transaction record sequences of the target contact, take the user reading record list of the last account in the target sequence as the target feature, acquire the transaction record sequences whose user reading record lists of the last account in all browsing record sequences are the same as the target feature, and aggregate the transaction record sequences into a feature record set;

[0078] Step S502: obtaining the difference value and retention index of each browsing record sequence in the feature record set, recording the average value of the difference value as the first evaluation coefficient, and recording the average value of the retention index as the second evaluation coefficient;

[0079] Step S503: taking the user reading record list of the last account of the target sequence as the feature label of the target sequence, and collecting the feature label, the first evaluation coefficient and the second evaluation coefficient to form an evaluation group;

[0080] Step S504: Gather the evaluation groups corresponding to all feature tags of the target contact to obtain an evaluation database of the target contact.

[0081] Step S600: The target user selects a product from the product database as the product to be shared, matches contacts according to the proportion of each product attribute introduction content in the introduction content of the product to be shared, and sorts the matched contacts by average residence time to obtain a sequence of recommended sharing;

[0082] Step S601: Obtain the product introduction content of the product to be shared, group all the feature words in the product introduction content into a set D0, collect the feature words according to the product attributes, and record the set composed of the feature words corresponding to the e-th product attribute of the product to be shared as De;

[0083] Step S602: Calculate the attribute ratio αe of the e-th commodity attribute, αe=num(De) / num(D0), where num is a function for obtaining the number of set feature words;

[0084] Step S603: Collect the attributes of each product to be shared and the corresponding attribute ratios to form an attribute label;

[0085] Step S604: obtaining evaluation groups corresponding to the feature tags that are the same as the attribute tags, and merging all evaluation groups into a first set to be shared;

[0086] One contact person has multiple evaluation groups. Each evaluation group evaluates the sales management of the contact person when the product has a certain characteristic. By analyzing the introduction content of the newly added products, the contact person who meets the characteristics of the newly added products is screened out, and the contacts are sorted according to the communication efficiency corresponding to the product characteristics.

[0087] Step S605: setting a judgment threshold F, merging the evaluation groups whose first evaluation coefficients are less than F in the first set to be shared into the second set to be shared, and sorting the second set to be shared in ascending order of the second evaluation coefficients to obtain a third set to be shared;

[0088] Step S606: sequentially obtain the accounts of each evaluation group in the third sharing set to form a contact sequence, and push the contact sequence to the target user.

[0089] A product attention tracking system based on user behavior analysis, the system includes: product management module, browsing record sequence management module, difference comparison module, retention time management module, feature evaluation module and recommendation sorting module;

[0090] Among them, the product management module is used to manage product information and the corresponding relationship between the product sharing link and the product;

[0091] The browsing record sequence management module is used to obtain the propagation record of the sharing link, and organize the network accounts into a browsing record sequence according to the propagation order of the sharing link. The browsing record sequence management module includes: an introduction content management unit, a propagation record management unit and a browsing record sequence management unit. The introduction content management unit is used to manage the introduction content of the product, the propagation record management unit is used to manage the propagation record of the sharing link, and the browsing record sequence management unit is used to capture the transaction record of the product to obtain the browsing record sequence.

[0092] The difference comparison module is used to compare the difference between the browsing characteristics of the last network account in the target sequence and the target sequence, wherein the difference comparison module includes: a feature value calculation unit, a user reading record management unit, a comparison management unit and a difference value calculation unit, wherein the feature value calculation unit is used to obtain the browsing time of the introduction content corresponding to each commodity attribute, and calculate the feature value of the introduction content of the corresponding category, the user reading record management unit is used to collect the evaluation groups corresponding to each feature value to obtain the user reading record list, the comparison management unit is used to select the average browsing ratio of the corresponding category from the target sequence according to the category of the introduction content browsed by the last account in the target sequence, and the difference value calculation unit is used to calculate the difference value between the target attribute set and the comparison set;

[0093] The retention time management module is used to calculate the average retention time of the shared links in each browsing record sequence, and the feature evaluation module is used to evaluate the browsing record sequence corresponding to the target contact. The retention time management module includes: a time collection unit, a repetition detection unit and a retention index calculation unit. The time collection unit is used to collect the retention time in any two network accounts, the repetition detection unit is used to detect the number of repetitions of the user reading record list corresponding to the last network account in the target sequence, and the retention index calculation unit is used to calculate the retention index of the target sequence.

[0094] The feature evaluation module is used to evaluate the browsing record sequence corresponding to the target contact, wherein the feature evaluation module includes: an evaluation coefficient management unit, an evaluation group management unit and an evaluation database management unit, wherein the evaluation coefficient management unit is used to collect the first evaluation coefficient and the second evaluation coefficient, the evaluation group management unit is used to collect the feature label, the first evaluation coefficient and the second evaluation coefficient to form an evaluation group, and the evaluation database management unit is used to manage the evaluation database of the target contact;

[0095] Among them, the recommendation sorting module is used to analyze the introduction content of the product, sort the matched contacts, and obtain a sequence of recommended sharing. Among them, the recommendation sorting module includes an attribute analysis unit, a feature matching unit, a threshold judgment unit and a sorting push unit. Among them, the attribute analysis unit is used to analyze the attribute ratio of various product attributes in the product introduction content, the feature matching unit is used to manage the first set to be shared, the threshold judgment unit is used to manage the second set to be shared, and the sorting push unit is used to manage the third set to be shared, and push contacts in sequence to the target user.

