Method and system for measuring advertising effect
By configuring Hbase and Druid storage structures, combining message queues and real-time log processing, the problem of difficulty in supporting fine-grained analysis in the existing technology is solved, and efficient advertising performance evaluation and multi-source conversion funnel analysis are achieved, which improves analysis Timeliness.
Patent Information
- Application Number
- CN202110170697.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-02-08
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2041-02-08
AI Technical Summary
When evaluating the effectiveness of advertising promotion, existing user conversion funnel models are difficult to support fine-grained analysis, such as attribution analysis by region, device type and designated population, and in the face of big data, analysis efficiency and poor timeliness.
By configuring Hbase and Druid storage structures, combining message queues and real-time log processing, a user conversion funnel model for multi-source and variable window periods is realized, supporting source statistics, population statistics and variable window period query, improving data processing and Query efficiency.
Provides calculation, storage and rapid query solutions for multi-source user conversion funnel analysis, reducing the query pressure of unnecessary data, improving user ID matching efficiency, real-time multi-source conversion funnel analysis and designated population analysis, and improving advertising effectiveness Timeliness of assessment.
Smart Images

Figure CN112837101B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data analysis technology. More specifically, the present invention mainly relates to a method and system for measuring advertising effect. Background Art
[0002] With the development of e-commerce in recent years, advertisements can be seen everywhere in life. As a form of media, advertisements play a unique and irreplaceable role in real life. Whether walking on the street or watching TV at home, reading newspapers or browsing the Internet, you will find advertisements everywhere. In this era of highly developed communication technology, advertisements have been everywhere, and various forms of advertisements such as TV advertisements, radio advertisements, newspaper advertisements, signboard advertisements, and online advertisements coexist. Advertisements have a huge impact on social culture and people's concepts and behaviors.
[0003] The user conversion funnel model is an effective tool for evaluating the effectiveness of advertising promotions, especially the user conversion funnel model that can be divided into sources, which can intuitively display the effectiveness of advertising promotions. The existing user conversion funnel model mainly attributes according to the promotion channel, with a fixed conversion window period, and poor support for user conversion attribution at a finer granularity. For example, it does not support analysis by user attributes such as region, IP, and device type, nor does it support attribution analysis by specified population. In addition, the existing solution must start with the calculation of the original detailed data for each query. Faced with a large amount of data, the analysis efficiency is too low and the timeliness is too poor. Summary of the invention
[0004] The embodiment of the present application provides a method for measuring advertising effectiveness, so as to at least solve the problem of subjective factors in the related art.
[0005] The present invention provides a method for measuring advertising effect, the method comprising the following steps:
[0006] Configuration steps: Configure storage structure;
[0007] Acquisition steps: After configuration is completed, obtain the conversion funnel table;
[0008] Query step: Based on the conversion funnel table, query the funnel data and evaluate the advertising effect.
[0009] In the above-mentioned method for measuring the advertising effect, the configuration step specifically includes the following steps:
[0010] Funnel configuration steps: Configure funnel storage;
[0011] Hbase configuration steps: configure Hbase storage structure;
[0012] Druid configuration steps: Configure the Druid storage structure.
[0013] In the above-mentioned method for measuring the advertising effect, the obtaining step specifically includes the following steps:
[0014] Log pulling step: pull real-time logs from the message queue;
[0015] Data processing step: performing data processing on the real-time log;
[0016] Status query step: extracting the visitor ID in the real-time log after data processing, querying the conversion funnel status corresponding to the visitor ID in the Hbase, and deserializing it into a local temporary cache object;
[0017] Judgment step: grouping the real-time logs according to the visitor IDs, and judging the real-time logs one by one in each group;
[0018] Deletion step: after traversal, if the conversion funnel state of the visitor ID has not been modified or newly created, deleting the data corresponding to the visitor ID from the local temporary cache object;
[0019] Writing step: serializing all data in the local temporary cache object and writing them to the Hbase;
[0020] Data reading step: reading data from the message queue in real time to obtain the conversion funnel table.
[0021] In the above-mentioned method for measuring the advertising effect, the data processing step specifically includes the following steps:
[0022] Cleaning step: performing data cleaning on the real-time log;
[0023] Filtering step: Filter abnormal logs.
[0024] In the above-mentioned method for measuring the advertising effect, the judging step specifically includes the following steps:
[0025] The first judgment step: judging whether the real-time log will update the current funnel state in the local temporary cache object, and if so, writing the latest state into the message queue;
[0026] The second judgment step: judging whether the real-time log will trigger a new funnel calculation, and if so, creating a conversion funnel state object with the real-time log as the trigger log, and updating it to the local temporary cache object.
