User behavior attribution analysis method and system

By setting up multiple attribution analysis models and visualization modules in user behavior attribution analysis, the problems of inflexibility and accuracy of analysis in the prior art are solved, and analysis efficiency and user experience are improved, and it is suitable for multiple business fields.

CN119046042APending Publication Date: 2024-11-29FUJIAN PUPU INFORMATION TECH CO LTD
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411141071.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-20
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The existing user behavior attribution analysis methods are inflexible and cannot adjust the algorithm model according to specific business scenarios, resulting in low analysis efficiency, poor accuracy, and poor user experience.

Method used

Provide a user behavior attribution analysis method. By setting up multiple attribution analysis model types, selecting the most appropriate model based on business scenarios, calculating the contribution of events to be attributable, and displaying the results through visual modules to improve analysis flexibility. and accuracy.

Benefits of technology

It improves the efficiency, flexibility and accuracy of user behavior attribution analysis, and is suitable for areas such as e-commerce, digital marketing and user experience optimization, and lowers the threshold for user use.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119046042A_ABST
    Figure CN119046042A_ABST
Patent Text Reader

Abstract

The invention discloses a user behavior attribution analysis method and system, and the method comprises the following steps: collecting behavior event data of a user, and constructing a user behavior path according to a time sequence; input attribution analysis parameters are received, the attribution analysis parameters comprise a target conversion event, a to-be-attributed event and the type of the attribution analysis model, and an effective user behavior path comprising the target conversion event and the to-be-attributed event is obtained; according to the type of the attribution analysis model, calculating the contribution degree of each event to be attributed in each effective user behavior path; summing up the sum of the contribution degrees of the events to be attributed in all the effective user behavior paths to obtain an attribution analysis result; by adopting the technology, the analyzed target conversion event, the to-be-attributed event and the type of the attribution analysis model can be selected according to the business scene and the demand, so that the efficiency, the flexibility and the accuracy of contribution attribution analysis of the to-be-attributed event to the target conversion event are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a user behavior attribution analysis method and system. Background Art

[0002] User behavior attribution analysis is an important method to evaluate the impact of different factors (marketing resource positions, advertising placement, page / functional modules, etc.) on results (such as sales results, conversion rates, user behavior, etc.). It is widely used in Internet digital marketing, user experience optimization and other fields.

[0003] Existing user behavior attribution analysis usually relies on data developers to develop fixed attribution analysis reports or manually collect data. The main method is to predetermine the target conversion event, perform attribution analysis on the achievement of the final target conversion event according to the final click, and calculate the contribution by finding the to-be-attributed event closest to the target conversion event. Since the target event and / or to-be-attributed event are fixed, it is impossible to flexibly adjust the algorithm model of attribution analysis according to specific business scenarios, that is, it is impossible to flexibly select the target conversion event to be evaluated and the to-be-attributed event to be attributed according to the specific scenario; if attribution is required according to other target events and / or to-be-attributed events, it is necessary to re-write a new algorithm model according to the new rules, and readjust the target event and / or to-be-attributed event. Writing a new algorithm model takes at least two weeks and at most one or two months, which is very inefficient and difficult to meet the ever-changing market rhythm. Moreover, the existing attribution analysis simply determines the contribution of the "to-be-attributed event" to the "target conversion event" based on the time series, lacks the association with the intermediate conversion events, and leads to poor attribution accuracy. In addition, there are problems such as inflexible attribution window setting, limitations in interaction and data display, deficiencies in data display and export functions, users are restricted when conducting in-depth analysis and data utilization, and the user threshold is high, affecting the user experience. Summary of the invention

[0004] To this end, it is necessary to provide a user behavior attribution analysis system and method to solve the problem that the existing attribution analysis algorithm is inflexible and cannot flexibly adjust the attribution analysis algorithm model according to specific business scenarios, and to provide a more efficient, flexible and accurate attribution analysis method.

[0005] To achieve the above object, the present invention provides a user behavior attribution analysis method, which comprises the following steps:

[0006] Collect user behavior event data and construct user behavior paths in time series;

[0007] Receive input attribution analysis parameters, where the attribution analysis parameters include target conversion events, events to be attributed, and types of attribution analysis models, and obtain effective user behavior paths including target conversion events and events to be attributed;

[0008] According to the type of attribution analysis model, calculate the contribution of each event to be attributed in each effective user behavior path;

[0009] The contribution of each event to be attributed in all valid user behavior paths is summed up to obtain the attribution analysis results.

