User Deep Interaction System and Method Based on Multi-Source Data Platform

Abstract

The present invention discloses a user in-depth interaction system and method based on a multi-source data platform, which relates to the technical field of user in-depth interaction. It includes obtaining the historical online interaction records and historical offline interaction records of users in a shopping mall, evaluating the interaction tendency of users towards the commodities in the shopping mall to obtain interaction tendency data; obtaining the interaction tendency data of each user in the shopping mall, analyzing the overall interaction tendency of each user towards the commodities in the shopping mall to obtain the target interaction tendency data of the shopping mall; obtaining the historical consumption records of users in the shopping mall, analyzing the consumption matching of commodities by users in the shopping mall to obtain the consumption matching data of users; obtaining the consumption matching data of each user in the shopping mall, analyzing the consumption matching degree between different commodities in the shopping mall, and combining the target interaction tendency data to optimize and adjust the interaction strategy of the shopping mall.

Description

Technical Field

[0001] The present invention relates to the technical field of in-depth user interaction, and specifically to an in-depth user interaction system and method based on a multi-source data platform. Background Art

[0002] A multi-source data platform integrates data from different channels and of different specifications. With the continuous development of technology, the sources and forms of users' information acquisition are diverse. Therefore, it has become increasingly common to use a multi-source data platform for in-depth user interaction. The specific advantages of using a multi-source data platform include, but are not limited to, the following: 1. Provide precise services. By integrating multi-source data such as social media and transaction records, data silos can be broken, and a full-dimensional view of user behavior and needs can be formed; 2. Optimize operation efficiency. Based on the multi-source data platform, various data can be obtained, and a prediction model can be used to predict the loss of users, and intervention strategies can be implemented in advance to optimize operations; 3. Enhance cross-channel collaboration. The interaction data of users on different platforms can be coordinated with each other to ensure the coherence of services.

[0003] There are numerous merchants in a large shopping mall. Every day, many users carry out various activities online and offline. Each time a user conducts an activity, an interaction record is generated. In reality, in most cases, these users' interaction records are isolated from each other. The large shopping mall cannot effectively analyze the interaction records of each user, lacks an understanding of users' consumption preferences, and cannot tap into users' consumption potential. This will not only lead to a decline in the revenue of the merchants in the mall, but also cause problems in the supply chains of the various merchants in the mall, bringing huge losses to the mall. Summary of the Invention

[0004] The purpose of the present invention is to provide an in-depth user interaction system and method based on a multi-source data platform to solve the problems raised in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An in-depth user interaction method based on a multi-source data platform, the method comprising:

[0006] Step S100: Obtain the historical online interaction records and historical offline interaction records of users in the shopping mall, evaluate the interaction tendency of users towards the goods in the shopping mall, and obtain interaction tendency data;

[0007] Step S200: Obtain the interaction tendency data of each user in the shopping mall, analyze the overall interaction tendency of each user towards the goods in the shopping mall, and obtain the target interaction tendency data of the shopping mall;

[0008] Step S300: Obtain the historical consumption records of the user in the mall, analyze the consumption collocation of the user's goods in the mall, and obtain the user's consumption collocation data;

[0009] Step S400: Obtain the consumption collocation data of each user in the mall, analyze the consumption collocation degree between different goods in the mall, and combine the target interaction tendency data to optimize and adjust the interaction strategy of the mall.

[0010] Further, step S100 includes:

[0011] Step S101: Obtain the historical online interaction records and historical offline interaction records of the user in the mall respectively, and obtain the total number S´ of the goods consumed by the user on the online platform in the mall sum , and obtain the total number S of the goods consumed by the user offline in the mall sum ;

[0012] Step S102: Calculate the characteristic proportion coefficient B = S sum / (S´ sum + S sum ), calculate the characteristic proportion coefficient B´ = S´ sum / (S´ sum + S sum );

[0013] Step S103: Obtain the commodity types of the goods with which the user interacts from the historical online interaction records and historical offline interaction records respectively, and calculate the interaction tendency values of the user for each commodity type. Among them, the interaction tendency value F of the user for the c-th commodity type c :

[0014] ,

[0015] Among them, A´ c represents the total number of the user's historical online interaction records when the commodity type of the interaction goods is the c-th commodity type; A c represents the total number of the user's historical offline interaction records when the commodity type of the interaction goods is the c-th commodity type; A´ sum represents the total number of the user's historical online interaction records; A sum represents the total number of the user's historical offline interaction records;

[0016] When the interaction tendency value F c is greater than the preset interaction tendency threshold, it is determined that the user has an interaction tendency for the goods of the c-th commodity type, and the c-th commodity type is recorded as the user's tendency commodity type;

[0017] Step S104: Obtain various types of user-preferred commodities and gather them to obtain the user's interaction preference data.