[0027] The above-mentioned method for measuring advertising effectiveness, wherein the query scheme for funnel data includes but is not limited to: statistical query by source, statistical query by population, and variable window period query.
[0028] Based on the same inventive concept, the present invention also discloses a system for measuring advertising effects based on the advertising effect measurement method disclosed by any invention.
[0029] The advertising effectiveness measurement system includes:
[0030] Configuration module, configuration storage structure;
[0031] Get the module. After configuration is completed, get the conversion funnel table.
[0032] The query module queries the funnel data based on the conversion funnel table and evaluates the advertising effect.
[0033] In the above-mentioned advertising effect measurement system, the configuration module specifically includes:
[0034] Funnel configuration unit, configures the funnel storage;
[0035] Hbase configuration unit, configures Hbase storage structure;
[0036] Druid configuration unit, configures the Druid storage structure.
[0037] In the above-mentioned advertising effect measurement system, the acquisition module specifically includes:
[0038] Log pulling unit, pulling real-time logs from the message queue;
[0039] A data processing unit, performing data processing on the real-time log;
[0040] A status query unit, which extracts the visitor ID in the real-time log after data processing, queries the conversion funnel status corresponding to the visitor ID in the Hbase, and deserializes it into a local temporary cache object;
[0041] A judgment unit, grouping the real-time logs according to the visitor ID, and judging the real-time logs one by one in each group;
[0042] A deleting unit, after traversal, if the conversion funnel state of the visitor ID has not been modified or newly created, deletes the data corresponding to the visitor ID from the local temporary cache object;
[0043] A writing unit, serializing all data in the local temporary cache object and writing them to the Hbase;
[0044] The data reading unit reads data from the message queue in real time to obtain the conversion funnel table.
[0045] In the above-mentioned advertising effect measurement system, the data processing unit specifically includes:
[0046] A cleaning unit, performing data cleaning on the real-time log;
[0047] Filter unit, filters exception logs.
[0048] Compared with the prior art, the present invention has the following beneficial effects:
[0049] 1. A method for measuring advertising effectiveness is proposed. According to the actual needs of customers, a user conversion funnel model with multiple sources and matching specified groups is provided to evaluate advertising effectiveness. The calculation, storage and fast query solutions of user conversion funnel analysis are also provided.
[0050] 2. By calculating and storing key data for conversion funnel queries, a large amount of data that does not need to be involved in conversion funnel queries is filtered out, effectively reducing storage and query pressure;
[0051] 3. Through the aggregation and bitmap indexing functions of the database, the matching efficiency of user IDs is greatly improved, thereby improving the timeliness of user conversion funnel data;
[0052] 4. Through the special table structure design, space is exchanged for time, and the funnel window period can be flexibly adjusted and quickly queried. It can easily complete the real-time multi-source conversion funnel analysis and the conversion funnel analysis that can match the specified population;
[0053] 5. Provides richer multi-source user conversion funnel analysis methods, provides user conversion funnel analysis methods for custom groups, and provides more efficient (real-time) user conversion funnel architecture, storage, calculation and query solutions;
[0054] 6. It has rich application scenarios and can be used to calculate other user conversion funnels, including user conversion funnel analysis scenarios for online shopping malls, educational websites, and other websites.
[0055] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS
[0056] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0057] Figure 1 This is an overall flow chart of a method for measuring advertising effectiveness provided by an embodiment of the present invention;
[0058] Figure 2 It is the overall architecture diagram of the technical solution disclosed in this embodiment;
[0059] Figure 3 yes Figure 1 The disclosed step S1 overall flow chart;
[0060] Figure 4 yes Figure 1 The overall flow chart of step S2 disclosed;
[0061] Figure 5 is a flow chart of the calculation scheme disclosed in this embodiment;
[0062] Figure 6 yes Figure 4 The overall flow chart of step S22 disclosed;
[0063] Figure 7 yes Figure 4 The overall flow chart of step S24 disclosed;
[0064] Figure 8 This is a structural framework diagram of an advertising effectiveness measurement system provided in this embodiment.