[0010] Furthermore, the user's behavior event data includes click event data, browsing event data, search event data and interaction event data.

[0011] Furthermore, the step of collecting the user's behavior event data to obtain an event data set includes the following steps:

[0012] Collect user behavior event data;

[0013] Pre-process the collected user behavior event data and mark each behavior event data with a timestamp;

[0014] Construct user behavior paths based on time series.

[0015] Furthermore, the step of receiving input attribution analysis parameters, wherein the attribution analysis parameters include target conversion events, events to be attributed, and types of attribution analysis models, and obtaining effective user behavior paths including target conversion events and events to be attributed, includes the following steps:

[0016] Receiving input attribution analysis parameters, wherein the attribution analysis parameters include target conversion events and events to be attributed;

[0017] Based on the target conversion event, traverse the user behavior path associated with the target conversion event;

[0018] The user behavior paths with the events to be attributed are screened out from the user behavior paths associated with the target conversion events, that is, the effective user behavior paths including the target conversion events and the events to be attributed.

[0019] Furthermore, in the step of receiving input attribution analysis parameters, which include target conversion events, events to be attributed and types of attribution analysis models, obtaining effective user behavior paths including target conversion events and events to be attributed also includes setting a window period to obtain effective user behavior paths of target conversion events and events to be attributed within the window period.

[0020] Furthermore, in the step of receiving input attribution analysis parameters, which include target conversion events, events to be attributed and the type of attribution analysis model, obtaining a valid user behavior path including the target conversion events and the events to be attributed, it also includes setting associated events to obtain a valid user behavior path including the target conversion events, events to be attributed and associated events.

[0021] Furthermore, the types of attribution analysis models include first touch point attribution model, last touch point attribution model, linear attribution model, and position attribution model.

[0022] Furthermore, the first-touch attribution model is to determine the contribution ratio of the first event to be attributed when there are multiple events to be attributed in the effective user behavior path that contribute to the same target conversion event, and the remaining events to be attributed share the remaining contribution ratio equally; or the last-touch attribution model is to determine the contribution ratio of the last event to be attributed when there are multiple events to be attributed in the effective user behavior path that contribute to the same target conversion event, and the remaining events to be attributed share the remaining contribution ratio equally; or the linear attribution model is to determine the contribution ratio of the last event to be attributed when there are multiple events to be attributed in the effective user behavior path that contribute to the same target conversion event, and the multiple events to be attributed share the contribution equally; or the position attribution model is to determine the contribution ratio of the first event to be attributed and the last event to be attributed when there are multiple events to be attributed in the effective user behavior path that contribute to the same target conversion event, and the remaining events to be attributed share the remaining contribution ratio equally.

[0023] A user behavior attribution analysis system, applying the above-mentioned user behavior attribution analysis method, comprises:

[0024] The collection module is used to collect user behavior event data and construct user behavior paths in time series;

[0025] An analysis module, used to receive input attribution analysis parameters and match effective user behavior paths according to the attribution analysis parameters;

[0026] The contribution calculation module calculates the contribution of each to-be-attributed event in each effective user behavior path in turn based on the attribution analysis model selected in the attribution analysis parameters;

[0027] The contribution aggregation module is used to aggregate the contributions of all events to be attributed in all valid user behavior paths to obtain attribution analysis results.

[0028] Furthermore, it also includes a visualization module for visualizing input attribution analysis parameters and / or visualizing output attribution analysis results.

[0029] Different from the prior art, the above technical solution mainly sets multiple different types of attribution analysis model types, in which the contribution ratios of attribution events at different positions in each attribution analysis model type are different; so that the most appropriate attribution model can be selected according to different business scenarios and needs; and then the target conversion events, events to be attributed and the type of attribution analysis model to be analyzed are selected according to the business scenarios and needs to improve the efficiency, flexibility and accuracy of the attribution analysis of the contribution of events to be attributed to the target conversion events, and is suitable for e-commerce, digital marketing, website and application user experience optimization and other fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 Schematic diagram of the process of the user behavior attribution analysis method of the embodiment;

[0031] Figure 2 Schematic diagram of the modules of the user behavior attribution analysis system of the embodiment.