[0018] Further, step S200 includes:

[0019] Step S201: Obtain the interaction preference data of each user in the mall, and obtain several types of user-preferred commodities from the interaction preference data;

[0020] Step S202: Analyze each user's overall interaction preference for commodities of various types in the mall. Among them, when analyzing each user's overall interaction preference for commodities of the h-th type in the mall, the specific process is as follows:

[0021] Obtain the maximum value W of the total number of consumer commodities of each user in the mall max , and calculate the overall interaction preference value L of each user for commodities of the h-th type in the mall h :

[0022] ,

[0023] where z represents the total number of users who mark the h-th type of commodity as a preferred commodity type; W e i represents the total number of consumer commodities of the i-th user who marks the h-th type of commodity as a preferred commodity type in the mall;

[0024] Step S203: When the overall interaction preference value L h is greater than the preset overall interaction preference threshold, it is determined that each user has an overall interaction preference for commodities of the h-th type, mark the h-th type of commodity as the target preferred commodity type of the mall, obtain several target preferred commodity types in the mall and gather them to obtain the target interaction preference data.

[0025] Further, step S300 includes:

[0026] Step S301: Record the online and offline consumption behaviors of users in the mall to obtain the historical consumption records of users, and obtain the commodity types of several commodities consumed by users from the historical consumption records;

[0027] Step S302: Analyze the consumption compatibility between various commodity types of commodities during the consumption process of users in the mall. Among them, when analyzing the consumption compatibility between commodities of the e-th type and commodities of the g-th type during the consumption process of users in the mall, the specific process is as follows:

[0028] Obtain the total number Q of historical consumption records of users that only contain commodities of the e-th type of commodity esum Obtain the total number Q of the historical consumption records of the user's commodities that only contain the g-th commodity type g sum Obtain the total number Q of the historical consumption records of the user's commodities that contain both the e-th commodity type and the g-th commodity type e,g sum Obtain the total number Q of the historical consumption records of the user's commodities that do not contain the e-th commodity type and the g-th commodity type sum ;

[0029] Step S303: Respectively set the random variable X of the e-th commodity type and the random variable Y of the g-th commodity type, set the eigenvalue k, where k > 0, X = {k, 0}, Y = {k, 0}, and calculate the joint probability p(k, 0) = Q of the commodity of the e-th commodity type and the commodity of the g-th commodity type e sum / n, p(0, k) = Q g sum / n, p(k, k) = Q e,g sum / n, p(0, 0) = Q sum / n;

[0030] Calculate the marginal probability p(k) = (Q e sum +Q e,g sum ) / n, p(0) = (Q g sum +Q sum ) / n;

[0031] Calculate the marginal probability p(k) = (Q g sum +Q e,g sum ) / n, p(0) = (Q e sum +Q sum ) / n;

[0032] Step S304: Obtain the total number n of the user's various historical consumption records, and calculate the consumption matching value R between the commodity of the e-th commodity type and the commodity of the g-th commodity type e,g :

[0033] ,

[0034] where p(x, y) is the joint probability of the commodity of the e-th commodity type and the commodity of the g-th commodity type; p(x) is the marginal probability of the commodity of the e-th commodity type; p(y) is the marginal probability of the commodity of the g-th commodity type;

[0035] Step S305: When the consumption matching value R e,g is greater than a preset consumption matching threshold, it is determined that there is a consumption matching between the commodity of the e-th commodity type and the commodity of the g-th commodity type during the user's consumption process in the mall. The e-th commodity type and the g-th commodity type are recorded as a consumption matching group of the user. Each consumption matching group of the user is obtained and aggregated to obtain consumption matching data.