[0096] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A commodity attention tracking method based on user behavior analysis, characterized in that: The method comprises the following steps: Step S100: Classify and collect the introduction contents of the products according to the product attributes, collect the introduction contents corresponding to several products and merchants to establish a product database, and set the sharing link corresponding to each product in the product database; Step S200: a user in the product database is taken as a target user. The target user sends a sharing link to a certain network account through the network. The certain network account sends the sharing link to another network account. The propagation order of the same sharing link in several network accounts is obtained. The account that places an order for the product is taken as the last network account. The network accounts are arranged into a browsing record sequence according to the propagation order. Step S300: Obtain browsing records of each network account in the browsing record sequence, record the time taken by the network account to browse the introduction content of different product attributes, take the time proportion of the network account browsing the introduction content of different product attributes of the same product as the browsing feature of the network account, and compare the difference in browsing features between the last network account in the target sequence and the target sequence; Step S400: collecting the propagation time of the shared links in the browsing record sequence, obtaining the length of the browsing record sequence, and calculating the average retention time of the shared links in each browsing record sequence; Step S500: Obtain the target user's history of sharing products, collect the network accounts of the target user who sent the sharing link in the history into the introduction content, take a contact in the target user's sharing contact set as the target contact, obtain all browsing record sequences of the target contact, classify the browsing record sequences according to browsing features, and evaluate various browsing features of the target contact through differences in browsing features and average retention time; Step S600: The target user selects a product from the product database as the product to be shared, matches contacts according to the proportion of each product attribute introduction content in the introduction content of the product to be shared, and sorts the matched contacts by average residence time to obtain a sequence of recommended sharing.

2. The method for tracking commodity attention based on user behavior analysis according to claim 1, characterized in that: Step S200 includes: Step S201: taking a product in the product database as a target product and obtaining the introduction content of the target product; Step S202: taking a certain sharing link shared by the target user as the target link, and sharing the target link from one account to another account is recorded as one transmission; Step S203: Arrange the network accounts that view the target product according to the propagation order of the target link, record the account that purchases the target product as the last network account, and form a target sequence with all network accounts from the first account shared by the target user to the last network account.

3. The method for tracking commodity attention based on user behavior analysis according to claim 2, characterized in that: Step S300 includes: Step S301: obtain the i-th account Ui in the target sequence, obtain the total time ta that the account Ui browses the introduction content of the target product in the target sequence, obtain the k-th introduction content in the introduction content of the target product, record the k-th product attribute as qk, obtain the time that the user Ui browses the k-th introduction content as tk, and calculate the characteristic value αik of the introduction content corresponding to the account Ui browsing the k-th product attribute, αik=tk / ta; Step S302: Combining the commodity attributes with the corresponding feature values, obtaining the feature values ​​corresponding to the k-th value to obtain the evaluation group (qk, αik), respectively obtaining the feature values ​​of all commodity attributes read by account Ui, and collecting the evaluation groups corresponding to the feature values ​​to obtain the user reading record list of account Ui; Step S303: collect the feature values ​​corresponding to the same product attribute of the target product, calculate the average value of the feature value of the same product attribute, import each product attribute and the corresponding average value into a reference state list, obtain the product attributes in the reference state list and import them into the reference attribute set R1; Step S304: obtaining a user reading record list of the last account in the target sequence, collecting the product attributes in the user reading record list into the target attribute set R2, obtaining the same product attributes as R2 from R1 as comparison attributes, obtaining the average values ​​corresponding to the comparison attributes from the reference status list, and collecting the comparison attributes and the corresponding average values ​​into the comparison set, wherein the arrangement order of the product attributes in the comparison set is the same as the arrangement order of the product attributes in the target attribute set R2; Step S305: Calculate the difference value w between the target attribute set R2 and the comparison set. Among them, q 1 p1 represents the average value of the attributes of the product in p1 in the comparison set, q 2 p1 The feature value corresponding to the p1th product attribute in the target attribute set R2, and p2 represents the number of product attributes in the target attribute set R2.

4. The method for tracking commodity attention based on user behavior analysis according to claim 3, characterized in that: Step S400 includes: Step S401: the length of time from when a certain account receives a target link to when the target link is shared is recorded as retention time, and the total retention time in the target sequence is calculated, and the total retention time is recorded as T; Step S402: Obtain user reading record lists of all accounts in the target sequence, count the number of user reading record lists that are the same as the user reading record list of account Ui, and record the number as M; Step S403: Calculate the retention index γ of the target sequence, γ=N×lg(M×T), where N represents the number of target link propagations in the target sequence, and lg represents a logarithmic function with base 10.