[0032] Description of reference numerals:

[0033] 10. Collection module; 20. Event matching module; 30. Contribution calculation module; 40. Contribution aggregation module; 50. Visualization module. DETAILED DESCRIPTION

[0034] In order to explain the technical content, structural features, achieved objectives and effects of the technical solution in detail, the following is a detailed description in conjunction with specific embodiments and accompanying drawings.

[0035] Reference to "embodiment" herein means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present application. The term "embodiment" appearing in various places in the specification does not necessarily refer to the same embodiment, nor does it particularly limit its independence or association with other embodiments. In principle, in the present application, as long as there is no technical contradiction or conflict, the various technical features mentioned in the embodiments can be combined in any way to form a corresponding implementable technical solution.

[0036] Unless otherwise defined, the technical terms used in this document have the same meanings as those generally understood by those skilled in the art to which this application belongs; the use of relevant terms in this document is only for describing specific embodiments and is not intended to limit this application.

[0037] In the description of this application, the term "and / or" is an expression used to describe the logical relationship between objects, indicating that three relationships may exist, for example, A and / or B, which means: A exists, B exists, and A and B exist at the same time. In addition, the character " / " in this article generally indicates that the objects before and after are in an "or" logical relationship.

[0038] In the present application, terms such as “first” and “second” are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship of quantity, priority or sequence between these entities or operations.

[0039] Without further limitations, in this application, the words "include", "comprises", "has" or other similar expressions used in the sentences are intended to cover non-exclusive inclusion. These expressions do not exclude the presence of additional elements in the process, method or product including the elements, so that the process, method or product including a series of elements may include not only those limited elements, but also other elements not explicitly listed, or also include elements inherent to such process, method or product.

[0040] Similar to the understanding in the Patent Examination Guidelines, in this application, expressions such as "greater than", "less than", "exceed" and the like are understood to exclude the number itself; expressions such as "above", "below", "within" and the like are understood to include the number itself. In addition, in the description of the embodiments of this application, "multiple" means more than two (including two), and similar expressions related to "multiple" are also understood in this way, such as "multiple groups", "multiple times", etc., unless otherwise clearly and specifically limited.

[0041] In the description of the embodiments of the present application, space-related expressions used, such as "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "vertical", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc., indicate the orientation or position relationship based on the orientation or position relationship shown in the specific embodiments or drawings, and are only for the convenience of describing the specific embodiments of the present application or facilitating the reader's understanding, and do not indicate or imply that the referred device or component must have a specific position, a specific orientation, or be constructed or operated in a specific orientation, and therefore cannot be understood as a limitation on the embodiments of the present application.

[0042] Unless otherwise expressly specified or limited, in the description of the embodiments of the present application, the terms such as "install", "connect", "connect", "fix", "set", etc. used should be understood in a broad sense. For example, the "connection" can be a fixed connection, a detachable connection, or an integrated setting; it can be a mechanical connection, an electrical connection, or a communication connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium; it can be the internal connection of two elements or the interaction relationship between two elements. For technicians in the technical field to which the present application belongs, the specific meanings of the above terms in the embodiments of the present application can be understood according to the specific circumstances.

[0043] See also Figure 1 As shown, the present invention provides a user behavior attribution analysis method, which belongs to the field of data analysis technology, and solves the problem that the existing attribution analysis algorithm is inflexible and the algorithm model of the attribution analysis cannot be flexibly adjusted according to the specific business scenario; it mainly sets a plurality of different types of attribution analysis model types, wherein the contribution proportions of attribution events at different positions in each attribution analysis model type are different; so that the most suitable attribution model can be selected according to different business scenarios and needs; and then the target conversion event, the event to be attributed and the type of attribution analysis model to be analyzed are selected according to the business scenario and needs to improve the efficiency, flexibility and accuracy of the attribution analysis of the contribution of the event to be attributed to the target conversion event, and is suitable for e-commerce, digital marketing, website and application user experience optimization and other fields.