[0036] Further, step S400 includes:

[0037] Step S401: Obtain the consumption matching data of each user in the mall, and analyze the consumption matching degree between the commodities of each commodity type in the mall. Among them, when analyzing the consumption matching degree between the commodity of the v-th commodity type and the commodity of the u-th commodity type in the mall, the specific process is as follows:

[0038] When the v-th commodity type and the u-th commodity type are in the same consumption matching group in the consumption matching data of a certain user, mark the certain user, and obtain several marked users;

[0039] Step S402: Calculate the characteristic consumption matching value T between the commodity of the v-th commodity type and the commodity of the u-th commodity type in the mall v,u :

[0040] ,

[0041] where β is the total number of several marked users; R α v,u is the consumption matching value between the commodity of the e-th commodity type and the commodity of the u-th commodity type in the α-th marked user;

[0042] Step S403: When the characteristic consumption matching value T v,u is greater than a preset characteristic consumption matching threshold, it is determined that there is a consumption matching between the commodity of the v-th commodity type and the commodity of the u-th commodity type in the mall, and the v-th commodity type is recorded as the matching commodity type of the u-th commodity type;

[0043] Step S404: Obtain several commodity types with consumption matching in the mall, obtain the target interaction tendency data of the mall in the current period, and obtain several target tendency commodity types of the mall from the target interaction tendency data;

[0044] Step S405: Denote several items of goods of the target preference commodity types as hot commodities, obtain several items of goods of the matching commodity types of the target preference commodity types, and also denote them as hot commodities. Adjust the procurement of goods in the mall during the current period, increase the procurement volume of each hot commodity during the current period, and reduce the procurement volume of other goods;

[0045] Step S406: On the online platform and offline platform of the mall, respectively increase the number of recommendations of the hot commodities during the current period, and increase the frequency of the hot commodities participating in activities, and adjust the interaction strategy of the mall;

[0046] In the above steps, by analyzing the consumption matching degree between the goods of various commodity types in the mall, it is possible to understand which two commodity types of goods will be purchased by users simultaneously during the actual consumption process. Therefore, when conducting activity promotion, the promotion of goods can be more targeted, thereby increasing the probability and frequency of in-depth interaction of users, and further increasing the sales volume of goods.

[0047] In order to better implement the above method, a user in-depth interaction system based on a multi-source data platform is also proposed. The system includes an interaction preference data module, a target interaction preference data module, a consumption matching analysis module, and an interaction strategy optimization module;

[0048] The interaction preference data module is used to evaluate the interaction preference of users for the goods in the mall to obtain interaction preference data;

[0049] The target interaction preference data module is used to analyze the overall interaction preference of each user for the goods in the mall to obtain the target interaction preference data of the mall;

[0050] The consumption matching analysis module is used to analyze the consumption matching of the goods of users in the mall to obtain the consumption matching data of users;

[0051] The interaction strategy optimization module is used to analyze the consumption matching degree between different goods in the mall, and combine the target interaction preference data to optimize and adjust the interaction strategy of the mall.

[0052] Furthermore, the interaction preference data module includes an interaction preference value unit and an interaction preference data unit;

[0053] The interaction preference value unit is used to calculate the interaction preference value of users for each commodity type;

[0054] The interaction preference data unit is used to evaluate the interaction preference of users for the goods in the mall according to the interaction preference value to obtain interaction preference data.

[0055] Further, the target interaction tendency data module includes an overall interaction tendency value unit and a target interaction tendency data unit;

[0056] The overall interaction tendency value unit is used to calculate the overall interaction tendency values of each user for the goods of various commodity types in the mall;

[0057] The target interaction tendency data unit is used to analyze the overall interaction tendency of each user for the goods of various commodity types in the mall according to the overall interaction tendency values, and obtain the target interaction tendency data.

[0058] Further, the consumption matching analysis module includes a consumption matching value unit and a consumption matching analysis unit;

[0059] The consumption matching value unit is used to calculate the consumption matching values between the goods of various commodity types;

[0060] The consumption matching analysis unit is used to analyze the consumption matching between the goods of various commodity types during the user's consumption process in the mall according to the consumption matching values, and obtain the consumption matching data.

[0061] Further, the interaction strategy optimization module includes a consumption matching degree analysis unit and an interaction strategy optimization unit;

[0062] The consumption matching degree analysis unit is used to obtain the consumption matching data of each user in the mall, and analyze the consumption matching degree between different goods in the mall;

[0063] The interaction strategy optimization unit is used to optimize and adjust the interaction strategy of the mall according to the consumption matching degree between different goods in the mall and in combination with the target interaction tendency data.