[0065] In the above picture:
[0066] 1. Configuration module; 2. Acquisition module; 3. Query module; 11. Funnel configuration unit; 12. Hbase configuration unit; 13. Druid configuration unit; 21. Log pulling unit; 22. Data processing unit; 23. Status query unit; 24. Judgment unit; 25. Deletion unit; 26. Writing unit; 27. Data reading unit; 221. Cleaning unit; 222. Filtering unit; 241. First judgment unit; 242. Second judgment unit. DETAILED DESCRIPTION
[0067] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0068] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a kind of", "the" and the like involved in this application do not indicate a quantitative limitation, and may represent the singular or plural. The terms "include", "comprise", "have" and any of their variations involved in this application are intended to cover non-exclusive inclusions. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are merely to distinguish similar objects, without waiting for the specific ordering of the objects.
[0069] The present invention is described in detail below in conjunction with the various embodiments shown in the accompanying drawings, but it should be noted that these embodiments are not limitations of the present invention, and any equivalent transformations or substitutions in functions, methods, or structures made by ordinary technicians in the field based on these embodiments are all within the scope of protection of the present invention.
[0070] Before elaborating on each embodiment of the present invention in detail, the core inventive concept of the present invention is summarized and elaborated in detail through the following several embodiments.
[0071] The present invention can evaluate the advertising effect based on the user conversion funnel model, and provide the calculation, storage and fast query solutions of the user conversion funnel analysis.
[0072] Embodiment 1:
[0073] Reference Figures 1 to 7 As shown, this example discloses a specific implementation method of a method for measuring advertising effectiveness (hereinafter referred to as "method").
[0074] Specifically, the overall architecture of the method shown is first introduced. The original log to be calculated is pulled from the message queue kafka, the intermediate data is stored in the Hbase database, the final result is stored in the Druid database, and the calculation component selects the spark-streaming component (log real-time calculation component) to complete the real-time calculation and warehousing of the real-time log. It should be noted that the database used in the method can be replaced by other databases of the same type, such as Mysql can be replaced by SQLserver, Hbase can be replaced by redis, Druid can be replaced by clickhouse, and the log real-time calculation component spark-streaming component can be replaced by storm or flink, etc., and the message queue kafka can be replaced by message queue components such as RabbitMQ, and the present invention is not limited to this.
[0075] Specifically, refer to Figure 1 and Figure 2 As shown, the method disclosed in this embodiment includes the following steps:
[0076] Step S1, configuring the storage structure.
[0077] Specifically, in some of the embodiments, reference is made to Figure 3 As shown, the step S1 specifically includes the following steps:
[0078] S11, configure funnel storage;
[0079] S12. Configure Hbase storage structure;
[0080] S13. Configure the Druid storage structure.
[0081] Specifically, in some of the embodiments, some of the fields involved in step S1 are explained as follows:
[0082] Field Name type explain visitorId String Visitor ID funnelId String Funnel ID reffer String Source information, which can be expanded into multiple fields beginDate long Funnel first step trigger date lastStep long Funnel previous state lastStepDate datetime Funnel previous step trigger date
[0083] Specifically, in some of the embodiments, when designing the funnel configuration storage, each time a user creates a new funnel, a unique funnel ID needs to be generated, and the funnel ID, conversion cycle and other information need to be stored as a record in Mysql. The specific configuration details are adjusted according to needs or scenarios.
[0084] Specifically, in some of the embodiments, Hbase plays the role of storing intermediate results, which include the user ID and the status of the conversion funnel calculation in which the user participates. The fields are designed as follows:
[0085]
[0086] Among them, FunnelReffer is a funnel source class composed of multiple source fields. LastStep refers to the Nth step of the funnel created by the user (reducing the storage volume). The specific information corresponding to this step must be matched in the Mysql database. The rowKey is visitorId, and the value is the latest status of the corresponding funnel under different funnel IDs in which the user participates. Each funnel ID stores the status of funnels with different start dates or different sources.
[0087] Specifically, in some of the embodiments, the final calculation results of the conversion funnel will be stored in Druid for query analysis. Since the query granularity is day, Druid's storage segmentation strategy is segmented by day and partitioned by the hash value of funnelId. In order to meet the needs of customers to flexibly select analysis time periods and population matching packages, Druid's table structure is designed as follows:
[0088]
[0089] Then execute step S2. After the configuration is completed, obtain the conversion funnel table.