[0044] The following provides a specific implementation of a user behavior attribution analysis method, which includes the following steps:

[0045] S1. Collect user behavior event data and construct user behavior paths in time series;

[0046] S2. receiving input attribution analysis parameters, wherein the attribution analysis parameters include target conversion events, events to be attributed, and types of attribution analysis models, and obtaining effective user behavior paths including target conversion events and events to be attributed;

[0047] S3. Calculate the contribution of each to-be-attributed event in each effective user behavior path according to the type of attribution analysis model;

[0048] S4. The contribution of each event to be attributed in all valid user behavior paths is summed up to obtain the attribution analysis result.

[0049] The above-mentioned collection of user behavior event data can be performed through a tracking script or API integrated in an application (app) or website application. The above-mentioned user behavior events include but are not limited to click events, browsing events, search events and interactive events, and the above-mentioned user behavior event data includes but is not limited to click event data, browsing event data, search event data and interactive event data. Among the attribution analysis parameters received as input, the input attribution analysis parameters can be manually selected and set by a data analyst, and the input method can include instructions, visualization and other methods. The above-mentioned target conversion event is one of the user's behavior events, usually a click event, an interactive event, etc., indicating that the user completes a certain business (such as adding to a purchase, placing an order); the above-mentioned to-be-attributed event is one of the user's behavior events, which is used to set the event that needs to be analyzed for the contribution to the "target conversion event", generally a click event (such as an ad position click), a browsing event, a search event and other user behavior events. In the effective user behavior path, the to-be-attributed event is located before the target conversion event. The above-mentioned attribution analysis results may include but are not limited to the name of the to-be-attributed behavior, the contribution of the to-be-attributed behavior to the target conversion event, etc. The attribution analysis results obtained above can be displayed in a visual way to detail the attribution analysis results. The visual display can be realized by selecting a suitable graph or table or a combination of graphs. For example, a bar graph or pie chart can be used to compare the contribution of different factors to the results; a line graph or area graph can be used to show the trend over time; and a table can be used to show the number of users and conversion rate at each stage of the attribution node in detail.

[0050] The user behavior event data collected in step S1 can be directly used to construct the user behavior path in time series, or the collected user behavior event data can be preprocessed and then the user behavior path can be constructed in time series. That is, the step of collecting the user behavior event data to obtain the event data set includes the following steps:

[0051] Collect user behavior event data;

[0052] Pre-process the collected user behavior event data and mark each behavior event data with a timestamp;

[0053] Construct user behavior paths based on time series.

[0054] The above preprocessing of the collected user behavior event data makes the data structured and standardized. The above preprocessing may include data cleaning, data conversion, feature selection, etc., wherein data cleaning may remove invalid or incomplete data, including deleting duplicate records, processing missing values, processing abnormal values, etc. It may be implemented using data processing tools, such as drop_duplicates(), fillna(), etc.

[0055] Mark each behavioral event data with a timestamp for subsequent time series analysis. Add a timestamp to each behavioral event data based on the time when the behavioral event data occurred. The timestamp can be a specific date and time, or a relative time (such as the time interval relative to a certain event). You can use the functions provided by the data processing tool, such as assign(), transform(), and timestamp calculation formulas to add timestamps to behavioral event data.

[0056] In the above-mentioned construction of the user behavior path based on the time series, the user behavior path is a complete behavior sequence of the user completing the target conversion event, which includes the target conversion event and at least one event to be attributed, and of course may also include associated events, etc.

[0057] In step S2, the attribution analysis parameters received as input include the target conversion event, the event to be attributed, and the type of the attribution analysis model, and the step of obtaining the effective user behavior path including the target conversion event and the event to be attributed can directly filter out the effective user behavior path with the target conversion event and the event to be attributed from all user behavior paths, or first filter out a certain event to be attributed, and then traverse forward (if the event to be attributed is not the first event to be attributed) to associate the event to be attributed and traverse backward to associate the event to be attributed (if the event to be attributed is not the last event to be attributed) and the user behavior path of the target conversion event; or first filter out the target conversion event, and then traverse forward to associate the target conversion event. Taking the last embodiment as an example, it specifically includes the following steps:

[0058] Receiving input attribution analysis parameters, wherein the attribution analysis parameters include target conversion events and events to be attributed;

[0059] Based on the target conversion event, traverse the user behavior path associated with the target conversion event in the event data set;

[0060] The user behavior paths with the events to be attributed are screened out from the user behavior paths associated with the target conversion events, that is, the effective user behavior paths including the target conversion events and the events to be attributed.