[0064] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention realizes in-depth interaction with users in the mall. Considering the different interaction tendency degrees of different users for different commodity types of goods, starting from the actual situation, according to the user's historical online interaction records and historical offline interaction records in the mall, the interaction tendency of the user for the goods in the mall is evaluated, so as to obtain the commodity types of the goods that the user is more inclined to interact with. And considering the actual situation that different commodity types of goods will be purchased in a matching manner during the user's purchase process, the consumption matching degree between commodity types is analyzed, so as to optimize and adjust the interaction strategy of the mall, and tap the consumption potential of users. This will not only increase the income of the merchants in the mall, but also make the supply chains of the merchants in the mall more reasonable, and avoid the shortage or backlog of goods. Description of the Drawings

[0065] Figure 1It is the flowchart of the method for in-depth user interaction based on the multi-source data platform of the present invention;

[0066] Figure 2 It is the schematic diagram of the modules of the system for in-depth user interaction based on the multi-source data platform of the present invention. Detailed implementation manners

[0067] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0068] Embodiment: As Figure 1 - Figure 2 shown, the present invention provides a technical solution, a method for in-depth user interaction based on a multi-source data platform, and the method includes:

[0069] Step S100: Obtain the historical online interaction records and historical offline interaction records of the user in the mall, evaluate the interaction tendency of the user with respect to the commodities in the mall, and obtain interaction tendency data;

[0070] Among them, step S100 includes:

[0071] Step S101: Respectively obtain the historical online interaction records and historical offline interaction records of the user, and obtain the total number S´ of the commodities consumed by the user in the online platform of the mall sum , and obtain the total number S of the commodities consumed by the user offline in the mall sum ;

[0072] Step S102: Calculate the characteristic proportion coefficient B = S sum / (S´ sum + S sum ), and calculate the characteristic proportion coefficient B´ = S´ sum / (S´ sum + S sum );

[0073] Step S103: Respectively obtain the commodity types of the commodities with which the user interacts from the historical online interaction records and historical offline interaction records, and calculate the interaction tendency values of the user with respect to each commodity type. Among them, the interaction tendency value F of the user with respect to the c-th commodity type c :

[0074] ,

[0075] Among them, A´ cWhen the commodity type of the interactive commodity is the c-th commodity type, the total number of the user's historical online interaction records; A c When the commodity type of the interactive commodity is the c-th commodity type, the total number of the user's historical offline interaction records; A´ sum The total number of the user's historical online interaction records; A sum The total number of the user's historical offline interaction records;

[0076] For example, A´ sum is 100; A sum is 100; B is 0.6; B´ is 0.4; A´3 is 30; A3 is 40;

[0077] Calculate the interaction tendency value F2 of the user for the 2nd commodity type:

[0078] ,

[0079] When the interaction tendency value F c is greater than the preset interaction tendency threshold, it is determined that the user has an interaction tendency for the commodity of the c-th commodity type, and the c-th commodity type is recorded as the user's tendency commodity type;

[0080] For example, each commodity type includes washing powder, sanitary napkins, toothbrushes, etc.;

[0081] Step S104: Obtain each user's tendency commodity types and pool them to obtain the user's interaction tendency data;

[0082] Step S200: Obtain the interaction tendency data of each user in the mall, analyze the overall interaction tendency of each user for the commodities in the mall, and obtain the target interaction tendency data of the mall;

[0083] Among them, step S200 includes:

[0084] Step S201: Obtain the interaction tendency data of each user in the mall, and obtain several tendency commodity types of the user from the interaction tendency data;

[0085] Step S202: Analyze the overall interaction tendency of each user for the commodities of each commodity type in the mall. Among them, when analyzing the overall interaction tendency of each user for the commodities of the h-th commodity type in the mall, the specific process is:

[0086] Obtain the maximum value W of the total number of consumer commodities of each user in the mall max , and calculate the overall interaction tendency value L of each user for the commodities of the h-th commodity type in the mall h :

[0087] ,

[0088] Among them, z represents the total number of users for whom the h-th item type of goods is marked as the preferred item type; W e i represents the total number of goods consumed by the i-th user for whom the h-th item type of goods is marked as the preferred item type, in the mall;

[0089] For example, z is 4; W 2 1 is 30; W 2 2 is 40; W 2 3 is 60; W 2 4 is 50; W max is 100;

[0090] Calculate the overall interaction preference value L2 of each user for the goods of the 2nd item type in the mall:

[0091] ,

[0092] Step S203: When the overall interaction preference value L h is greater than the preset overall interaction preference threshold, it is determined that each user has an overall interaction preference for the goods of the h-th item type, and the h-th item type is recorded as the target preference item type of the mall. Obtain several target preference item types in the mall and gather them to obtain the target interaction preference data;