[0090] Specifically, in some of the embodiments, reference is made to Figure 4 and Figure 5 As shown, step S2 specifically includes the following steps:
[0091] S21, pull real-time logs from the message queue;
[0092] S22, performing data processing on the real-time log;
[0093] S23, extracting the visitor ID in the real-time log after data processing, querying the conversion funnel state corresponding to the visitor ID in the Hbase, and deserializing it into a local temporary cache object;
[0094] S24, grouping the real-time logs according to the visitor IDs, and judging the real-time logs one by one in each group;
[0095] S25. After the traversal is completed, if the conversion funnel state of the visitor ID has not been modified or newly created, deleting the data corresponding to the visitor ID from the local temporary cache object;
[0096] S26, serializing all data in the local temporary cache object and writing them to the Hbase;
[0097] S27. Read data from the message queue in real time to obtain the conversion funnel table.
[0098] Specifically, in some of the embodiments, reference is made to Figure 6 As shown, the step S22 specifically includes the following steps:
[0099] S221, performing data cleaning on the real-time log;
[0100] S222. Filter abnormal logs.
[0101] Specifically, in some of the embodiments, reference is made to Figure 7 As shown, the step S24 specifically includes the following steps:
[0102] S241, determining whether the real-time log will update the current funnel state in the local temporary cache object, and if so, writing the latest state into the message queue;
[0103] S242: Determine whether the real-time log will trigger a new funnel calculation. If so, create a conversion funnel state object using the real-time log as a trigger log, and update the object to the local temporary cache.
[0104] Specifically, the overall calculation process is as follows: the real-time calculation program pulls a batch of real-time logs from Kafka. First, the logs must be cleaned and abnormal logs must be filtered out. Then, all visitorIds (visitor IDs) in the current batch are taken out to form a rowKeyList (rowkey, i.e., row key, is required to read data from Hbase. Multiple rowkeys form a rowKeyList, which can improve the average query efficiency). Then, the conversion funnel status corresponding to these visitorIds is queried in Hbase, and deserialized into a local temporary cache object (a mapping relationship with rowKey (i.e., visitorId) as key and multiple funnel status as value); then the logs are grouped by visitorId, and each log in each group is judged. First, it is judged whether the current log will update the current funnel status in the local temporary cache. If it will be updated, the latest status is written to Kafka; secondly, according to the data read from Mysql (synchronized to broadcast variables by subscription) The funnel configuration determines whether the current log will trigger a new funnel calculation. If so, a conversion funnel state object is created with the current log as the trigger log and updated to the local temporary cache object. After traversing the logs of each visitorId group, if the conversion funnel state of the visitorId has not been modified or created, the corresponding key and value are deleted from the local temporary cache. After calculating the current batch of data, all data in the local temporary cache object (if there is no data, it does not need to be written back) are serialized and written to Hbase. Finally, Druid reads each data record from Kafka in real time and stores it in the conversion funnel table for subsequent queries.
[0105] Then, step S3 is executed to query the funnel data based on the conversion funnel table and evaluate the advertising effect.
[0106] Specifically, in some embodiments, the query schemes for the funnel data include but are not limited to: source-based statistical query, population-based statistical query, and variable window period query. For example, the source can be divided into advertising activities, IP segments, media, social categories, etc., and the population can be divided into retained users, new users, special groups (students, office workers), etc.
[0107] Specifically, taking the statistics of the achievement of each step in the conversion funnel from 20200701 to 20200707 (the overall data query range is automatically extended backward by one window period by default, and the first step of the funnel must be triggered within the date range of 20200701 to 20200707. Here, assuming that the default window period is 3 days, the actual query period is from 20200701 to 20200710) as an example, the query solution is explained in three ways:
[0108] 1. Statistics by source. Here we assume that the source is 'reffer'. In the actual production environment, there may be multiple fields. You can connect them with the and operator. As follows:
[0109] select count(distinct(visitorId))as uv,lastStep from (selectvisitorId,beginDate,max(lastStep)as lastStep from tb where beginDate>=20200701 and beginDate<=20200707 and lastStepDate<= 20200710 andTIMESTAMPDIFF(DAY,beginDate,lastStepDate)<=3 and funnelId='fid'and reffer='a'groupby beginDate,visitorId)ta group by lastStep
[0110] 2. Count by population. Here, the population refers to multiple specific visitorIds, which are replaced by the "..." symbol in SQL. As follows:
[0111] select count(distinct(visitorId))as uv,lastStep from (selectvisitorId,beginDate,max(lastStep)as lastStep from tb where beginDate>=20200701 and beginDate<=20200707 and lastStepDate<= 20200710 andTIMESTAMPDIFF(DAY,beginDate,lastStepDate)<=3 and funnelId='fid'andvisitorId in(...)group by beginDate,visitorId)ta group by lastStep
[0112] 3. Variable window period query, the maximum window period is 180 days, and it can be adjusted at will within 180 days. Here, it is assumed that the window period is set to 7 days. As follows:
[0113] select count(distinct(visitorId))as uv,lastStep from (selectvisitorId,beginDate,max(lastStep)as lastStep from tb where beginDate>=20200701 and beginDate<=20200707 and lastStepDate<= 20200714 andTIMESTAMPDIFF(DAY,beginDate,lastStepDate)<=7 and funnelId='fid'group bybeginDate,visitorId)ta group by lastStep
[0114] Through a method for measuring advertising effectiveness disclosed in an embodiment of the present application, a multi-source user conversion funnel model that can match a specified population is provided to evaluate advertising effectiveness in response to actual customer needs, and a calculation, storage, and fast query solution for user conversion funnel analysis is provided; by calculating and storing key data for conversion funnel queries, a large amount of data that does not need to participate in conversion funnel queries is filtered out, effectively reducing storage and query pressure; through the aggregation and bitmap indexing functions of the database, the matching efficiency of user IDs is greatly improved, thereby improving the timeliness of user conversion funnel data; through a special table structure design, space is exchanged for time, and flexible adjustment and fast query of the funnel window period are achieved, so that real-time multi-source conversion funnel analysis and conversion funnel analysis that can match a specified population can be easily completed; a richer multi-source user conversion funnel analysis method is provided, a user conversion funnel analysis method for a custom population is provided, and a more efficient (can be used in real time) user conversion funnel architecture, storage, calculation, and query solution are provided; the application scenarios are rich, and can be used for the calculation of other user conversion funnels, including user conversion funnel analysis scenarios for websites such as online shopping malls and educational websites.
[0115] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent.
[0116] Embodiment 2:
[0117] In combination with the method for measuring advertising effectiveness disclosed in the first embodiment, this embodiment discloses a specific implementation example of a system for measuring advertising effectiveness (hereinafter referred to as the “system”).
[0118] Reference Figure 8 As shown, the system comprises:
[0119] Configuration module 1, configuration storage structure;
[0120] Get module 2. After configuration is completed, get the conversion funnel table;
[0121] The query module 3 queries the funnel data based on the conversion funnel table and evaluates the advertising effect.
[0122] Specifically, in some embodiments, the configuration module 1 specifically includes:
[0123] A funnel configuration unit 11, configuring the funnel storage;
[0124] Hbase configuration unit 12, configures Hbase storage structure;
[0125] Druid configuration unit 13 configures the Druid storage structure.
[0126] Specifically, in some embodiments, the acquisition module 2 specifically includes:
[0127] The log pulling unit 21 pulls the real-time log from the message queue;
[0128] A data processing unit 22 performs data processing on the real-time log;
[0129] The state query unit 23 extracts the visitor ID in the real-time log after data processing, queries the conversion funnel state corresponding to the visitor ID in the Hbase, and deserializes it into a local temporary cache object;
[0130] The judging unit 24 groups the real-time logs according to the visitor IDs, and judges the real-time logs one by one in each group;
[0131] The deleting unit 25 deletes the data corresponding to the visitor ID from the local temporary cache object if the conversion funnel state of the visitor ID has not been modified or newly created after the traversal.
[0132] A writing unit 26 serializes all data in the local temporary cache object and writes it to the Hbase;
[0133] The data reading unit 27 reads data from the message queue in real time to obtain the conversion funnel table.
[0134] Specifically, in some embodiments, the data processing unit 22 specifically includes:
[0135] A cleaning unit 221 performs data cleaning on the real-time log;
[0136] The filtering unit 222 filters the abnormal logs.
[0137] Specifically, in some embodiments, the determining unit 24 specifically includes:
[0138] The first judging unit 241 judges whether the real-time log will update the current funnel state in the local temporary cache object, and if so, writes the latest state into the message queue;
[0139] The second judgment unit 242 judges whether the real-time log will trigger a new funnel calculation. If yes, a conversion funnel state object is created with the real-time log as a trigger log, and updated to the local temporary cache object.
[0140] For the technical solutions of the other parts of the advertising effect measurement system disclosed in this embodiment and the advertising effect measurement method disclosed in the first embodiment that are the same as those in the first embodiment, please refer to the description in the first embodiment and will not be repeated here.