[0061] Step S2 can also support dynamic adjustment of the window period, that is, in the step of receiving the input attribution analysis parameters, the attribution analysis parameters include the target conversion event, the event to be attributed and the type of the attribution analysis model, and obtaining the effective user behavior path including the target conversion event and the event to be attributed, it also includes setting the window period, and obtaining the effective user behavior path of the target conversion event and the event to be attributed within the window period. The above-mentioned window period refers to a specific time period set in the funnel analysis, which is used to limit the total time for the user to complete each event in the funnel. It can be dynamically adjusted according to the analysis requirements, and the window period for different attribution analyses can be set differently, such as a few minutes, hours or days, etc., which is suitable for different types of user behavior analysis. So that the analysis is not limited to the static funnel model, but can be dynamically adjusted according to the actual situation, so as to better understand and analyze the user's behavior pattern within a specific time.

[0062] In step S2, in order to increase the direct correlation between the event to be attributed and the target conversion event and make the calculation of contribution attribution more accurate, the relevant attributes of the associated event can be input by adding associated events. The above-mentioned associated event belongs to any one of the user's behavioral events, specifically an event that must be triggered before completing the target conversion event, to ensure that the event to be attributed and the target conversion event are related. In the effective user behavior path, it can be located before or after the event to be attributed. The event before the event to be attributed is called a forward associated event, and the event after the event to be attributed is called a backward associated event. The relevant attributes of the associated event include but are not limited to the name of the user's behavioral event, the user registration channel, the target conversion event and the associated event are the same product, etc. Specifically, in receiving the input attribution analysis parameters, the attribution analysis parameters include the target conversion event, the event to be attributed and the type of the attribution analysis model, and the step of obtaining the effective user behavior path including the target conversion event and the event to be attributed, it also includes setting the associated event to obtain the effective user behavior path including the target conversion event, the event to be attributed and the associated event.

[0063] The attribution analysis model in step S2 determines the contribution of the events to be attributed at different locations based on rules, algorithms or statistics. Common attribution models include linear attribution models, time decay models, proportion attribution models, etc. A suitable attribution model can be selected according to actual conditions. In some embodiments, types including the first contact attribution model, the last contact attribution model, the linear attribution model, the position attribution model, etc. are provided for data analysts to choose.

[0064] The first touch point attribution model is to determine the contribution ratio of the first event to be attributed when there are multiple events to be attributed in the effective user behavior path that contribute to the same target conversion event, and the remaining contribution ratios are equally shared by the remaining events to be attributed; if it can be determined that the contribution ratio of the first event to be attributed is 100%, the remaining contribution ratios are equally shared by the remaining events to be attributed, which is 0.

[0065] The last-touch attribution model is to determine the contribution ratio of the last event to be attributed when there are multiple events to be attributed in the effective user behavior path that contribute to the same target conversion event, and the remaining contribution ratios are evenly divided among the remaining events to be attributed; if it can be determined that the contribution ratio of the last event to be attributed is 100%, the remaining contribution ratios are evenly divided among the remaining events to be attributed, which is 0.

[0066] The linear attribution model is that when there are multiple to-be-attributed events in the effective user behavior path that contribute to the same target conversion event, the multiple to-be-attributed events evenly distribute the contribution;

[0067] The position attribution model is that when there are multiple to-be-attributed events in the effective user behavior path that contribute to the same target conversion event, the contribution ratio of the first to-be-attributed event and the last to-be-attributed event is determined, and the remaining contribution ratio is evenly divided among the remaining to-be-attributed events. For example, if the contribution ratio of the first to-be-attributed event and the last to-be-attributed event is determined to be 40% each, the remaining contribution ratio is evenly divided among the remaining to-be-attributed events, which is 20%.

[0068] In order to further understand the present invention, the following specific embodiments are provided.

