CN118134611A - Cross-border electronic commerce big data analysis system applied to digital economy - Google Patents
Cross-border electronic commerce big data analysis system applied to digital economy Download PDFInfo
- Publication number
- CN118134611A CN118134611A CN202410560448.2A CN202410560448A CN118134611A CN 118134611 A CN118134611 A CN 118134611A CN 202410560448 A CN202410560448 A CN 202410560448A CN 118134611 A CN118134611 A CN 118134611A
- Authority
- CN
- China
- Prior art keywords
- event
- user
- module
- data set
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000007405 data analysis Methods 0.000 title claims abstract description 19
- 238000006243 chemical reaction Methods 0.000 claims abstract description 67
- 238000013507 mapping Methods 0.000 claims abstract description 23
- 238000010276 construction Methods 0.000 claims abstract description 7
- 238000012795 verification Methods 0.000 claims description 15
- 238000012360 testing method Methods 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 11
- 238000000034 method Methods 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 6
- 230000000007 visual effect Effects 0.000 claims description 5
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000000638 solvent extraction Methods 0.000 claims 1
- 238000005516 engineering process Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a cross-border electronic commerce big data analysis system applied to digital economy, which relates to the technical field of business data prediction and comprises a construction module, a data analysis module and a data analysis module, wherein the construction module is used for constructing an event sequence data set; the determining module is used for obtaining the actual user event path of each user; the dividing module is used for comparing the standard user event path with the actual user event path to confirm the typical event path and preference of the user, determining the loyalty of the user and dividing the user into different user groups; the recommendation module is used for determining the conversion rate of the conversion stage and constructing a personalized recommendation model according to the conversion rate, the mapping list of the user ID and the typical event path-preference and the event sequence data set; and the display module is used for visually displaying the personalized content. Redundancy brought by the past time series data or other dimension data is reduced, and the effectiveness of the data is ensured. The event behavior mode of the user is accurately reflected. The adaptability and the accuracy of personalized recommendation are ensured.
Description
Technical Field
The application relates to the technical field of business data prediction, in particular to a cross-border electronic commerce big data analysis system applied to digital economy.
Background
The method is applied to large data analysis of a digital economic cross-border electronic commerce, and relates to key technologies such as cloud computing, large data, artificial intelligence, the Internet of things, block chains, mobile communication, GIS, social networks, network security, electronic commerce and the like. The technology provides powerful data processing and analysis capability for cross-border electronic commerce enterprises, and is beneficial to improving the operation efficiency and competitiveness of the enterprises. Through the technologies, cross-border e-commerce enterprises can better understand user demands, conduct accurate marketing, optimize logistics distribution routes, guarantee the safety and transparency of transactions, and provide more personalized services. With the continuous development of technologies, cross-border e-commerce enterprises can better utilize the technologies to realize continuous growth and optimization of services.
In the prior art, because the business data has a plurality of types and a large quantity, time series data are often adopted, the data redundancy degree is high, and the accuracy and the adaptability of personalized recommendation of big data analysis are low.
Therefore, how to improve the accuracy and adaptability of personalized recommendation is a technical problem to be solved at present.
Disclosure of Invention
The invention provides a cross-border electronic commerce big data analysis system applied to digital economy, which is used for solving the technical problems of low accuracy and adaptability of personalized recommendation in the prior art. The method comprises the following steps:
The construction module is used for defining events on the E-commerce platform, determining a standard user event path, acquiring E-commerce data of a user, extracting event information from the E-commerce data of the user and constructing an event sequence data set;
The determining module is used for selecting a plurality of events from the standard user event paths as key events, intercepting an event sequence data set by taking the key events as nodes to obtain an event subsequence data set, screening frequent subsequences in the event subsequence data set, and obtaining the actual user event paths of each user by splicing the frequent subsequences corresponding to different key events;
The dividing module is used for comparing the standard user event path and the actual user event path to confirm the typical event path and preference of the user, and establishing a mapping list of user ID and typical event path-preference so as to determine the user loyalty and divide the user into different user groups;
The recommendation module is used for defining a conversion stage related to personalized recommendation of the user group, determining the conversion rate of the conversion stage, and constructing a personalized recommendation model according to the conversion rate, a mapping list of the user ID and typical event path-preference and an event sequence data set;
the display module is used for recommending the personalized content to the user through the personalized recommendation model, mapping the recommended personalized content and displaying the personalized content in a visual mode.