[0141] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0142] In summary, the beneficial effects of the present invention are that a method for measuring advertising effectiveness is proposed, and a multi-source user conversion funnel model that can match a specified population is provided to evaluate advertising effectiveness in response to actual customer needs, and a calculation, storage, and fast query solution for user conversion funnel analysis is provided; by calculating and storing key data for conversion funnel queries, a large amount of data that does not need to participate in conversion funnel queries is filtered out, effectively reducing storage and query pressure; through the aggregation and bitmap indexing functions of the database, the matching efficiency of user IDs is greatly improved, thereby improving the timeliness of user conversion funnel data; through a special table structure design, space is exchanged for time, and flexible adjustment and fast query of the funnel window period are achieved, and real-time multi-source conversion funnel analysis and conversion funnel analysis that can match a specified population can be easily completed; a richer multi-source user conversion funnel analysis method is provided, a user conversion funnel analysis method for a custom population is provided, and a more efficient (can be used in real time) user conversion funnel architecture, storage, calculation, and query solution are provided; the application scenarios are rich, and can be used for the calculation of other user conversion funnels, including user conversion funnel analysis scenarios for websites such as online shopping malls and educational websites.
[0143] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.
Claims
1. A method for measuring advertising effectiveness, characterized in that , including the following steps: Configuration steps: Configure storage structure; Acquisition steps: After configuration is completed, obtain the conversion funnel table; Query step: based on the conversion funnel table, query the funnel data and evaluate the advertising effect; The configuration step specifically includes the following steps: Funnel configuration steps: Configure funnel storage; Hbase configuration steps: configure Hbase storage structure; Druid configuration steps: Configure Druid storage structure; The obtaining step specifically includes the following steps: Log pulling step: pull real-time logs from the message queue; Data processing step: performing data processing on the real-time log; Status query step: extracting the visitor ID in the real-time log after data processing, querying the conversion funnel status corresponding to the visitor ID in the Hbase, and deserializing it into a local temporary cache object; Judgment step: grouping the real-time logs according to the visitor IDs, and judging the real-time logs one by one in each group; Deletion step: after traversal, if the conversion funnel state of the visitor ID has not been modified or newly created, deleting the data corresponding to the visitor ID from the local temporary cache object; Writing step: serializing all data in the local temporary cache object and writing them to the Hbase; Data reading step: reading data from the message queue in real time to obtain the conversion funnel table.
2. The method for measuring the advertising effect according to claim 1, characterized in that: The data processing step specifically includes the following steps: Cleaning step: performing data cleaning on the real-time log; Filtering step: Filter abnormal logs.
3. The method for measuring the advertising effect according to claim 1, characterized in that: The judging step specifically includes the following steps: The first judgment step: judging whether the real-time log will update the current funnel state in the local temporary cache object, and if so, writing the latest state into the message queue; The second judgment step: judging whether the real-time log will trigger a new funnel calculation, and if so, creating a conversion funnel state object with the real-time log as the trigger log, and updating it to the local temporary cache object.
4. The method for measuring the advertising effect as claimed in claim 1, characterized in that: The query schemes for the funnel data include, but are not limited to: statistical query by source, statistical query by population, and query by variable window period.
5. A system for measuring advertising effectiveness, characterized in that ,include: Configuration module, configuration storage structure; Get the module. After configuration is completed, get the conversion funnel table. A query module, based on the conversion funnel table, queries funnel data and evaluates advertising effects; The configuration module specifically includes: Funnel configuration unit, configures the funnel storage; Hbase configuration unit, configures Hbase storage structure; Druid configuration unit, configures the Druid storage structure; Wherein, the acquisition module specifically includes: Log pulling unit, pulling real-time logs from the message queue; A data processing unit, performing data processing on the real-time log; A status query unit, which extracts the visitor ID in the real-time log after data processing, queries the conversion funnel status corresponding to the visitor ID in the Hbase, and deserializes it into a local temporary cache object; A judgment unit, grouping the real-time logs according to the visitor ID, and judging the real-time logs one by one in each group; A deleting unit, after traversal, if the conversion funnel state of the visitor ID has not been modified or newly created, deletes the data corresponding to the visitor ID from the local temporary cache object; A writing unit, serializing all data in the local temporary cache object and writing them to the Hbase; The data reading unit reads data from the message queue in real time to obtain the conversion funnel table.
6. The advertising effectiveness measurement system according to claim 5, characterized in that: The data processing unit specifically includes: A cleaning unit, performing data cleaning on the real-time log; Filter unit, filters exception logs.