[0069] The following provides the user behavior event data collected on a certain day, and constructs the user behavior path in time series:

[0070] User behavior path 1: Home page → G1 → Home page → G2 → Scenario page 1 → ST1 → Product details page 1 → S1 → Shopping cart page → C1 → O1 → P1;

[0071] User behavior path 2: Homepage → G3 → Scenario page 2 → ST2 → SG4 → Scenario page 3 → ST3 → Product details page 3 → S3 → Shopping cart page → C2 → O2 → P2;

[0072] User behavior path 3: Home page → G2 → Scenario page 2 → ST4 → Business details page 5 → S5 → Business details page 6 → S6 → Shopping cart page → C3 → O3 → P3;

[0073] User behavior path 4: Home page → G3 → Scenario page 2 → ST4 → Business details page 5 → S5 → Business details page 6 → S6 → Shopping cart page → C4 → O4 → P4;

[0074] G stands for click on home page ad slot, G1 stands for click on home page ad slot 1, G2 stands for click on home page ad slot 2, and G3 stands for click on home page ad slot 3;

[0075] ST represents the click of the product card on the scene page. ST1 is the click of the product card on scene page 1, ST2 is the click of the product card on scene page 2, ST3 is the click of the product card on scene page 3; ST4 is the click of the product card on scene page 4;

[0076] SG represents the scene page click ad slot, SG4 represents the scene page 2 click ad slot 4;

[0077] S represents the purchase of items on the product detail page. S1 is the purchase of item 1 on the product detail page; S3 is the purchase of item 6 on the product detail page; S5 is the purchase of item 5 on the product detail page; S6 is the purchase of item 6 on the product detail page;

[0078] C represents the shopping cart to order items. C1 is the shopping cart to order item 1, C2 is the shopping cart to order item 2 + item 3, C3 is the shopping cart to order item 6, and C4 is the shopping cart to order item 9.

[0079] O represents the product order event, O1 is the product 1 order event, O2 is the product 2 + product 3 order event, O3 is the product 6 order event; O4 is the product 9 order event;

[0080] P stands for payment.

[0081] Enter the parameters to be attributed:

[0082] The events to be attributed are: G click on the ad space, ST click on the product card, S add to cart on the product detail page;

[0083] Window period: the same day;

[0084] Target conversion time: O product order event;

[0085] Forward association event: ST product card click + associated attribute (product id) or S product detail page add to cart + associated attribute (product id)

[0086] Traverse the user paths within the window period from the set "target conversion event" (order placement) to filter out the paths that contain "forward association events" (ST product card click or S product detail page purchase) and meet the association attribute conditions (the target conversion event and the forward association event are the same product).

[0087] According to the above set rules, it can be seen that among the above four user behavior paths, user behavior paths 1, 2, and 3 are valid user behavior paths, and the forward association event association attribute condition of user behavior path 4 is not met.

[0088] It can be seen that the unattributed events that contribute to user behavior path 1 are G1, G2, and ST1; the unattributed events that contribute to user behavior path 2 are G3, ST2, SG4, and ST3; the unattributed events that contribute to user behavior path 3 are G2, ST4, and S5;

[0089] Select the first touchpoint attribution model to calculate the contribution of each event to be attributed in each effective user behavior path:

[0090] The contribution of the first event to be attributed is determined to be 100%, and the remaining contributions of the remaining events to be attributed are equally divided and are 0.

[0091] Then the contribution allocation result of the unattributed event of user behavior path 1 is: G1=1, the contribution allocation result of the unattributed event of user behavior path 2 is: G3=1, and the contribution allocation result of the unattributed event of user behavior path 3 is: G2=1;

[0092] The total contribution of each event to be attributed in all valid user behavior paths is summed up to obtain the attribution analysis result:

[0093] Event Name Ad slot name Total goal conversions Contribution Ad click G1 1 33.33% G2 1 33.33% G3 1 33.33%

[0094] Select the last touch attribution model to calculate the contribution of each event to be attributed in each effective user behavior path:

[0095] The contribution of the last event to be attributed is determined to be 100%, and the remaining contributions of the remaining events to be attributed are equally divided to be 0.

[0096] Then the contribution allocation result of the unattributed event of user behavior path 1 is: ST1=1, the contribution allocation result of the unattributed event of user behavior path 2 is: ST3=1, and the contribution allocation result of the unattributed event of user behavior path 3 is: S5=1;

[0097] The total contribution of each event to be attributed in all valid user behavior paths is summed up to obtain the attribution analysis result:

[0098]

[0099]

[0100] Select the linear attribution model to calculate the contribution of each event to be attributed in each effective user behavior path:

[0101] Multiple events to be attributed share this contribution evenly.