In some embodiments of the application, the building block is configured to:
determining a starting point event and an end point event according to the type of the electronic commerce ring segment, and analyzing a fixed event and a non-fixed event between the starting point event and the end point event;
and determining the fixed event sequence and the non-fixed event sequence between the starting point event and the end point event according to the sequence and the occurrence frequency of the fixed event and the non-fixed event of the user historically, so as to obtain a standard user event path.
In some embodiments of the present application, the determining module is configured to:
Calculating the contribution degree of each event in the starting point event and the end point event to the next event and the contribution degree of each event in the end point event, carrying out weighted summation on the two contribution degrees to obtain the contribution degree index of each event in the starting point event and the end point event, sequencing from small to large, and taking the event with the contribution degree index exceeding the median as a key event.
In some embodiments of the present application, the determining module is further configured to:
Splitting the event sequence data set according to a complete electronic commerce ring type period to obtain an event sequence data set representing the complete electronic commerce ring type period;
Calculating the event complexity corresponding to each event sequence data set;
;
Wherein, For the x-th time series dataset to correspond to event complexity,AndComplex conversion coefficients corresponding to fixed and non-fixed events in the x-th time series dataset,AndThe number of fixed event categories and the number of non-fixed event categories in the x-th time series data set,As a complex weight for the kth stationary event,For the frequency of occurrence of the kth stationary event,For the duration of the kth fixation event,For complex weights of the g-th non-stationary event,For the frequency of occurrence of the g-th non-stationary event,Duration of the non-fixation event of type g;
Intercepting event sequence data sets by taking key events as nodes to obtain event subsequence data sets corresponding to three time node sections of a time node between a starting point event and the key event, a time node between a plurality of key events and the key event and a time node between the key event and an end point event, and distributing event complexity according to the lengths of the three time node sections to obtain event complexity of each event subsequence data set;
Determining the number of event types and a support threshold value through the event complexity of each event subsequence data set, finding a plurality of alternative subsequences in each event subsequence data set according to the number of event types, calculating the support degree of each alternative subsequence in the event subsequence data set corresponding to three time node segments in a plurality of periods, and taking the alternative subsequence with the support degree exceeding the corresponding support degree threshold value as a frequent subsequence.
In some embodiments of the present application, the determining module is further configured to:
And splicing different frequent subsequences in event subsequence data sets corresponding to the three time node sections of the time node between the starting point event and the key event, the time node between a plurality of key events and the key event and the time node between the key event and the terminal event to obtain a complete event subsequence data set, and determining the actual user event path of each user by analyzing the event subsequence data set.
In some embodiments of the present application, the dividing module is further configured to:
Obtaining purchase frequency, purchase amount, commodity preference index, platform using time length and platform registration time length according to the user ID and the mapping list of typical event paths-preferences;
Determining user loyalty through purchase frequency, purchase amount, commodity preference index, platform use time length and platform registration time length;
;
Wherein P is the user loyalty, In order to purchase a relative loyalty weight for the frequency,In order to purchase the frequency of the frequency,For the relative loyalty weight of the purchase amount,In order to purchase the amount of money,For the relative loyalty weights that are preferred for the merchandise,In order to provide a commodity preference index,For the duration of the use of the platform,A is a preset constant for the registration duration of the platform;
the users are divided into different user groups through different intervals to which the user loyalty belongs.
In some embodiments of the present application, the recommendation module is configured to:
The event to which the personalized recommendation service is applied is referred to as a recommendation event, the conversion process of the recommendation event and the adjacent event in the actual user event path of each user group is referred to as a first conversion stage, the conversion process of the recommendation event and the key event in the actual user event path of each user group is referred to as a second conversion stage, and the conversion rates of the first conversion stage and the second conversion stage are calculated.
In some embodiments of the present application, the recommendation module is further configured to:
Respectively integrating the conversion rates of the first conversion stage and the second conversion stage, and carrying out weighted summation to obtain a conversion rate index of each user group;
Determining respective event sequence data sets through typical event paths and preferences of each user, determining dividing ratios of a training set, a testing set and a verification set according to conversion rate indexes of each user group, and dividing the two event sequence data sets into the training set, the testing set and the verification set according to the dividing ratios;
and constructing a personalized recommendation model through the training set, the testing set and the verification set.