[0093] Step S300: Obtain the historical consumption records of users in the mall, analyze the consumption collocation of goods by users in the mall, and obtain the consumption collocation data of users;

[0094] Among them, Step S300 includes:

[0095] Step S301: Record the online and offline consumption behaviors of users in the mall to obtain the historical consumption records of users, and obtain the item types of several goods consumed by users from the historical consumption records;

[0096] Step S302: Analyze the consumption collocation between various item types of goods during the consumption process of users in the mall. Among them, analyze the consumption collocation between the goods of the e-th item type and the goods of the g-th item type during the consumption process of users in the mall. The specific process is as follows:

[0097] Obtain the total number Q of historical consumption records of users that only contain the goods of the e-th item type e sum , obtain the total number Q of historical consumption records of users that only contain the goods of the g-th item type g sum , obtain the total number Q of historical consumption records of users that contain both the goods of the e-th item type and the goods of the g-th item typee,g sum , obtain the total number Q of the historical consumption records of the user's commodities excluding the e-th commodity type and the g-th commodity type sum ;

[0098] Step S303: Respectively set the random variable X of the e-th commodity type and the random variable Y of the g-th commodity type, set the eigenvalue k, where k > 0, X = {k, 0}, Y = {k, 0}, and calculate the joint probability p(k, 0) = Q of the commodity of the e-th commodity type and the commodity of the g-th commodity type e sum / n, p(0, k) = Q g sum / n, p(k, k) = Q e,g sum / n, p(0, 0) = Q sum / n;

[0099] Calculate the marginal probability p(k) = (Q e sum +Q e,g sum ) / n, p(0) = (Q g sum +Q sum ) / n;

[0100] Calculate the marginal probability p(k) = (Q g sum +Q e,g sum ) / n, p(0) = (Q e sum +Q sum ) / n;

[0101] Step S304: Obtain the total number n of the user's various historical consumption records, and calculate the consumption matching value R between the commodity of the e-th commodity type and the commodity of the g-th commodity type e,g :

[0102] ,

[0103] where p(x, y) is the joint probability of the commodity of the e-th commodity type and the commodity of the g-th commodity type; p(x) is the marginal probability of the commodity of the e-th commodity type; p(y) is the marginal probability of the commodity of the g-th commodity type;

[0104] Step S305: When the consumption matching value R e,gIf it is greater than a preset consumption matching threshold, it is determined that there is a consumption matching between the commodity of the e-th commodity type and the commodity of the g-th commodity type during the user's shopping process in the mall. The e-th commodity type and the g-th commodity type are recorded as a consumption matching group of the user. Obtain and collect each consumption matching group of the user to obtain consumption matching data;

[0105] Step S400: Obtain the consumption matching data of each user in the mall, analyze the consumption matching degree between different commodities in the mall, and combine the target interaction tendency data to optimize and adjust the interaction strategy of the mall;

[0106] Among them, step S400 includes:

[0107] Step S401: Obtain the consumption matching data of each user in the mall, and analyze the consumption matching degree between the commodities of each commodity type in the mall. Among them, to analyze the consumption matching degree between the commodity of the v-th commodity type and the commodity of the u-th commodity type in the mall, the specific process is as follows:

[0108] When the v-th commodity type and the u-th commodity type are in the same consumption matching group in the consumption matching data of a certain user, mark the certain user, and obtain several marked users;

[0109] Step S402: Calculate the characteristic consumption matching value T between the commodity of the v-th commodity type and the commodity of the u-th commodity type in the mall v,u :

[0110] ,

[0111] Among them, β is the total number of several marked users; R α v,u is the consumption matching value between the commodity of the e-th commodity type and the commodity of the u-th commodity type in the α-th marked user;

[0112] Step S403: When the characteristic consumption matching value T v,u is greater than the preset characteristic consumption matching threshold, it is determined that there is a consumption matching between the commodity of the v-th commodity type and the commodity of the u-th commodity type in the mall, and the v-th commodity type is recorded as the matching commodity type of the u-th commodity type;

[0113] Step S404: Obtain several commodity types with consumption matching in the mall, obtain the target interaction tendency data of the mall in the current period, and obtain several target tendency commodity types of the mall from the target interaction tendency data;