[0102] Then the contribution distribution result of the unattributed events in user behavior path 1 is: G1, G2, ST1 = 0.33, the contribution distribution result of the unattributed events in user behavior path 2 is: G3, ST2, SG4, ST3 = 0.25, and the contribution distribution result of the unattributed events in user behavior path 3 is: G2, ST4, S5 = 0.33;

[0103] The total contribution of each event to be attributed in all valid user behavior paths is summed up to obtain the attribution analysis result:

[0104] Event Name Ad slot name List name Total goal conversions Contribution Ad click G1 - 0.33 11% G2 - 0.66 22% G3 - 0.25 8.33% SG4 - 0.25 8.33% Product card click - ST1 0.33 11% - ST2 0.25 8.33% - ST3 0.25 8.33% - ST4 0.33 11% - S5 0.33 11%

[0105] Select the position attribution model to calculate the contribution of each event to be attributed in each effective user behavior path:

[0106] It is determined that the contribution of the first event to be attributed and the last event to be attributed is 40% each, and the remaining events to be attributed share the remaining contribution of 20% in equal parts.

[0107] Then the contribution distribution result of the unattributed events in user behavior path 1 is: G1=0.4, G2=0.2, ST1=0.4, the contribution distribution result of the unattributed events in user behavior path 2 is: G3=0.4, ST2=0.1, SG4=0.1, ST3=0.4, the contribution distribution result of the unattributed events in user behavior path 3 is: G2=0.4, ST4=0.2, S5=0.4;

[0108] The total contribution of each event to be attributed in all valid user behavior paths is summed up to obtain the attribution analysis result:

[0109] Event Name Ad slot name List name Total goal conversions Contribution Ad click G1 - 0.4 13.33% G2 - 0.6 20% G3 - 0.4 13.33% SG4 - 0.1 3.33% Product card click - ST1 0.4 13.33% - ST2 0.1 3.33% - ST3 0.4 13.33% - ST4 0.2 6.66% - S5 0.4 13.33%

[0110] See also Figure 2 As shown, the present invention also provides a user behavior attribution analysis system, which includes a collection module 10, which is used to collect user behavior event data to obtain a data set;

[0111] An event matching module 20, configured to receive input attribution analysis parameters and match effective user behavior paths according to the attribution analysis parameters;

[0112] Contribution calculation module 30, based on the attribution analysis model selected in the attribution analysis parameters, sequentially calculates the contribution of each to-be-attributed event in each effective user behavior path;

[0113] The contribution aggregation module 40 is used to aggregate the contribution of each event to be attributed in all valid user behavior paths to obtain an attribution analysis result.

[0114] The above-mentioned user behavior attribution analysis system may also include a visualization module 50 for visualizing the input attribution analysis parameters and / or visualizing the output attribution analysis results. In some embodiments, the visualization module 50 is used to visualize the input attribution analysis parameters, provide a friendly graphical interface, and allow data analysts to customize the target conversion events and events to be attributed through intuitive operations such as dragging and dropping, selecting, etc. The interface supports complex event combinations and nesting, so that the attribution calculation can accurately reflect the complex user behavior path; in some embodiments, the visualization module 50 is used to visualize the output attribution analysis results, and selects a suitable combination of graphs or tables or charts to achieve instant data analysis and result feedback. In some embodiments, the visualization module 50 is used to visualize the input attribution analysis parameters and visualize the output attribution analysis results, so that it has the advantages of the above-mentioned embodiments.

[0115] It should be noted that, although the above embodiments have been described in this article, the patent protection scope of the present invention is not limited thereby. Therefore, based on the innovative concept of the present invention, changes and modifications made to the embodiments described herein, or equivalent structures or equivalent process changes made using the contents of the present invention specification and drawings, directly or indirectly applying the above technical solutions to other related technical fields, are all included in the patent protection scope of the present invention.