By applying the technical scheme, the construction module is used for defining the event on the E-commerce platform, determining a standard user event path, acquiring the E-commerce data of the user, extracting event information from the E-commerce data of the user and constructing an event sequence data set; the determining module is used for selecting a plurality of events from the standard user event paths as key events, intercepting an event sequence data set by taking the key events as nodes to obtain an event subsequence data set, screening frequent subsequences in the event subsequence data set, and obtaining the actual user event paths of each user by splicing the frequent subsequences corresponding to different key events; the dividing module is used for comparing the standard user event path and the actual user event path to confirm the typical event path and preference of the user, and establishing a mapping list of user ID and typical event path-preference so as to determine the user loyalty and divide the user into different user groups; the recommendation module is used for defining a conversion stage related to personalized recommendation of the user group, determining the conversion rate of the conversion stage, and constructing a personalized recommendation model according to the conversion rate, a mapping list of the user ID and typical event path-preference and an event sequence data set; the display module is used for recommending the personalized content to the user through the personalized recommendation model, mapping the recommended personalized content and displaying the personalized content in a visual mode. By constructing the event sequence data set, the application reduces redundancy brought by the past time sequence data or other dimension data and ensures the validity of the data. And by splicing the frequent subsequences corresponding to different key events, the actual user event path of each user is obtained, and the event behavior mode of the user is accurately reflected. And constructing a personalized recommendation model according to the conversion rate, the mapping list of the user ID and the typical event path-preference and the event sequence data set, so that the adaptability and the accuracy of personalized recommendation are ensured.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 shows a schematic structural diagram of a cross-border electronic commerce big data analysis system applied to digital economy according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
The embodiment of the application provides a cross-border electronic commerce big data analysis system applied to digital economy, which is shown in fig. 1 and comprises the following contents:
The construction module is used for defining the event on the E-commerce platform, determining a standard user event path, acquiring the E-commerce data of the user, extracting event information from the E-commerce data of the user and constructing an event sequence data set.
In this embodiment, events on the e-commerce platform are defined according to user behavior, for example, browsing merchandise: the user accesses the merchandise page to view the merchandise details. Added to shopping carts: the user selects and adds items of interest to the shopping cart. Accessing a shopping cart: the user returns to the shopping cart page to view the added merchandise. And determining different standard user event paths according to different link target requirements.
In this embodiment, event information (user ID, timestamp, event type, commodity ID, etc.) is extracted from the e-commerce data of the user, an event sequence data set is constructed, the behavior record of each user is converted into events, it is ensured that each event has a definite time sequence, and corresponding data corresponding to various events are marked on the sequence in the time sequence. The event sequence data typically records a series of interactions of the user on the platform, such as browsing merchandise, adding to a shopping cart, placing orders, paying, etc. These data reflect the actual operation of the user and thus redundant information that may be present in the time series data can be avoided to some extent. In terms of data validity, the event sequence data provides more detailed behavior information, which helps to understand the behavior patterns and preferences of the user.
In some embodiments of the application, the building block is configured to:
determining a starting point event and an end point event according to the type of the electronic commerce ring segment, and analyzing a fixed event and a non-fixed event between the starting point event and the end point event;
and determining the fixed event sequence and the non-fixed event sequence between the starting point event and the end point event according to the sequence and the occurrence frequency of the fixed event and the non-fixed event of the user historically, so as to obtain a standard user event path.
In this embodiment, for example, when the type of the electronic commerce ring is a purchasing link, the starting point event is browsing goods, and the ending point event may be a payment event or a confirmation receiving event. A fixed event is an event that must exist between a start event and an end event, for example, when the e-commerce link type is a purchasing link, then an order is placed for the event that must exist between the start event and the end event. A non-stationary event is an event that does not necessarily exist between a start event and an end event, such as adding a shopping cart.
In this embodiment, the fixed event sequence and the non-fixed event sequence between the start event and the end event are determined according to the sequences and the occurrence frequency of the fixed event and the non-fixed event of the user in history, the sequences of all the users on the fixed event and the non-fixed event are compared, and the sequence with the highest occurrence frequency is used as the sequence of the occurrence of the fixed event and the non-fixed event.