[0114] Step S405: Denote several items of goods of the target tendency commodity types as popular goods, obtain several items of goods of the matching commodity types of the target tendency commodity types, and also denote them as popular goods. Adjust the procurement of goods in the mall during the current period, increase the procurement volume of each popular good during the current period, and reduce the procurement volume of other goods;

[0115] Step S406: On the online platform and offline platform of the mall, respectively increase the number of recommendations of the popular goods during the current period, and increase the frequency of the popular goods participating in activities, and adjust the interaction strategy of the mall;

[0116] In order to better implement the above method, a user deep interaction system based on a multi-source data platform is also proposed. The system includes an interaction tendency data module, a target interaction tendency data module, a consumption matching analysis module, and an interaction strategy optimization module;

[0117] The interaction tendency data module is used to evaluate the interaction tendency of users with respect to the goods in the mall to obtain interaction tendency data;

[0118] The target interaction tendency data module is used to analyze the overall interaction tendency of each user with respect to the goods in the mall to obtain the target interaction tendency data of the mall;

[0119] The consumption matching analysis module is used to analyze the consumption matching of users with respect to the goods in the mall to obtain the consumption matching data of users;

[0120] The interaction strategy optimization module is used to analyze the consumption matching degree between different goods in the mall, and combine the target interaction tendency data to optimize and adjust the interaction strategy of the mall;

[0121] Among them, the interaction tendency data module includes an interaction tendency value unit and an interaction tendency data unit;

[0122] The interaction tendency value unit is used to calculate the interaction tendency value of users with respect to each commodity type;

[0123] The interaction tendency data unit is used to evaluate the interaction tendency of users with respect to the goods in the mall according to the interaction tendency value to obtain interaction tendency data;

[0124] Among them, the target interaction tendency data module includes an overall interaction tendency value unit and a target interaction tendency data unit;

[0125] The overall interaction tendency value unit is used to calculate the overall interaction tendency value of each user with respect to the goods of each commodity type in the mall;

[0126] A target interaction tendency data unit, which is used to analyze the overall interaction tendency of each user with respect to the products of various product types in the mall according to the overall interaction tendency value, so as to obtain target interaction tendency data;

[0127] Among them, the consumption collocation analysis module includes a consumption collocation value unit and a consumption collocation analysis unit;

[0128] The consumption collocation value unit is used to calculate the consumption collocation values between the products of various product types;

[0129] The consumption collocation analysis unit is used to analyze the consumption collocation between the products of various product types during the consumption process of users in the mall according to the consumption collocation values, so as to obtain consumption collocation data;

[0130] Among them, the interaction strategy optimization module includes a consumption collocation degree analysis unit and an interaction strategy optimization unit;

[0131] The consumption collocation degree analysis unit is used to obtain the consumption collocation data of each user in the mall and analyze the consumption collocation degree between different products in the mall;

[0132] The interaction strategy optimization unit is used to optimize and adjust the interaction strategy of the mall according to the consumption collocation degree between different products in the mall and in combination with the target interaction tendency data.

[0133] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claims involved.