Claims

1. A user behavior attribution analysis method, characterized in that: The following steps are involved: Collect user behavior event data and construct user behavior paths in time series; Receive input attribution analysis parameters, where the attribution analysis parameters include target conversion events, events to be attributed, and types of attribution analysis models, and obtain effective user behavior paths including target conversion events and events to be attributed; According to the type of attribution analysis model, calculate the contribution of each event to be attributed in each effective user behavior path; The contribution of each event to be attributed in all valid user behavior paths is summed up to obtain the attribution analysis results.

2. The user behavior attribution analysis system and method according to claim 1, characterized in that: The user's behavior event data includes click event data, browsing event data, search event data and interaction event data.

3. The user behavior attribution analysis system and method according to claim 1, characterized in that: The step of collecting the user's behavior event data to obtain an event data set includes the following steps: Collect user behavior event data; Pre-process the collected user behavior event data and mark each behavior event data with a timestamp; Construct user behavior paths based on time series.

4. The user behavior attribution analysis method according to claim 1, characterized in that: The step of receiving input attribution analysis parameters, wherein the attribution analysis parameters include target conversion events, events to be attributed, and types of attribution analysis models, and obtaining effective user behavior paths including target conversion events and events to be attributed, comprises the following steps: Receiving input attribution analysis parameters, wherein the attribution analysis parameters include target conversion events and events to be attributed; Based on the target conversion event, traverse the user behavior path associated with the target conversion event; The user behavior paths with the events to be attributed are screened out from the user behavior paths associated with the target conversion events, that is, the effective user behavior paths including the target conversion events and the events to be attributed.

5. The user behavior attribution analysis method according to claim 1, characterized in that: In the step of receiving input attribution analysis parameters, the attribution analysis parameters include the target conversion event, the event to be attributed and the type of the attribution analysis model, obtaining the effective user behavior path including the target conversion event and the event to be attributed, also includes setting a window period, and obtaining the effective user behavior path of the target conversion event and the event to be attributed within the window period.

6. A user behavior attribution analysis method according to claim 1, characterized in that: In the step of receiving input attribution analysis parameters, the attribution analysis parameters include target conversion events, events to be attributed and the type of attribution analysis model, obtaining a valid user behavior path including the target conversion events and the events to be attributed, it also includes setting associated events to obtain a valid user behavior path including the target conversion events, events to be attributed and associated events.

7. The user behavior attribution analysis method according to claim 1, characterized in that: The types of attribution analysis models include first touch attribution model, last touch attribution model, linear attribution model, and position attribution model.

8. The user behavior attribution analysis method according to claim 1, characterized in that: The first touch attribution model is that when there are multiple events to be attributed in the effective user behavior path that contribute to the same target conversion event, the contribution ratio of the first event to be attributed is determined, and the remaining contribution ratios are evenly divided among the remaining events to be attributed; or the last touch attribution model is that when there are multiple events to be attributed in the effective user behavior path that contribute to the same target conversion event, the contribution ratio of the last event to be attributed is determined, and the remaining contribution ratios are evenly divided among the remaining events to be attributed; or the linear attribution model is that when there are multiple events to be attributed in the effective user behavior path that contribute to the same target conversion event, the multiple events to be attributed evenly distribute the contribution; Or the position attribution model is such that when there are multiple events to be attributed in the effective user behavior path that contribute to the same target conversion event, the contribution ratios of the first event to be attributed and the last event to be attributed are determined, and the remaining contribution ratios are equally divided among the remaining events to be attributed.

9. A user behavior attribution analysis system, applying the user behavior attribution analysis method according to any one of claims 1 to 8, characterized in that: include The collection module is used to collect user behavior event data and construct user behavior paths in time series; An analysis module, used to receive input attribution analysis parameters and match effective user behavior paths according to the attribution analysis parameters; The contribution calculation module calculates the contribution of each to-be-attributed event in each effective user behavior path in turn based on the attribution analysis model selected in the attribution analysis parameters; The contribution aggregation module is used to aggregate the contributions of all events to be attributed in all valid user behavior paths to obtain attribution analysis results.

10. The user behavior attribution analysis system according to claim 9, characterized in that: The method further includes a visualization module for visualizing input attribution analysis parameters and / or visualizing output attribution analysis results.

Citation Information

Cited By

  • Accurate user traceability attribution method and system oriented to high throughput

    CN121391357A

  • Short video traffic data analysis method and system

    CN121418614A

  • AI-driven intelligent analysis system and method for user session data

    CN122390792A