The determining module is used for selecting a plurality of events from the standard user event paths as key events, intercepting an event sequence data set by taking the key events as nodes to obtain an event subsequence data set, screening frequent subsequences in the event subsequence data set, and obtaining the actual user event paths of each user by splicing the frequent subsequences corresponding to different key events.
In this embodiment, whether a plurality of events are key events is determined by contribution degree, and event sequence data sets are intercepted by taking the key events as nodes, so as to obtain event subsequence data sets of the same plurality of event node segments of users in a plurality of periods, event subsequence data sets of the same plurality of event node segments of the users in different periods are compared, frequently occurring screening frequent subsequences are screened, and frequent subsequences corresponding to different key events are spliced, so that an actual user event path of each user is obtained.
In some embodiments of the present application, the determining module is configured to:
Calculating the contribution degree of each event in the starting point event and the end point event to the next event and the contribution degree of each event in the end point event, carrying out weighted summation on the two contribution degrees to obtain the contribution degree index of each event in the starting point event and the end point event, sequencing from small to large, and taking the event with the contribution degree index exceeding the median as a key event.
In this embodiment, the contribution degree is calculated by calculating historical data between two events, and calculating probability of occurrence of one event and probability of occurrence of two events simultaneously after occurrence of the other event, so as to convert the probability into the contribution degree.
In some embodiments of the present application, the determining module is further configured to:
Splitting the event sequence data set according to a complete electronic commerce ring type period to obtain an event sequence data set representing the complete electronic commerce ring type period;
Calculating the event complexity corresponding to each event sequence data set;
;
Wherein, For the x-th time series dataset to correspond to event complexity,AndComplex conversion coefficients corresponding to fixed and non-fixed events in the x-th time series dataset,AndThe number of fixed event categories and the number of non-fixed event categories in the x-th time series data set,As a complex weight for the kth stationary event,For the frequency of occurrence of the kth stationary event,For the duration of the kth fixation event,For complex weights of the g-th non-stationary event,For the frequency of occurrence of the g-th non-stationary event,Duration of the non-fixation event of type g;
Intercepting event sequence data sets by taking key events as nodes to obtain event subsequence data sets corresponding to three time node sections of a time node between a starting point event and the key event, a time node between a plurality of key events and the key event and a time node between the key event and an end point event, and distributing event complexity according to the lengths of the three time node sections to obtain event complexity of each event subsequence data set;
Determining the number of event types and a support threshold value through the event complexity of each event subsequence data set, finding a plurality of alternative subsequences in each event subsequence data set according to the number of event types, calculating the support degree of each alternative subsequence in the event subsequence data set corresponding to three time node segments in a plurality of periods, and taking the alternative subsequence with the support degree exceeding the corresponding support degree threshold value as a frequent subsequence.
In this embodiment, the event sequence data set is split according to a complete e-commerce ring type period (start point event-end point event), so as to obtain an event sequence data set representing a complete e-commerce ring type period, i.e. different event sequence data sets correspond to different periods of the user.
In this embodiment, the event complexity (the activity level or the behavior complexity of the user on the platform) corresponding to the event sequence data set is different, and the support threshold value corresponding to the difference is also different, and the mapping relationship can be determined according to the historical data.
In this embodiment, the event complexity is allocated according to the lengths of the three time node segments, so as to obtain the event complexity of each event sub-sequence data set, that is, the three time segments allocate a complete event complexity according to a preset proportion, so as to obtain the event complexity of each event sub-sequence data set.
In this embodiment, the event type number and the support threshold are determined by the event complexity of each event sub-sequence data set, and when each event complexity is mapped to one support threshold, one event type number is also mapped at the same time. For example, the number of event categories is 3, then there will be a sequence of 3 consecutive events as an alternative subsequence. The support is calculated by comparing a number of alternative subsequences to a number of corresponding event subsequences, here a number of identical event subsequence data sets of different periods of the same user.
It should be noted that, the number of event types determined by the event complexity is at least 2, because at least two different events are continuous to represent the user behavior pattern to some extent.
In some embodiments of the present application, the determining module is further configured to:
And splicing different frequent subsequences in event subsequence data sets corresponding to the three time node sections of the time node between the starting point event and the key event, the time node between a plurality of key events and the key event and the time node between the key event and the terminal event to obtain a complete event subsequence data set, and determining the actual user event path of each user by analyzing the event subsequence data set.