Claims

1. A method for in-depth user interaction based on a multi-source data platform, characterized in that The method includes: Step S100: Obtain the user's historical online interaction records and historical offline interaction records in the mall, evaluate the user's interaction tendency towards the goods in the mall, and obtain interaction tendency data; Step S200: Obtain the interaction tendency data of each user in the mall, analyze the overall interaction tendency of each user towards the goods in the mall, and obtain the target interaction tendency data of the mall; Step S300: Obtain the user's historical consumption records in the mall, analyze the consumption compatibility of the goods the user has in the mall, and obtain the user's consumption compatibility data; Step S400: Obtain the consumption compatibility data of each user in the mall, analyze the consumption compatibility degree between different goods in the mall, and combine the target interaction tendency data to optimize and adjust the interaction strategy of the mall; The step S100 includes: Step S101: Obtain the historical online interaction records and historical offline interaction records of the user in the mall respectively, and obtain the total number S' of consumer goods of the user on the online platform in the mall sum , and obtain the total number S of consumer goods of the user offline in the mall sum ; Step S102: Calculate the characteristic proportion coefficient B of the user's historical offline interaction records, where B = S sum / (S' sum +S sum ), and calculate the characteristic proportion coefficient B' of the user's historical online interaction records, where B' = S' sum / (S' sum +S sum ); Step S103: Obtain the commodity types of the commodities with which the user interacts from the historical online interaction records and the historical offline interaction records respectively, and calculate the interaction tendency values of the user for each commodity type. Among them, the interaction tendency value F of the user for the c-th commodity type c : ， Among them, A' c represents the total number of the user's historical online interaction records when the commodity type of the interactive commodity is the c-th commodity type; A c represents the total number of the user's historical offline interaction records when the commodity type of the interactive commodity is the c-th commodity type; A' sum represents the total number of the user's historical online interaction records; A sum represents the total number of the user's historical offline interaction records; When the interaction tendency value F c is greater than a preset interaction tendency threshold, it is determined that the user has an interaction tendency towards the commodity of the c-th commodity type, and the c-th commodity type is recorded as the user's tendency commodity type; Step S104: Obtain and collect various types of goods with the user's tendency to obtain the user's interaction tendency data; The step S300 includes: Step S301: Record the user's online and offline consumption behaviors in the mall to obtain the user's historical consumption records, and obtain the types of goods of several goods consumed by the user from the historical consumption records; Step S302: Analyze the consumption compatibility between various types of goods of the goods during the user's consumption process in the mall. Among them, when analyzing the consumption compatibility between the goods of the e-th type of goods and the goods of the g-th type of goods during the user's consumption process in the mall, the specific process is: Obtain the total number Q of the historical consumption records of the goods of the user that only contain the e-th item type of goods e sum , obtain the total number Q of the historical consumption records of the goods of the user that only contain the g-th item type of goods g sum , obtain the total number Q of the historical consumption records of the goods of the user that contain both the e-th item type and the g-th item type of goods e,g sum , obtain the total number Q of the historical consumption records of the goods of the user that do not contain the e-th item type and the g-th item type of goods sum ; Step S303: Respectively set the random variable X of the e-th commodity type and the random variable Y of the g-th commodity type, and set the eigenvalue k, where k > 0, X = {k, 0}, Y = {k, 0}, and calculate the joint probability p(k, 0) = Q of the commodity of the e-th commodity type and the commodity of the g-th commodity type e sum / n, p(0, k) = Q g sum / n, p(k, k) = Q e,g sum / n, p(0, 0) = Q sum / n; Calculate the marginal probability p(k) of the products of the e-th product type as p(k) = (Q e sum + Q e,g sum ) / n, p(0) = (Q g sum + Q sum ) / n; Calculate the marginal probability p(k) of the products of the g-th product type as p(k) = (Q g sum + Q e,g sum ) / n, p(0) = (Q e sum + Q sum ) / n; Step S304: Obtain the total number n of all historical consumption records of the user, and calculate the consumption matching value R between the commodity of the e-th commodity type and the commodity of the g-th commodity type e,g : ， Where p(x,y) is the joint probability of the goods of the e-th type of goods and the goods of the g-th type of goods; p(x) is the marginal probability of the goods of the e-th type of goods; p(y) is the marginal probability of the goods of the g-th type of goods; Step S305: When the consumption matching value R e,g is greater than a preset consumption matching threshold, it is determined that there is a consumption matching relationship between the commodity of the e-th commodity type and the commodity of the g-th commodity type during the user's consumption process in the shopping mall. The e-th commodity type and the g-th commodity type are recorded as a consumption matching group of the user, and each consumption matching group of the user is obtained and aggregated to obtain consumption matching data.

2. The user deep interaction method based on the multi-source data platform according to claim 1, characterized in that The step S200 includes: Step S201: Obtain the interaction tendency data of each user in the mall, and obtain several types of goods with the user's tendency from the interaction tendency data; Step S202: Analyze the overall interaction tendency of each user towards the goods of each type of goods in the mall. Among them, when analyzing the overall interaction tendency of each user towards the goods of the h-th type of goods in the mall, the specific process is: Obtain the maximum value W of the total number of consumer goods of each user in the mall max , calculate the overall interaction tendency value L of each user for the goods of the h-th commodity type in the mall h : ， where z represents the total number of users for whom the h-th item type is marked as a preferred item type; W e i represents the total number of items consumed by the i-th user for whom the h-th item type is marked as a preferred item type in the mall; Step S203: When the overall interaction tendency value L h is greater than a preset overall interaction tendency threshold, it is determined that each user has an overall interaction tendency towards the products of the h-th product type, and the h-th product type is recorded as the target tendency product type of the shopping mall. Obtain several target tendency product types in the shopping mall and gather them to obtain target interaction tendency data.