5. The method for tracking commodity attention based on user behavior analysis according to claim 4, characterized in that: Step S500 includes: Step S501: Acquire all transaction record sequences of the target contact, take the user reading record list of the last account in the target sequence as the target feature, acquire the transaction record sequences whose user reading record lists of the last account in all browsing record sequences are the same as the target feature, and aggregate the transaction record sequences into a feature record set; Step S502: obtaining the difference value and retention index of each browsing record sequence in the feature record set, recording the average value of the difference value as the first evaluation coefficient, and recording the average value of the retention index as the second evaluation coefficient; Step S503: taking the user reading record list of the last account of the target sequence as the feature label of the target sequence, and collecting the feature label, the first evaluation coefficient and the second evaluation coefficient to form an evaluation group; Step S504: Gather the evaluation groups corresponding to all feature tags of the target contact to obtain an evaluation database of the target contact.

6. The method for tracking commodity attention based on user behavior analysis according to claim 5, characterized in that: Step S600 includes: Step S601: Obtain the product introduction content of the product to be shared, group all the feature words in the product introduction content into a set D0, collect the feature words according to the product attributes, and record the set composed of the feature words corresponding to the e-th product attribute of the product to be shared as De; Step S602: Calculate the attribute ratio αe of the e-th commodity attribute, αe=num(De) / num(D0), where num is a function for obtaining the number of set feature words; Step S603: Collect the attributes of each product to be shared and the corresponding attribute ratios to form an attribute label; Step S604: obtaining evaluation groups corresponding to the feature tags that are the same as the attribute tags, and merging all evaluation groups into a first set to be shared; Step S605: setting a judgment threshold F, merging the evaluation groups whose first evaluation coefficients are less than F in the first set to be shared into the second set to be shared, and sorting the second set to be shared in ascending order of the second evaluation coefficients to obtain a third set to be shared; Step S606: sequentially obtain the accounts of each evaluation group in the third sharing set to form a contact sequence, and push the contact sequence to the target user.

7. A product attention tracking system based on user behavior analysis, used to execute the product attention tracking method based on user behavior analysis according to any one of claims 1 to 6, characterized in that: The system includes: The product management module, the browsing record sequence management module, the difference comparison module, the retention time management module, the feature evaluation module and the recommendation sorting module, wherein the product management module is used to manage the product information and the corresponding relationship between the product sharing link and the product, the browsing record sequence management module is used to obtain the propagation record of the sharing link, and organize the network accounts into a browsing record sequence according to the propagation order of the sharing link, the difference comparison module is used to compare the difference between the browsing features of the last network account in the target sequence and the target sequence, the retention time management module is used to calculate the average retention time of the sharing link in each browsing record sequence, the feature evaluation module is used to evaluate the browsing record sequence corresponding to the target contact, and the recommendation sorting module is used to analyze the introduction content of the product, sort the matched contacts, and obtain the recommended sharing sequence.

8. The commodity attention tracking system based on user behavior analysis according to claim 7 is characterized by: The browsing record sequence management module includes: an introduction content management unit, a propagation record management unit and a browsing record sequence management unit, wherein the introduction content management unit is used to manage the introduction content of the product, the propagation record management unit is used to manage the propagation record of the sharing link, and the browsing record sequence management unit is used to capture the transaction record of the product to obtain the browsing record sequence; The difference comparison module includes: a characteristic value calculation unit, a user reading record management unit, a comparison management unit and a difference value calculation unit, wherein the characteristic value calculation unit is used to obtain the browsing time of the introduction content corresponding to each commodity attribute and calculate the characteristic value of the introduction content of the corresponding category, the user reading record management unit is used to collect the evaluation groups corresponding to each characteristic value to obtain the user reading record list, the comparison management unit is used to select the average browsing ratio of the corresponding category from the target sequence according to the category of the introduction content browsed by the last account in the target sequence, and the difference value calculation unit is used to calculate the difference value between the target attribute set and the comparison set; The retention time management module includes: a time collection unit, a repetition detection unit and a retention index calculation unit, wherein the time collection unit is used to collect the retention time in any two network accounts, the repetition detection unit is used to detect the number of repetitions of the user reading record list corresponding to the last network account in the target sequence, and the retention index calculation unit is used to calculate the retention index of the target sequence.

9. The commodity attention tracking system based on user behavior analysis according to claim 7, characterized in that: The feature evaluation module includes: an evaluation coefficient management unit, an evaluation group management unit and an evaluation database management unit, wherein the evaluation coefficient management unit is used to collect the first evaluation coefficient and the second evaluation coefficient, the evaluation group management unit is used to collect the feature label, the first evaluation coefficient and the second evaluation coefficient to form an evaluation group, and the evaluation database management unit is used to manage the evaluation database of the target contact.

10. The commodity attention tracking system based on user behavior analysis according to claim 7, characterized in that: The recommendation sorting module includes an attribute analysis unit, a feature matching unit, a threshold judgment unit and a sorting push unit, wherein the attribute analysis unit is used to analyze the attribute ratio of various product attributes in the product introduction content, the feature matching unit is used to manage the first set to be shared, the threshold judgment unit is used to manage the second set to be shared, and the sorting push unit is used to manage the third set to be shared and push contacts in order to the target user.

Images