And the dividing module is used for comparing the standard user event path and the actual user event path to confirm typical event paths and preferences of the user, and establishing a mapping list of user IDs and typical event paths and preferences so as to determine the loyalty of the user and divide the user into different user groups.
In this embodiment, the typical event path and the preference path (partial event path) of the user are confirmed by comparing the standard user event path and the actual user event path.
In some embodiments of the present application, the dividing module is further configured to:
Obtaining purchase frequency, purchase amount, commodity preference index, platform using time length and platform registration time length according to the user ID and the mapping list of typical event paths-preferences;
Determining user loyalty through purchase frequency, purchase amount, commodity preference index, platform use time length and platform registration time length;
;
Wherein P is the user loyalty, In order to purchase a relative loyalty weight for the frequency,In order to purchase the frequency of the frequency,For the relative loyalty weight of the purchase amount,In order to purchase the amount of money,For the relative loyalty weights that are preferred for the merchandise,In order to provide a commodity preference index,For the duration of the use of the platform,A is a preset constant for the registration duration of the platform;
the users are divided into different user groups through different intervals to which the user loyalty belongs.
In this embodiment, purchase frequency, purchase amount, commodity preference index, platform use time length and platform registration time length are obtained according to previous data corresponding to a mapping list of a user ID and a typical event path-preference, where the commodity preference index is obtained by comprehensively analyzing commodity click rate, purchase rate, browsing time length, score and the like to affect user loyalty.
The recommendation module is used for defining conversion stages related to personalized recommendation of the user group, determining conversion rate of the conversion stages, and constructing a personalized recommendation model according to the conversion rate, the mapping list of the user ID and typical event path-preference and the event sequence data set.
In some embodiments of the present application, the recommendation module is configured to:
The event to which the personalized recommendation service is applied is referred to as a recommendation event, the conversion process of the recommendation event and the adjacent event in the actual user event path of each user group is referred to as a first conversion stage, the conversion process of the recommendation event and the key event in the actual user event path of each user group is referred to as a second conversion stage, and the conversion rates of the first conversion stage and the second conversion stage are calculated.
In this embodiment, the conversion rate is the ratio of the number of users between the completion of two events.
In some embodiments of the present application, the recommendation module is further configured to:
Respectively integrating the conversion rates of the first conversion stage and the second conversion stage, and carrying out weighted summation to obtain a conversion rate index of each user group;
Determining respective event sequence data sets through typical event paths and preferences of each user, determining dividing ratios of a training set, a testing set and a verification set according to conversion rate indexes of each user group, and dividing the two event sequence data sets into the training set, the testing set and the verification set according to the dividing ratios;
and constructing a personalized recommendation model through the training set, the testing set and the verification set.
In this embodiment, the personalized recommendation model is input as an event sequence data set, and output as personalized recommendation indexes of different users for different recommended contents.
In this embodiment, the dividing ratio of the training set, the testing set and the verification set is determined according to the conversion rate index of each user group, and different conversion rate indexes correspond to three ratios of different training sets, testing sets and verification sets, so as to obtain the personalized recommendation model through the data.
In the process of verifying the model by the verification set, the loss coefficient is adjusted, and after the verification data set is input into the personalized recommendation model, an output predicted value and an output true value aiming at each verification data are obtained, and the loss coefficient is corrected;
;
Wherein, The loss coefficient of the personalized recommendation index for the recommended content after the i+1st correction is calculated,For the i-th personalized recommendation index true value,For the i-th personalized recommendation index prediction value,For the i-th modified loss coefficient, exp is an exponential function,AndThe value ranges of the first contribution factor and the second contribution factor which are respectively corresponding to the event data circulation efficiency under the personalized recommendation function are between 0 and 1.
The output test value and the output true value corresponding to each verification data can correct the loss coefficient once; and outputting the finally obtained corrected loss coefficients as the loss coefficients of the personalized recommended indexes according to all the calculated corrected loss coefficients.
The display module is used for recommending the personalized content to the user through the personalized recommendation model, mapping the recommended personalized content and displaying the personalized content in a visual mode.