3. The user in-depth interaction method based on a multi-source data platform according to claim 2, wherein The step S400 includes: Step S401: Obtain the consumption compatibility data of each user in the mall, analyze the consumption compatibility degree between the goods of each type of goods in the mall. Among them, when analyzing the consumption compatibility degree between the goods of the v-th type of goods and the goods of the u-th type of goods in the mall, the specific process is: When the v-th type of goods and the u-th type of goods are in the same consumption compatibility group in the consumption compatibility data of a certain user, mark the certain user, and obtain several marked users; Step S402: Calculate the characteristic consumption matching value T between the v-th type of commodity and the u-th type of commodity in the mall v,u : ， where β is the total number of the marked users; R α v,u is the consumption matching value between the product of the e-th product type in the α-th marked user and the product of the u-th product type Step S403: When the characteristic consumption matching value T v,u is greater than a preset characteristic consumption matching threshold, it is determined that there is a consumption matching between the commodity of the v-th commodity type and the commodity of the u-th commodity type in the mall, and the v-th commodity type is recorded as the matching commodity type of the u-th commodity type; Step S404: Obtain several types of goods with consumption collocability in the mall, obtain the target interaction tendency data of the mall in the current period, and obtain several target tendency goods types of the mall from the target interaction tendency data; Step S405: Denote the goods of the several target tendency goods types as popular goods, obtain the goods of the collocation goods types of the several target tendency goods types, and also denote them as popular goods. Adjust the procurement of goods in the mall in the current period, increase the procurement quantity of each popular good in the current period, and reduce the procurement quantity of other goods; Step S406: On the online platform and offline platform of the mall, respectively increase the recommendation times of the popular goods in the current period, and increase the frequency of the popular goods participating in activities, and adjust the interaction strategy of the mall.

4. A user in-depth interaction system based on a multi-source data platform, which is used to execute the method for user in-depth interaction based on a multi-source data platform described in any one of claims 1-3, and is characterized in that, The system includes an interaction tendency data module, a target interaction tendency data module, a consumption collocation analysis module, and an interaction strategy optimization module; The interaction tendency data module is used to evaluate the interaction tendency of the user with the goods in the mall to obtain interaction tendency data; The target interaction tendency data module is used to analyze the overall interaction tendency of each user with the goods in the mall to obtain the target interaction tendency data of the mall; The consumption collocation analysis module is used to analyze the consumption collocability of the user with the goods in the mall to obtain the consumption collocation data of the user; The interaction strategy optimization module is used to analyze the consumption collocation degree between different goods in the mall, and combine the target interaction tendency data to optimize and adjust the interaction strategy of the mall.

5. The user in-depth interaction system based on the multi-source data platform according to claim 4, characterized in that The interaction tendency data module includes an interaction tendency value unit and an interaction tendency data unit; The interaction tendency value unit is used to calculate the interaction tendency value of the user for each type of goods; The interaction tendency data unit is used to evaluate the interaction tendency of the user with the goods in the mall according to the interaction tendency value to obtain interaction tendency data.

6. The user in-depth interaction system based on a multi-source data platform according to claim 4, characterized in that The target interaction tendency data module includes an overall interaction tendency value unit and a target interaction tendency data unit; The overall interaction tendency value unit is used to calculate the overall interaction tendency value of each user with the goods of each type of goods in the mall; The target interaction tendency data unit is used to analyze the overall interaction tendency of each user with the goods of each type of goods in the mall according to the overall interaction tendency value to obtain target interaction tendency data.

7. The user in-depth interaction system based on a multi-source data platform according to claim 4, characterized in that The consumption collocation analysis module includes a consumption collocation value unit and a consumption collocation analysis unit; The consumption collocation value unit is used to calculate the consumption collocation value between the goods of each type of goods; The consumption collocation analysis unit is used to analyze the consumption collocability between the goods of each type of goods in the consumption process of the user in the mall according to the consumption collocation value to obtain consumption collocation data.

8. The user deep interaction system based on a multi-source data platform according to claim 4, characterized in that, The interaction strategy optimization module includes a consumption collocation degree analysis unit and an interaction strategy optimization unit; The consumption matching degree analysis unit is used to obtain the consumption matching data of each user in the mall and analyze the consumption matching degree between different commodities in the mall; The interaction strategy optimization unit is used to optimize and adjust the interaction strategy of the mall according to the consumption matching degree between different commodities in the mall and in combination with the target interaction tendency data.

Images