By applying the technical scheme, the construction module is used for defining the event on the E-commerce platform, determining a standard user event path, acquiring the E-commerce data of the user, extracting event information from the E-commerce data of the user and constructing an event sequence data set; the determining module is used for selecting a plurality of events from the standard user event paths as key events, intercepting an event sequence data set by taking the key events as nodes to obtain an event subsequence data set, screening frequent subsequences in the event subsequence data set, and obtaining the actual user event paths of each user by splicing the frequent subsequences corresponding to different key events; the dividing module is used for comparing the standard user event path and the actual user event path to confirm the typical event path and preference of the user, and establishing a mapping list of user ID and typical event path-preference so as to determine the user loyalty and divide the user into different user groups; the recommendation module is used for defining a conversion stage related to personalized recommendation of the user group, determining the conversion rate of the conversion stage, and constructing a personalized recommendation model according to the conversion rate, a mapping list of the user ID and typical event path-preference and an event sequence data set; the display module is used for recommending the personalized content to the user through the personalized recommendation model, mapping the recommended personalized content and displaying the personalized content in a visual mode. By constructing the event sequence data set, the application reduces redundancy brought by the past time sequence data or other dimension data and ensures the validity of the data. And by splicing the frequent subsequences corresponding to different key events, the actual user event path of each user is obtained, and the event behavior mode of the user is accurately reflected. And constructing a personalized recommendation model according to the conversion rate, the mapping list of the user ID and the typical event path-preference and the event sequence data set, so that the adaptability and the accuracy of personalized recommendation are ensured.
From the above description of the embodiments, it will be clear to those skilled in the art that the present invention may be implemented in hardware, or may be implemented by means of software plus necessary general hardware platforms. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (may be a CD-ROM, a U-disk, a mobile hard disk, etc.), and includes several instructions for causing a computer device (may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective implementation scenario of the present invention.
Those skilled in the art will appreciate that the modules in the system in the implementation scenario may be distributed in the system in the implementation scenario according to the implementation scenario description, or that corresponding changes may be located in one or more systems different from the implementation scenario. The modules of the implementation scenario may be combined into one module, or may be further split into a plurality of sub-modules.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present application, and are not limiting; although the application has been described in detail with reference to the foregoing embodiments, it will be appreciated by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not drive the essence of the corresponding technical solutions to depart from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims (8)
1. A cross-border electronic commerce big data analysis system for digital economy, comprising:
The construction module is used for defining events on the E-commerce platform, determining a standard user event path, acquiring E-commerce data of a user, extracting event information from the E-commerce data of the user and constructing an event sequence data set;
The determining module is used for selecting a plurality of events from the standard user event paths as key events, intercepting an event sequence data set by taking the key events as nodes to obtain an event subsequence data set, screening frequent subsequences in the event subsequence data set, and obtaining the actual user event paths of each user by splicing the frequent subsequences corresponding to different key events;
The dividing module is used for comparing the standard user event path and the actual user event path to confirm the typical event path and preference of the user, and establishing a mapping list of user ID and typical event path-preference so as to determine the user loyalty and divide the user into different user groups;
The recommendation module is used for defining a conversion stage related to personalized recommendation of the user group, determining the conversion rate of the conversion stage, and constructing a personalized recommendation model according to the conversion rate, a mapping list of the user ID and typical event path-preference and an event sequence data set;
the display module is used for recommending the personalized content to the user through the personalized recommendation model, mapping the recommended personalized content and displaying the personalized content in a visual mode.
2. The cross-border e-commerce big data analysis system for use in a digital economy as claimed in claim 1, wherein the building module is configured to:
determining a starting point event and an end point event according to the type of the electronic commerce ring segment, and analyzing a fixed event and a non-fixed event between the starting point event and the end point event;
and determining the fixed event sequence and the non-fixed event sequence between the starting point event and the end point event according to the sequence and the occurrence frequency of the fixed event and the non-fixed event of the user historically, so as to obtain a standard user event path.
3. The cross-border e-commerce big data analysis system for use in a digital economy as claimed in claim 2, wherein the determining module is configured to:
Calculating the contribution degree of each event in the starting point event and the end point event to the next event and the contribution degree of each event in the end point event, carrying out weighted summation on the two contribution degrees to obtain the contribution degree index of each event in the starting point event and the end point event, sequencing from small to large, and taking the event with the contribution degree index exceeding the median as a key event.
4. The cross-border e-commerce big data analysis system for use in a digital economy of claim 2 wherein the determination module is further configured to:
Splitting the event sequence data set according to a complete electronic commerce ring type period to obtain an event sequence data set representing the complete electronic commerce ring type period;
Calculating the event complexity corresponding to each event sequence data set;
;
Wherein, Event complexity for the x-th time series dataset,/>And/>Complex conversion coefficients corresponding to fixed and non-fixed events, respectively,/>, in the xth time series datasetAnd/>Fixed event type number and non-fixed event type number,/>, respectively, in the xth time series datasetComplex weights for kth fixed event,/>For the frequency of occurrence of the kth stationary event,/>For the duration of the kth stationary event,/>Complex weights for the g-th non-stationary event,/>For the frequency of occurrence of the g-th non-stationary event,/>Duration of the non-fixation event of type g;
Intercepting event sequence data sets by taking key events as nodes to obtain event subsequence data sets corresponding to three time node sections of a time node between a starting point event and the key event, a time node between a plurality of key events and the key event and a time node between the key event and an end point event, and distributing event complexity according to the lengths of the three time node sections to obtain event complexity of each event subsequence data set;
Determining the number of event types and a support threshold value through the event complexity of each event subsequence data set, finding a plurality of alternative subsequences in each event subsequence data set according to the number of event types, calculating the support degree of each alternative subsequence in the event subsequence data set corresponding to three time node segments in a plurality of periods, and taking the alternative subsequence with the support degree exceeding the corresponding support degree threshold value as a frequent subsequence.
5. The cross-border e-commerce big data analysis system for use in a digital economy of claim 4 wherein the determination module is further configured to:
And splicing different frequent subsequences in event subsequence data sets corresponding to the three time node sections of the time node between the starting point event and the key event, the time node between a plurality of key events and the key event and the time node between the key event and the terminal event to obtain a complete event subsequence data set, and determining the actual user event path of each user by analyzing the event subsequence data set.
6. The cross-border e-commerce big data analysis system for use in a digital economy as claimed in claim 1, wherein the partitioning module is further configured to:
Obtaining purchase frequency, purchase amount, commodity preference index, platform using time length and platform registration time length according to the user ID and the mapping list of typical event paths-preferences;
Determining user loyalty through purchase frequency, purchase amount, commodity preference index, platform use time length and platform registration time length;
;
Wherein P is the user loyalty, For the relative loyalty of purchase frequency,/>To purchase frequency,/>For relative loyalty weight of purchase amount,/>For purchase amount,/>Relative loyalty weights for merchandise preference,/>For commodity preference index,/>For the duration of platform use,/>A is a preset constant for the registration duration of the platform;
the users are divided into different user groups through different intervals to which the user loyalty belongs.
7. The cross-border e-commerce big data analysis system for use in a digital economy as claimed in claim 3, wherein said recommendation module is configured to:
The event to which the personalized recommendation service is applied is referred to as a recommendation event, the conversion process of the recommendation event and the adjacent event in the actual user event path of each user group is referred to as a first conversion stage, the conversion process of the recommendation event and the key event in the actual user event path of each user group is referred to as a second conversion stage, and the conversion rates of the first conversion stage and the second conversion stage are calculated.
8. The cross-border e-commerce big data analysis system for use in a digital economy of claim 7 wherein the recommendation module is further configured to:
Respectively integrating the conversion rates of the first conversion stage and the second conversion stage, and carrying out weighted summation to obtain a conversion rate index of each user group;
Determining respective event sequence data sets through typical event paths and preferences of each user, determining dividing ratios of a training set, a testing set and a verification set according to conversion rate indexes of each user group, and dividing the two event sequence data sets into the training set, the testing set and the verification set according to the dividing ratios;
and constructing a personalized recommendation model through the training set, the testing set and the verification set.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410560448.2A CN118134611B (en) | 2024-05-08 | 2024-05-08 | Cross-border electronic commerce big data analysis system applied to digital economy |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410560448.2A CN118134611B (en) | 2024-05-08 | 2024-05-08 | Cross-border electronic commerce big data analysis system applied to digital economy |
Publications (2)
Publication Number | Publication Date |
---|---|
CN118134611A true CN118134611A (en) | 2024-06-04 |
CN118134611B CN118134611B (en) | 2024-07-09 |
Family
ID=91240773
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410560448.2A Active CN118134611B (en) | 2024-05-08 | 2024-05-08 | Cross-border electronic commerce big data analysis system applied to digital economy |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN118134611B (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107729552A (en) * | 2017-11-03 | 2018-02-23 | 广州优视网络科技有限公司 | User behavior method for tracing, device, user terminal and server |
CN112819336A (en) * | 2021-02-03 | 2021-05-18 | 国家电网有限公司 | Power monitoring system network threat-based quantification method and system |
CN114374595A (en) * | 2022-01-13 | 2022-04-19 | 平安普惠企业管理有限公司 | Event node attribution analysis method and device, electronic equipment and storage medium |
WO2023029589A1 (en) * | 2021-08-31 | 2023-03-09 | 上海商汤智能科技有限公司 | Neural network compilation method and apparatus, device, and storage medium |
CN117874086A (en) * | 2024-01-15 | 2024-04-12 | 北京工业大学 | Dynamic sharing method for multi-aggregation query in complex event processing |
-
2024
- 2024-05-08 CN CN202410560448.2A patent/CN118134611B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107729552A (en) * | 2017-11-03 | 2018-02-23 | 广州优视网络科技有限公司 | User behavior method for tracing, device, user terminal and server |
CN112819336A (en) * | 2021-02-03 | 2021-05-18 | 国家电网有限公司 | Power monitoring system network threat-based quantification method and system |
WO2023029589A1 (en) * | 2021-08-31 | 2023-03-09 | 上海商汤智能科技有限公司 | Neural network compilation method and apparatus, device, and storage medium |
CN114374595A (en) * | 2022-01-13 | 2022-04-19 | 平安普惠企业管理有限公司 | Event node attribution analysis method and device, electronic equipment and storage medium |
CN117874086A (en) * | 2024-01-15 | 2024-04-12 | 北京工业大学 | Dynamic sharing method for multi-aggregation query in complex event processing |
Non-Patent Citations (1)
Title |
---|
江琴;刘琳岚;苏曦;蔡轲;: "基于事件权重的GUI测试路径生成方法", 计算机应用, no. 05, 1 May 2009 (2009-05-01) * |
Also Published As
Publication number | Publication date |
---|---|
CN118134611B (en) | 2024-07-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20180341898A1 (en) | Demand forecast | |
US20200234218A1 (en) | Systems and methods for entity performance and risk scoring | |
CN105488163B (en) | Information-pushing method and device | |
US20230230081A1 (en) | Account identification method, apparatus, electronic device and computer readable medium | |
CN118134611B (en) | Cross-border electronic commerce big data analysis system applied to digital economy | |
CN115841345B (en) | Cross-border big data intelligent analysis method, system and storage medium | |
CN115293291B (en) | Training method and device for sequencing model, sequencing method and device, electronic equipment and medium | |
CN117132315A (en) | Active user prediction method, device, equipment and storage medium | |
CN109189810B (en) | Query method, query device, electronic equipment and computer-readable storage medium | |
WO2020150597A1 (en) | Systems and methods for entity performance and risk scoring | |
CN110807687A (en) | Object data processing method, device, computing equipment and medium | |
CN114943563A (en) | Rights and interests pushing method and device, computer equipment and storage medium | |
US11422993B2 (en) | Duplicate table identification in enterprise database systems for data storage optimization | |
Guiffard | Valuing the virtual: The impact of fiber to the home on property prices in France | |
Herrmann | How to increase profits through predictive analytics when only few competitors’ bids are known | |
CN113052509A (en) | Model evaluation method, model evaluation apparatus, electronic device, and storage medium | |
US10891664B2 (en) | System and method for facilitating non-parametric weighted correlation analysis | |
KR102290798B1 (en) | Goods management server and method capabel of managing goods based on quantum theory | |
CN113763065B (en) | Method, apparatus, device and computer readable medium for recommending commodity | |
Ouamani et al. | A Hybrid Model for Demand Forecasting Based on the Combination of Statistical and Machine Learning Methods | |
US20230325840A1 (en) | Transaction-risk evaluation by resource-limited devices | |
CN115409636A (en) | Product risk prediction method, device, equipment and medium | |
CN112785197A (en) | Information recommendation method, device, computing equipment, medium and program product | |
US20120278218A1 (en) | Ascertaining market value through a competitive valuation process | |
KR20240028046A (en) | Loan interest rate, loan limit, and product recommendation service system using AI deep learning method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |