CN108694174B

CN108694174B - Content delivery data analysis method and device

Info

Publication number: CN108694174B
Application number: CN201710218077.XA
Authority: CN
Inventors: 石瑞超; 黄浩; 邹正勇; 李学凯; 冯喆; 陈禺伶; 苏麒匀
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2017-04-05
Filing date: 2017-04-05
Publication date: 2022-12-23
Anticipated expiration: 2037-04-05
Also published as: CN108694174A

Abstract

The application discloses a method for analyzing content delivery data, which comprises the following steps: providing data of a first page to a client for display; receiving, from the client, a delivery indicator and a content attribute dimension to be analyzed, which are configured by the client in response to a user operation on the control in the first page; acquiring historical release data corresponding to each information item under the content attribute dimension to be analyzed; determining the value of the release index to be analyzed at the specified time according to the historical release data corresponding to each information item; and determining the influence degree of each information item on the putting index or the influence degree of each information item on the change of the putting index according to the historical putting data corresponding to each information item and the value of the putting index to be analyzed. Providing the influence degree of each information item to the client to cause the client to present the influence degree of each information item in the first page or the second page. The application also provides a corresponding device.

Description

Content delivery data analysis method and device

Technical Field

The invention relates to the technical field of internet, in particular to a method and a device for analyzing content delivery data.

Background

With the development of internet technology, more and more media contents (including text, pictures, audio, video, etc.) are pushed to various users through the internet. Such as: when browsing a web page using a terminal device such as a mobile phone or a PC, a user may receive various media contents pushed by a network side, such as: advertisements in picture or video format, public service promotional information, news, etc. Thus, the user can know the time information, the interested contents and the like in time. Such media content may be referred to as push information or push content, etc.

Disclosure of Invention

The application example provides a method for analyzing content delivery data, which comprises the following steps:

providing data of a first page to a client so that the client displays the first page, wherein a control used for configuring a release index to be analyzed and a content attribute dimension is displayed in the first page;

receiving, from the client, a delivery indicator and a content attribute dimension to be analyzed, which are configured by the client in response to a user operation on the control in the first page;

obtaining historical delivery data corresponding to each information item under the dimension of the content attribute to be analyzed,

determining the value of the release index to be analyzed at the specified time according to the historical release data corresponding to each information item;

and determining the influence degree of each information item on the putting index or the influence degree of each information item on the change of the putting index according to the historical putting data corresponding to each information item and the value of the putting index to be analyzed.

Providing the influence degree of each information item to the client so that the client presents the influence degree of each information item in the first page or the second page.

The present application further provides an analysis apparatus for content delivery data, including:

the system comprises a sending unit, a receiving unit and a processing unit, wherein the sending unit is used for providing data of a first page to a client so as to enable the client to display the first page, and a control used for configuring a release index to be analyzed and a content attribute dimension is displayed in the first page;

a receiving unit, configured to receive, from the client, a delivery index and a content attribute dimension to be analyzed, which are configured by the client in response to an operation of the control in the first page by a user;

an influence degree determination unit configured to:

An influence sending unit, configured to provide the influence degree of each information item to the client, so that the client displays the influence degree of each information item in the first page or the second page.

By adopting the scheme provided by the application, the problem troubleshooting can be automated, and the labor and the time are saved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.

FIG. 1 is a system architecture diagram to which an example of the present application relates;

FIG. 2 is a flow diagram of a method for analyzing content delivery data according to an example embodiment of the present application;

FIG. 3 is a diagram of an advertisement impact analysis page according to an example of the present application;

FIG. 4 is a diagram of an example advertisement transaction analysis page of the present application;

FIG. 5 is a diagram of an example advertisement shadow trend analysis page of the present application;

FIG. 6 is a schematic diagram of an analysis apparatus for content delivery data according to an example of the present application; and

FIG. 7 is a block diagram of a computing device in an example of the present application.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.

The application provides a method for analyzing content delivery data, which can be applied to the system architecture shown in fig. 1. As shown in fig. 1, the system architecture includes: the system comprises an Application (APP) client 101, a background manager client 103, a delivery platform 102 and a media content server 105, wherein the Application (APP) client 101, the background manager client 103, the delivery platform 102 and the media content server 105 communicate through the internet 104.

End users may access the media content server 105 using the application client 101, such as: browsing a web page, viewing an online article, or viewing an online video, etc. When a user accesses the media content server 105 using the application client 101, the application client 101 may issue a content delivery request to the delivery platform 102, and the delivery platform 102 may push media content (e.g., pictures, videos, etc.) matching the content delivery request to the application client 101. The database of the delivery platform 102 records log data of content delivery, where the log data includes a provider of delivered media content, delivery times, time of each delivery, a target user of the delivery, whether the delivered media content is clicked by a user, and the like. When the delivered content is an advertisement, the log data includes data such as an advertiser, a product type, an advertisement ID, an advertisement slot, CPM (Cost Per thousand people, average Cost Per thousand people for a delivery of an advertisement), download rate, consumption (click rate of an advertisement within a set time range), exposure, click rate, and the like. When a user browses page content using an application client 101 (e.g., a microblog, a WeChat, a QQ, a news client, a browser, etc.), the content released by the release platform 102 to the application client 101 may be: videos, articles, news, advertisements, and the like.

The background manager accesses the server providing the data analysis service in the delivery platform 102 through the client 103, and the server providing the data analysis service can obtain the delivered log data in the database of the delivery platform 102, so that the background manager analyzes and manages the indexes of the delivered content according to the delivered log data by accessing the server.

In some examples, various problems of fluctuation of the effect of the delivered content are encountered in the delivery process, and these problems often involve a large amount of data, have multiple dimensions and are difficult to analyze, and usually require more manpower to analyze in combination with multi-dimensional log data.

Based on the above technical problem, the present application provides an analysis method for content delivery data, which can be applied to the delivery platform 102, and in particular, can be applied to a server in the delivery platform 102 for providing data analysis services. As shown in fig. 2, the method comprises the steps of:

step 201: providing data of a first page to a client to enable the client to display the first page, wherein a control used for configuring a release index and a content attribute dimension to be analyzed is displayed in the first page.

The delivery platform 102 provides the first page data to the client 101, so that the client displays the first page data. The client 101 may be a browser client. In some examples, when a user (e.g., a backend administrator) logs into a content delivery data analysis website provided by delivery platform 102 through a browser client, a server of the website in delivery platform 102 (i.e., a server providing data analysis services) sends data of a first page to the browser client, so that the first page is presented on the browser. The first page may be an influence degree analysis page, that is, an analysis of the index of the delivered content at one time point, an anomaly analysis page, that is, a change in the index of the delivered content at two time points, or a trend analysis page, that is, a change in the index of the delivered content over a period of time. The influence analysis page, the transaction analysis page and the trend analysis page have different websites, a user selects the corresponding analysis page by logging in different websites, and links of other analysis pages can be displayed on one analysis page, so that the user can conveniently log in other analysis pages on the current analysis page.

The control of the release index to be analyzed and the control of the content attribute dimension are displayed on the first page, the control of the release index to be analyzed is used for enabling a user to select the index to be analyzed of the release content, and the control of the content attribute dimension is used for enabling the user to select the attribute dimension of the release content. For example, for the delivered content being an advertisement, the indicators to be analyzed include: CPM (Cost Per thousand mile, average Cost Per thousand delivered ads), download rate, consumption, exposure, click-through rate, etc., and the content attribute dimensions of the ads include: ad ID, advertiser, ad spot, type of goods, ID of goods, etc.

Step 202: and receiving a release index and a content attribute dimension to be analyzed, which are configured by the client in response to the operation of the control in the first page by the user, from the client.

The user operates the release index control and the content attribute dimension control to be analyzed in the first page, and the client responds to the operation of the user to obtain the release index and the content attribute dimension of the release content and sends the release index and the content attribute dimension to the server providing the data analysis service in the release platform 102. For example, for the delivered content being an advertisement, fig. 3 shows an influence analysis page diagram of the advertisement, in which an analysis dimension control 301 and an analysis index control 302 are shown. Analysis dimension control 301 can be an input box, a user inputs a content property dimension, control 301 can also be a drop-down list, a user selects a content property dimension in the drop-down list, and control 301 can also be a combined list. The analysis index control 302 may be an input box, and the user inputs an index to be analyzed of the advertisement, or may be a drop-down list, and the user selects an index to be analyzed of the advertisement in the drop-down list, or the control 302 may be a combined list. After a user inputs a to-be-analyzed index (CPM) and a content attribute dimension (commodity type) of an advertisement, a browser where the influence analysis page is located sends the to-be-analyzed index (CPM) and the analysis dimension (commodity type) to a server providing a data analysis service in the delivery platform 102. Fig. 4 is a page diagram of advertisement transaction analysis, where an analysis dimension control 401 and an analysis index control 402 are displayed in the page diagram, and a user inputs a content attribute dimension of an advertisement in an input box of the analysis dimension control 401: ad ID, or select a content attribute dimension in the drop down list of analysis dimension control 401: advertisement ID, the user enters the index to be analyzed for the advertisement in the input box of the analyze index control 402: download rate, or select the ad's metrics to be analyzed in the drop-down list of the analyze metrics control 402: the download rate. Control 401 and control 402 may also be a combo box. Fig. 5 is a page diagram of trend analysis of an advertisement, in which an analysis dimension control 501 and an analysis index control 502 are displayed, and a user inputs a content attribute dimension of the advertisement in an input box of the analysis dimension control 501: the advertiser, or in the drop-down list of the analyze dimension control 501, selects the content attribute dimension: the advertiser, user, enters the ad's metrics to be analyzed in the input box of the analysis metrics control 502: CPM, or select the ad's metrics to be analyzed in a drop-down list of the analyze metrics control 502: and (4) CPM. Control 501 and control 502 may also be a combo box.

Step 203: and acquiring historical delivery data corresponding to each information item under the content attribute dimension to be analyzed.

The server providing the data analysis service in the delivery platform 102 obtains different information items in the content attribute dimension according to the received content attribute dimension from the client, and obtains historical delivery data corresponding to the different information items at the same time, for example, when the delivered content is an advertisement, when the content attribute dimension of the advertisement is an advertiser, each information item in the content attribute dimension of the advertisement is a different advertiser, and the historical delivery data corresponding to each information item is CPM, download rate, consumption, exposure, click rate and the like of the advertisement of each advertiser in the delivery history process.

Step 204: and determining the value of the release index to be analyzed at the specified time according to the historical release data corresponding to each information item.

And a server providing data analysis service in the launching platform 102 determines the overall value of the launching index at the specified time according to the historical launching data corresponding to each information item. For example, for the case that the delivered content is an advertisement, when the delivery index to be analyzed is consumption and the content attribute dimension is an advertiser, each information item in the content attribute dimension corresponds to each advertiser, and the historical delivery data of each advertiser includes: CPM, download rate, consumption, exposure, click rate and other data of the advertisement of each advertiser in the historical process of delivery. And determining an overall consumption value according to consumption data in the historical delivery data of each advertiser at the specified time, for example, taking the sum of the consumption values of the advertisers as the overall consumption value of the delivery index to be analyzed.

Step 205: and determining the influence degree of each information item on the putting index or the influence degree of each information item on the change of the putting index according to the historical putting data corresponding to each information item and the value of the putting index to be analyzed.

When the specified time is a single time point, the influence degree of each information item on the release index can be determined, and when the specified time is two time points or a period of time is, the influence degree of each information item on the change of the release index can be determined. When the index to be analyzed of the advertisement is CPM, and when the influence degree of the advertisement is analyzed, the influence degree of each advertiser on the whole CPM is determined according to the data of CPM, consumption, exposure and the like of the advertisement of each advertiser. When the advertising is analyzed differently and simultaneously or trend, the influence degree of each advertiser on the change of the whole CPM at two time points or the change of the whole CPM within a period of time is determined according to the historical advertisement delivery data of each advertiser.

The analysis method of the content delivery data in the embodiment of the application is to analyze different information items under the content attribute dimension of a delivered content to find the information item which has the largest influence on the index or has the largest influence on the index change. The information items are different options in the content attribute dimension, for example, the information items in the age content attribute dimension are 1-100 years old, the information items in the gender content attribute dimension are male and female, the information items in the ad slot content attribute dimension are ad slot id1, ad slot id2 and the like, wherein the ad slot refers to a specific display position of the advertisement when the advertisement is delivered. The influence degree of each information item is expressed by import (imp) in the present application, and the larger the influence degree is, the larger the influence of the corresponding information item on the overall index or the change in the overall index is. For example, the consumption of an advertisement on a single day mobile intranet site is 1 ten thousand yuan, and analysis from the ad slot content attribute dimension can find out which ad slot contributes to larger consumption, or analysis from the age dimension can see that people in that age group contribute to larger consumption. When the influence degree of each information item is determined, the influence degree of different information items under the same content attribute dimension on the delivery index or the influence of different information items on the delivery index change is only compared, and cross comparison is not performed on the information items under different content attribute dimensions.

Step 206: providing the influence degree of each information item to the client so that the client presents the influence degree of each information item in the first page or the second page.

After the server providing the data analysis service in the delivery platform 102 determines the influence degree of each information item in the previous step, the influence degree of each information item is sent to the client for display. When the influence degree analysis is performed on the putting indexes to be analyzed, the influence degrees of the information items on the putting indexes to be analyzed are displayed on the influence degree analysis page or a new page, so that a user can conveniently analyze the formation reason of a certain index at a certain specific time point, for example, the CPM of an advertisement space in 2016, 12, and 10 is a formation reason of 35 yuan. When the advertisement is analyzed in a transaction manner, the influence degree of change of each information item on a to-be-analyzed delivery index at two time points is displayed on a transaction analysis page or a new page, so that a user can analyze the reason of change of a certain index at the two time points conveniently, for example, the CPM of an advertisement position yesterday is 30 yuan, the CPM of the previous day is 40 yuan, and the reason of 10 yuan reduction of yesterday compared with the previous day is analyzed. When the putting index to be analyzed is subjected to trend analysis, the influence degree of the change of each information item on the index to be analyzed in a period of time is displayed on a trend analysis page or a new page, so that a user can conveniently analyze the reason of the change of a certain index in a period of time. For example, the reasons why a certain index fluctuates continuously over a period of time, such as the fact that the CPM of an ad slot has fallen from the last month to the month, are analyzed.

By adopting the analysis method of the content delivery data provided by the embodiment of the application, the delivery index to be analyzed and the content attribute dimension input by the user are received, the influence degree of different information items on the delivery index or the influence degree on the change of the delivery index under the content attribute dimension can be automatically calculated, so that the user can conveniently analyze the formation reason of the delivery index or the change reason of the delivery index, the automation of problem troubleshooting is realized, and the labor and the time are saved.

In some examples, a time control for configuring a time to be analyzed is further shown in the first page;

the method further comprises the following steps: receiving, from the client, a time to analyze that the client has configured in response to a user operation of the control in the first page; wherein the time to be analyzed is the specified time.

The time control to be analyzed is used for the user to input the time to be analyzed, which is the time specified in the above step 601. When the influence degree of the index to be analyzed is analyzed, the time to be analyzed is one time point, when the index to be analyzed is analyzed in a different motion mode, the time to be analyzed is two time points, and when the index to be analyzed is analyzed in a trend mode, the time to be analyzed is one time period. The time control to be analyzed may be an input box, the time to be analyzed is input in the input box by the user, or may be a drop-down list, the time to be analyzed is selected from the drop-down list by the user, or may be a combination box including both the input box and the drop-down list. For example, for the delivered content being an advertisement, in the influence analysis page in fig. 3, the time control to be analyzed 303 is for the user to input or select one time point, in the transaction analysis page in fig. 4, the time control to be analyzed 403 is for the user to input or select two time points, and in the trend analysis page in fig. 5, the time control to be analyzed 503 is for the user to input or select a time period.

In some examples, wherein the specified time comprises a point in time. And aiming at a content attribute dimension, calculating the influence degree of each information item under the content attribute dimension on the release index at the time point. For example, the CPM of a certain ad slot is 35-element, and when the content attribute dimension is a commodity type, the analysis is mainly to analyze which advertisements of the commodity type on the ad slot contribute more to the overall CPM value of the ad slot.

When influence degree analysis is carried out on the putting index to be analyzed, the method mainly comprises the following steps:

step S101: when the historical putting data corresponding to each information item is obtained: and acquiring historical release data corresponding to each information item at the time point.

As shown in fig. 3, the dimension of the content attribute input by the user is a commodity type, and the time point is 20170315, then the server providing the data analysis service in the delivery platform 102 acquires the delivery data of each commodity type at the time point.

Step S102: when the value of the putting index is determined: and calculating the value of the release index at the time point according to the historical release data corresponding to each information item at the time point.

As also described in the above example, the value of the delivery index is obtained according to the obtained delivery data of each commodity type at the time point, that is, the overall value of CPM at the time point is obtained, the consumption of each commodity type at the time point 20170315 is summed to obtain the overall consumption value, the exposure of each commodity type at the time point 20170315 is summed to obtain the overall exposure, and the overall consumption and the overall exposure are determined as the overall value of CPM at the time point. In some examples, the CPM is composed of index consumption and exposure, i.e., CPM = consumption/exposure, so that the value of the overall consumption value and the value of the overall exposure value at the time point can be used as the value of the release index. The sum of the values of the consumption of each commodity type at the time point 20170315 can be determined as the overall value of the consumption at the time point, and the sum of the values of the exposure of each commodity type at the time point 20170315 can be determined as the overall value of the exposure at the time point.

Step S103: and determining the index value of each information item on the time point to the contribution of the release index according to the historical release data corresponding to each information item on the time point.

As shown in fig. 3, the placement index to be analyzed is CPM, the content attribute dimension is a commodity type, and the historical placement data corresponding to each commodity type includes CPM, consumption, exposure, download rate, click rate, and other data in the advertisement placement process for each commodity type. And the index value of the contribution of each commodity type to the overall CPM is the CPM value of each commodity type, and the CPM is obtained from the historical putting data of each commodity type. In some examples, since the CPM is composed of a consumption index and an exposure index, the index value contributed by each commodity type may be a consumption value and an exposure value of each commodity type, which are also obtained from historical release data of each commodity type.

Step S104: when the influence degree of each information item on the putting index is determined: and determining the influence degree of each information item on the release index according to the index value of each information item on the release index contribution at the time point and the value of the release index at the time point.

Also as described above, the influence of each product type on the overall CPM at time point 20170315 is determined based on the consumption value and exposure value of the advertisement for each product type at time point 20170315, the overall value consumed at time point 20170315, and the overall value of exposure.

And dividing the putting indexes to be analyzed into a single index and a composite index according to the calculation mode of the influence degree of each information item. The single index is an index which can only be used for polymerization, and the composite index comprises more than two single indexes which are polymerized by the single indexes. For example, for an advertisement, the ad's metrics such as consumption, exposure, etc. are a single metric, and the ad's CPM = consumption/exposure, is a composite metric.

In some examples, when the placement metric is a single metric, the determining the degree of influence of each information item on the placement metric includes: and taking the ratio of the index value of each information item to the contribution of the release index to the value of the release index as the influence degree of each information item on the release index.

When the release index to be analyzed is a single index, the influence degree of the ith information item on the release index under the content attribute dimension is calculated through the following formula (1)

imp _i ＝index _i /index (1)

Wherein index represents the value of the launch index, index _i An index value representing the contribution of the i-th information item to the placement index, e.g. for the advertisement index CPM, index is the CPM value of the whole, index _i The CPM value of an advertisement of a certain commodity type can be calculated, when the overall CPM value is calculated, the consumption of each commodity type can be summed to obtain overall consumption, the exposure of each commodity type is summed to obtain overall exposure, and the overall CPM is obtained through the overall consumption and the overall exposure.

For example, if the index is the overall consumption of a current advertisement space, and the analysis dimension is the type of a commodity, the index is _i Which refers to how much money is consumed for a particular type of product at the spot. The larger the ratio of the consumption value of a certain commodity type to the overall consumption value is, the more money is consumed by the commodity type relatively, and the influence on the overall consumption is larger.

In some examples, when the placement index includes a single index a and a single index B, the index value contributed by each information item includes an index value contributed by each information item to the single index a and the single index B, respectively; the value of the putting index comprises a value of a single index A and a value of a single index B;

determining the influence degree of each information item by the following formula (2):

wherein, imp _i Is the influence degree of the ith information item on the placement indicator in the content attribute dimension, index is the value of the placement indicator,

a is the value of the single index A, B is the value of the single index B, index _i The indicator value contributed for the ith information item,

A _i for the index value, B, of the i-th information item contributing to the index A _i The index value that the ith information item contributes to the index B.

As also shown in fig. 3, the placement index to be analyzed is CPM, the content attribute dimension is a commodity type, and the historical placement data corresponding to each commodity type includes CPM, consumption, exposure, download rate, click rate, and other data in the advertisement placement process for each commodity type. Wherein, A is the overall value consumed at the time point 20170315, B is the overall value exposed at the time point 20170315, and the overall value of CPM at the time point 20170315, i.e. index, can be obtained according to A and B. A. The _i Consumption value for i-th merchandise type, B _i Exposure value for ith type of goods according to A _i And B _i The CPM value index of the contribution of the ith commodity type can be obtained _i . Therefore, the influence degree of each commodity type on the whole CPM can be obtained according to the formula (2).

In formula (2), in the calculation of the influence of each information item on the placement index, the value index of the placement index at the time point and the index value index of the ith information item contributing to the placement index are considered _i Also, it is important to consider a single drop indicator included in the drop indicatorThe index B has a greater influence on the placement index because the information item having a large index value has a greater influence on the placement index, for example, for the compliance index: download rate = number of downloads/number of clicks. The number of clicks here is B in the formula. The click number is considered because, for the advertisement with a large click number, even if the fluctuation of the download rate is small, the whole download rate of the advertisement system is influenced; for the advertisement with a small number of clicks, even if the download rate fluctuates greatly, the overall download rate may not be affected.

In the page diagram of the impact analysis of ads shown in FIG. 3, the querying user enters or selects CPM in the analysis metrics control 302, types of goods in the analysis dimension control 401, and a point in time 20170315 in the date control 403, while entering or selecting an ad slot in the filter options control 304: feeds Huang Jinwei, which analyzes which commercial type advertisement on Feeds Huang Jinwei contributes most to CPM on the Feeds gold, where CMP = consumption/exposure, is a composite index, and the influence is calculated using the above equation (2). The results of the impact analysis obtained are shown in table 1:

analysis results
					Type of goods	Actual consumption of	Number of exposures	CPM	impact
Connections outside the Tencent domain (1000)	2793127.323	134190605	20.8146	-0.051
					Open platform mobile APP (12)	3258658.796	148175729	21.9919	-0.0378
Authentication space-authentication space front page (3)	83620.3995	6246523	13.3867	-0.0073
					Jingdong POP merchant direct-casting advertisement (25)	760089.0703	32992527	23.0382	-0.0047
Public comment shop (26)	44560.97	3556471	12.5295	－0.0045
					0	16120.9563	1153317	13.9779	-0.0013
41	419.58	32908	12.7501	0
					Apple App Store-Apple applications (19)	2461655.3467	59752665	41.1974	0.1066

TABLE 1

The method comprises the steps that a first column is a commodity type, namely advertisements of different commodity types on a Feeds gold bit, a second column is consumption values corresponding to the advertisements of the different commodity types, a third column is exposure values corresponding to the advertisements of the different commodity types, a fourth column is CPM values corresponding to the advertisements of the different commodity types, a fifth column is influence values of the advertisements of the different commodity types on the whole CPM on the Feeds gold bit, the influence values are obtained through calculation, the influence values are sequenced from large to small, and the larger the influence value is, the corresponding commodity type has on the CPM.

In some examples, the specified time includes: and the first time point and the second time point are used for carrying out transaction analysis on the putting index to be analyzed. The analysis of the variation is used for analyzing the fluctuation reason of the delivery index at two different time points, for example, cpm of one advertisement position yesterday is 40 yuan, cpm of the previous day is 30 yuan, and the reason that yesterday is 10 yuan lower than the previous day is analyzed. When carrying out transaction analysis on the release indexes to be analyzed, the method mainly comprises the following steps

Step S201: when the historical putting data corresponding to each information item is obtained: and acquiring historical release data corresponding to each information item at the first time point and historical release data corresponding to each information item at the second time point.

In the example shown in fig. 4, the time points of input are two time points: 20161128 and 20161110, where the content attribute dimension is an advertisement ID, and a server in the delivery platform 102 providing a data analysis service obtains historical delivery data of an advertisement represented by each advertisement ID at a time point 20161128 and a time point 20161110, where the historical delivery data includes: download number, number of clicks, consumption, exposure, click rate, etc.

Step S202: when the value of the putting index is determined: determining a value of a release index at the first time point according to historical release data corresponding to each information item at the first time point, and determining a value of a release index at the second time point according to historical release data corresponding to each information item at the second time point;

also as described in the above example, the analysis index input by the query user is the download rate, among others. Download rate = download number/click number, and the total download number and the value of the click number at time point 20161128 are calculated from the download number and the click number in the historical delivery data of each advertisement ID at time point 20161128, for example, the sum of the download numbers of each advertisement ID at time point 20161128 is regarded as the total download number at time point 20161128, the sum of the click numbers of each advertisement ID at time point 20161128 is regarded as the total click number at time point 20161128, and the total download rate at the time point is determined from the total download number and the total click number at time point 20161128. The corresponding overall download rate, overall download number, overall click number and overall download rate at the time point 20161110 can be obtained. In some examples, the value of the placement metric at the first time point includes a total download count and a total click count at the first time point, and the value of the placement metric at the second time point includes a total download count and a total click count at the second time point.

Step S203: determining a first index value of each information item at the first time point contributing to the release index according to the historical release data corresponding to each information item at the first time point; determining a second index value of each information item on the second time point for contribution of the release index according to historical release data corresponding to each information item on the second time point;

also as shown in the above example, a second index value of the contribution of the advertisement ID to the overall download rate at time 20161128 is calculated according to the download number and the click number of the advertisement represented by a certain advertisement ID in the historical delivery data of time 20161128, and the second index value includes the download number and the click number of the advertisement ID at the time, and a first index value of the contribution of the advertisement ID to the overall download rate at time 20161110 is calculated according to the download number and the click number of the advertisement in the historical delivery data of time 20161110, and the first index value includes the download number and the click number of the advertisement ID at the time.

Step S204: when the influence degree of each information item on the change of the release index is determined: and determining the influence degree of each information item on the change of the release index according to the first index value contributed by each information item, the second index value contributed by each information item, the value of the release index at the first time point and the value of the release index at the second time point.

As shown in the above example, the influence of each advertisement ID on the change of the total download rate at

time points

20161128 and 20161110 is calculated from the number of downloads and the number of clicks of each advertisement ID at

time points

20161128 and 20161110, and the number of downloads and the number of clicks of each advertisement ID at

time points

20161128 and 20161110. In this example, the download rate = download number/click number, which is a composite index, and the calculation method of the composite index in the transaction analysis is described in detail below

In some examples, when the placement index includes a single index a and a single index B, the first index value contributed by each information item at the first time point includes an index a first index value and an index B first index value contributed by each information item at the first time point; the second index value contributed by each information item at the second time point comprises an index A second index value and an index B second index value contributed by each information item at the second time point; the value of the release index at the first time point comprises a value of a single index A and a value of a single index B at the first time point, the value of the release index at the second time point comprises a value of a single index A and a value of a single index B at the second time point, and the influence degree of each information item on the change of the release index is determined by the following formula (3):

wherein, imp _i -the influence degree for the i-th information item in the content attribute dimension, index being the value of a drop indicator at the first point in time,

a is the value of a single index A at a first time point, B is the value of a single index B at the first time point,

a 'is the value of the single index A at the second time point, B' is the value of the single index B at the second time point, index _i The first indicator value contributed for the ith information item at the first point in time,

A _i for the index A, the first index value, B of the contribution of the ith information item at the first point in time _i A first index value, index, for the index B contributed by the ith information item at said first point in time _i ' the second index value contributed by the i-th information item at the second point in time,

A _i ' index A second index value, B of the contribution of the ith information item at the second point in time _i ' is the ith information item at the second point in timeThe upper contribution index B is the second index value.

Also as an example shown in fig. 4, in formula (3), where a is the overall download number at time point 20161110 and B is the overall click number at time point 20161110, the overall download rate index at time point 20161110 can be obtained from A, B. A 'is the total download number at the time point 20161128, B' is the total click number at the time point 20161128, and the total download rate at the time point 20161128 can be obtained from A 'and B'. A. The _i Is the number of i-th advertisement ID downloads at time point 20161110, B _i The number of clicks for the ith advertisement ID at time point 20161110, from A _i 、B _i The ith advertisement ID download rate index at time point 20161110 can be obtained _i 。A _i ' is the number of downloads of the ith advertisement ID at time point 20161128, B _i ' is the number of clicks of the ith advertisement ID at time point 20161128, from A _i ′、B _i ' can get the download rate of the ith ad ID at time point 20161128. Thus, the influence of each advertisement ID on the overall download rate from time 20161110 to time 20161128 can be obtained according to equation (3).

In formula (3), in the calculation of the influence of each information item on the placement index, the value index of the placement index at the time point and the index value index of the ith information item contributing to the placement index are considered _i A single index B included in the placement index is also heavily considered because an information item having a large index value of the index B has a greater influence on the placement index, for example, for a match index: download rate = number of downloads/number of clicks. The number of clicks here is B in the formula. The click number is considered because, for the advertisement with a large click number, even if the fluctuation of the download rate is small, the whole download rate of the advertisement system is influenced; for the advertisement with a small click number, even if the download rate fluctuates greatly, the overall download rate may not be affected.

As shown in fig. 4, in the advertisement transaction analysis example, a query user inputs or selects a download rate in an analysis indicator control 402, inputs or selects an advertisement ID in an analysis dimension control 401, inputs or selects two time points, 20161110 and 20161128, in a time control 403, and simultaneously inputs a site and an advertisement product ID in a filter control 404, that is, the overall download rate of the advertisement of the product on the site is analyzed, and it is analyzed which advertisements have a large influence on the change of the overall download rate. The results of the transaction analysis are shown in table 2 below:

TABLE 2

The first column is advertisement ID, the second column is the influence degree of each advertisement ID on the whole download rate, the influence degrees are ranked from high to low, and the larger the influence degree is, the larger the influence of the advertisement corresponding to the corresponding advertisement ID on the whole download rate is. The third column weight represents the result of the influence degree normalization processing corresponding to the advertisement id, and the change type value refers to the change type of the advertisement corresponding to the advertisement id, and comprises three types of normal operation, stop operation and new operation. The click number ratio represents the ratio of the click number of an advertisement id corresponding to the time point to be analyzed to the whole click number, the delta click number ratio represents the difference between the value of the ratio at the first time point to be analyzed and the value at the second time point to be analyzed, the download rate represents the ratio of the advertisement id corresponding to the download quantity of the advertisement id at the time point to be analyzed to the whole download quantity, and the delta download rate represents the difference between the value of the ratio at the first time point to be analyzed and the value at the second time point to be analyzed.

In some examples, when the delivery index to be analyzed is analyzed differently, and when the delivery index to be analyzed is a single index, and the influence degree of the ith information item on the change of the delivery index in the content attribute dimension is calculated, the influence degree of the ith information item on the change of the index is calculated by the following formula (4)

imp _i ＝Δindex _i /Δindex (4)

Wherein, delta index = index' -index, delta index _i ＝index′ _i -index _i Index is the value of the index delivered at the first time point, index' is the value of the index delivered at the second time point, index _i Is a first index value, index'Is the second index value contributed for the ith information item.

In some examples, the specified time includes: a time period. Namely, trend analysis is carried out on the putting indexes to be analyzed. Trend analysis is used to analyze the reason that a certain indicator fluctuates continuously over a period of time, such as the CPM of an ad spot falling from the last month to the month, and the reason that the CPM of the ad spot falls continuously over the period of time. When trend analysis is carried out on the putting indexes to be analyzed, the method mainly comprises the following steps:

step S301: when the historical putting data corresponding to each information item is obtained: and acquiring historical release data corresponding to each information item in the time period.

In the trend analysis example of the advertisement shown in fig. 5, the placement index to be analyzed is CPM, the content attribute dimension is an advertiser, and the time period is 20170308-20170310, that is, the degree of influence of the advertisement of the advertiser on the change of the CPM in the whole time period is analyzed to be relatively large. The information items correspond to advertisers, historical advertisement data of the advertisements of the advertisers in the time period are obtained, the historical advertisement data comprise a plurality of time points in the time period, in some examples, the time points in the time period are spaced by days, in the time period 20170308-20170310, the historical advertisement data of 20170308, 20170309 and 20170310 at three time points exist, the historical advertisement data corresponding to each time point is obtained, and the historical advertisement data mainly comprise data such as the downloading number, the clicking number, the consumption, the exposure, the clicking rate and the like of the advertisements of the advertisers.

Step S302: when the value of the release index is determined: and determining the value of the release index in the time period according to the historical release data corresponding to each information item in the time period.

As also shown in the above example, the value of the placement indicator to be analyzed is determined according to the historical placement data of each advertiser advertisement, the placement indicator to be analyzed is CPM, CPM = consume/expose, and according to the time period: 20170308-20170310, the consumption and exposure data of the advertisements of the advertisers at a plurality of time points determine the overall consumption value and the overall exposure value of the advertisements at the plurality of time points in the time period, for example, the sum of the consumption values of the advertisers at the time points is used as the overall consumption value, the sum of the exposure values of the advertisers at the time points is used as the overall exposure value, and the overall CPM value at the time points can be determined according to the overall consumption value and the overall exposure value at the time points. And the value of the release index in the time period comprises the whole consumption value and the whole exposure value at each time point.

Step S303: the determining the influence degree of each information item on the change of the release index comprises: and determining the influence degree of each information item on the change of the release index in the time period according to the historical release data corresponding to each information item in the time period and the value of the release index in the time period. Also as shown in the above example, according to the acquired time period: and determining the influence degree of each advertiser on the change of the whole CPM in the 20170308-20170310 time period by the consumption value and the exposure value of the advertisement of each advertiser at each time point in the 20170308-20170310 as well as the whole consumption value and the whole exposure value at each time point in the time period.

In some examples, the method for analyzing content delivery data provided by the present application, when performing trend analysis on a delivery indicator to be analyzed, further includes the following steps:

step S1: a first point in time and a second point in time in the time period are determined.

In the application, when trend analysis is performed on the release indexes to be analyzed, the calculation mode of the influence degree is the same as that of the transaction analysis, namely, the influence degree of each information item on the change of the release indexes at two time points is calculated, and the difference is that for the trend analysis, the first index value and the second index value of the release indexes contributed by each information item at the two time points, and the values of the release indexes at the two time points are obtained through linear fitting. Two time points need to be determined as the time points to be analyzed within the above-mentioned time period. As also shown in the trend analysis example of an advertisement in FIG. 5, the user inputs the time period: 20170308-20170310, and determining a first time point and a second time point in the time period as time points to be analyzed. In some examples, the time points at the two ends of the time period can be respectively used as the first time point and the second time point, namely 20170308 is used as the first time point, and 20170310 is used as the second time point. In other examples, other time points may be selected as the first time point and the second time point in the time period.

Step S2: determining an index value of each information item in the time period for the contribution of the release index according to historical release data corresponding to each information item in the time period;

for any information item, the historical placement data includes a plurality of time points in the time period and historical placement data corresponding to the information item at each time point. As described above, since the advertisement consumption data and exposure data of each advertiser are determined based on the historical advertisement placement data of each advertiser at a plurality of time points in the time period, the placement index CPM to be analyzed is composed of consumption and exposure, and thus the consumption value and exposure value of each advertiser at each time point in the time periods 20170308 to 20170310 constitute index values that contribute to the overall CPM.

And step S3: performing linear fitting on the index values contributed by the information items in the time period, and determining a first index value contributed by each information item to the release index at the first time point and a second index value contributed by each information item to the release index at the second time point according to an obtained fitting straight line;

as also shown in the above example, for any advertiser, the consumption data of the advertiser in the time periods 20170308 to 20170310 are linearly fitted to obtain a fitted straight line, and the consumption value at the first time point and the consumption value at the second time point are read from the fitted straight line. And simultaneously carrying out linear fitting on the exposure data of the advertiser in the time periods 20170308-20170310 to obtain a fitted straight line, and reading the exposure value of the first time point and the exposure value of the second time point on the fitted straight line. The consumption value and the exposure value at the first time point constitute a first index value of the advertiser's contribution to the overall CPM, and the consumption value and the exposure value at the second time point constitute a second index value of the advertiser's contribution to the overall CPM.

And step S4: performing linear fitting on the value of the release index in the time period, and determining the value of the release index at the first time point and the value of the release index at the second time point according to the obtained fitting straight line;

as also shown in the above example, the input index to be analyzed is CPM, which is a composite index composed of single index consumption and single index exposure, and the value of the input index in the time period includes the whole value of consumption and the whole value of exposure in the time period. And performing linear fitting on the overall consumption value at each time point in the period of time to obtain a fitting straight line, and reading the overall consumption value at the first time point and the overall consumption value at the second time point on the obtained fitting straight line. And simultaneously performing linear fitting on the exposure overall value at each time point in the period of time to obtain a fitting straight line, and reading the exposure overall value at the first time point and the exposure overall value at the second time point on the obtained fitting straight line. The value of the release index at the first time point comprises a consumption overall value and an exposure overall value at the first time point, and the value of the release index at the second time point comprises a consumption overall value and an exposure overall value at the second time point.

Step S5: the determining the influence degree of each information item on the change of the release index in the time period comprises: and determining the influence degree of each information item on the change of the release index in the time period according to the first index value contributed by each information item, the second index value contributed by each information item, the value of the release index at the first time point and the value of the release index at the second time point.

As also shown in the above example, the influence of each advertiser on the change of the whole CPM in the time periods 20170308-20170310 is determined according to the consumption value and the exposure value contributed by each advertiser at the first time point, the consumption value and the exposure value contributed by each advertiser at the second time point, the consumption overall value and the exposure overall value at the first time point, and the consumption overall value and the exposure overall value at the second time point.

In the application, when the drop index to be analyzed is subjected to transaction analysis and trend analysis, the influence degree of each information item is calculated in the same manner, namely the influence degree of each information item on the change of the drop index at two time points is calculated, and the difference is that in the calculation of the influence degree in the trend analysis, the first index value, the second index value, the value of the drop index at the first time point and the value of the drop index at the second time point, which are contributed by each information item, are obtained through linear fitting. The delivery indexes to be analyzed can be divided into two types according to the calculation mode of influence degree: single index and composite index, the single index refers to the index only making aggregation, such as advertisement index consumption, exposure, etc.; a composite indicator comprises two or more single indicators, such as CPM = consume/expose.

In some examples, when the placement metric is a single metric, the determining the degree of influence of each information item includes:

step S1: acquiring a first difference value between the first index value contributed by each information item at the first time point and the second index value contributed at the second time point;

step S2: acquiring a second difference value between the value of the release index at the first time point and the value of the release index at the second time point;

and step S3: taking a ratio of the first difference to the second difference as the influence degree of each information item.

When the influence degree of the ith information item on the change of the release index under the dimension of the content attribute is calculated, when the index is a single index, the influence degree of the ith information item on the change of the index is calculated through a formula (4)

imp _i ＝Δindex _i /Δindex (4)

Wherein Δ index = index' -index, Δ index _i ＝index′ _i -index _i Index is the value of the delivered index at the first time point, index' is the value of the delivered index at the second time point, index _i Index' is a second index value contributed for the ith information item.

In some examples, when the placement index includes a single index a and a single index B, the first index value contributed by each information item at the first time point includes an index a first index value and an index B first index value contributed by each information item at the first time point; the second index value contributed by each information item at the second time point comprises an index A second index value and an index B second index value contributed by each information item at the second time point; the value of the release index at the first time point comprises a value of a single index A and a value of a single index B at the first time point, the value of the release index at the second time point comprises a value of a single index A and a value of a single index B at the second time point, and the influence degree of each information item on the change of the release index is determined through a formula (3):

wherein, imp _i An index being the influence of the i-th information item in the content attribute dimension, the index being the value of a delivery indicator at the first point in time,

A _i for the index A, the first index value, B, of the contribution of the ith information item at the first point in time _i A first index value, index, for the index B contributed by the ith information item at said first point in time _i ' the second index value contributed by the i-th information item at the second point in time,

A _i ' index A second index value, B of the contribution of the ith information item at the second point in time _i ' is the indicator B second indicator value that the ith information item contributes at the second point in time.

Also as an example shown in fig. 5, in equation (3), where a is the overall consumption at time point 20170308 and B is the overall exposure at time point 20170308, the overall CPM at time point 20170308 can be obtained from A, B: and index. A 'is the overall consumption at time point 20170310, B' is the overall exposure at time point 20170310, and A 'and B' can obtain the overall CPM at time point 20170310. A. The _i Is the consumption of the ith advertiser at time point 20170308, B _i For the exposure of the ith advertiser at time point 20170308, from A _i 、B _i The CPM of the ith advertiser at time point 20170308 may be obtained: index _i 。A _i ' consumption of the ith advertiser at time point 20170310, B _i ' is the exposure of the ith advertiser at time point 20170310, from A _i ′、B _i ' the CPM of the ith advertiser at time point 20170310 may be obtained. Therefore, according to the formula (3), the influence degree of each advertiser on the whole CPM change from the time point 20170308 to the time point 20170310 can be obtained.

In formula (3), in the calculation of the influence of each information item on the placement index, the value index of the placement index at the time point and the index value index of the ith information item contributing to the placement index are considered _i A single index B included in the placement index is also considered emphatically, because the information item with a large index value of the index B has a greater influence on the placement index, for example, for a match index: download rate = number of downloads/number of clicks. The number of clicks here is B in the formula. The click number is considered because, for the advertisement with a large click number, even if the fluctuation of the download rate is small, the whole download rate of the advertisement system is influenced; for the advertisement with a small number of clicks, even if the download rate fluctuates greatly, the overall download rate may not be affected.

Also as shown in the trend analysis example of an ad in FIG. 5, a query user enters or selects CPM in the analyze metrics control 502, enters or selects an advertiser in the analyze dimensions control 501, enters or selects a time period in the time control 503, 20170308-20170310, while entering or selecting ad spots in the filter control 504: feeds Huang Jinwei, namely, the overall CPM on the ad slot is analyzed, and the influence of which advertisers on the overall CPM change is analyzed. The results of the transaction analysis are shown in table 3 below:

TABLE 3

The first column is an advertiser, the second column is the calculated influence degree of each advertiser on the whole CPM, the influence degrees are ranked from high to low, and the larger the influence degree is, the larger the influence of the corresponding advertiser on the whole CPM is. The third column of weights represents the result of the advertiser's corresponding influence degree normalization process. The change type value refers to the change type of the advertisement corresponding to the advertisement id, and comprises three types of normal, stop-casting and new-entering. The exposure number ratio represents a ratio of an exposure number corresponding to a time point to be analyzed of an advertiser to an overall exposure number, wherein the exposure number and the overall exposure number are results obtained through linear fitting, and the delta exposure number ratio represents a difference between a value of the ratio at a first time point to be analyzed and a value at a second time point to be analyzed. The CPM represents a ratio of CPM and overall CPM contributed by an advertiser at a time point to be analyzed, wherein the CPM and the overall CPM contributed by each advertiser are results obtained through linear fitting, the Δ CPM represents a difference value between a value of the ratio at a first time point to be analyzed and a value at a second time point to be analyzed, and the trend list is CPM variation trend of the advertiser in the time period.

In some examples, wherein a control for configuring a filtering option is shown in the first page; the method further comprises the steps of:

and S1, receiving a filtering condition configured by the client in response to the operation of the user on the filtering option control in the first page from the client.

The filtering control is used for filtering the historical release data when the historical release data of each information item is obtained. As in the impact analysis of advertisements shown in FIG. 3, the user enters or selects a Feeds gold spot at the filter control 304.

S2, acquiring historical release data corresponding to each information item comprises the following steps: and acquiring historical release data corresponding to each information item according to the filtering condition.

As shown in the above example, if the content attribute dimension is a product type, when the user acquires the historical placement data of each product type, only the historical placement data of each product type at the Feeds gold position is acquired.

In some examples, in step 204 above, performing the influence of providing information items to the client comprises:

step S1: and sequencing the influence degrees in a descending order.

And a server providing data analysis service in the launching platform 102 sorts the determined influence degrees of the information items according to a descending order to obtain a sorting result.

Step S2: and providing the sequencing result to the client so that the client can display the influence degree corresponding to each information item according to the sequencing result.

And the server providing data analysis service in the delivery platform 102 sends the determined sequencing result to the client, so that the client displays the influence degree corresponding to each information item according to the sequencing result. For example, in table 1, the influence value of each commodity type on the overall CPM is displayed in a descending order, the influence of each advertisement ID on the change of the overall download rate in table 2 is displayed in a descending order, and the influence of each advertiser on the change of the overall CPM in table 3 is displayed in a descending order.

In some examples, the method for analyzing content delivery data provided by the present application further includes:

and providing parameters related to the influence degree of each information item to the client so that the client displays the related parameters in the first page or the second page.

The client may also present a parameter related to the degree of influence when presenting the degree of influence of the information items. As in table 1, when determining the influence of each commodity type on the overall CPM, consumption, exposure, and CPM data for each commodity type are required, and these data are presented together. In table 2, when determining the influence of each advertisement ID on the overall download rate change, data such as the occupancy, the Δ hit count occupancy, and the download rate need to be calculated using points, and these data are displayed together with the influence. In table 3, when determining the influence of each advertiser on the change of CPM over a period of time, data such as exposure count ratio, Δ exposure count ratio, CPM, and Δ CPM are required, and these data are displayed together with the influence. The trend of the CPM for each advertiser over the time period is also shown in Table 3. The data displayed together with the influence degree can be conveniently analyzed by a user.

The present embodiment further provides an apparatus 600 for analyzing content delivery data, as shown in fig. 6, including

A sending unit 601, configured to provide data of a first page to a client, so that the client displays the first page, where a control for configuring a release index and a content attribute dimension to be analyzed is displayed in the first page;

a receiving unit 602, configured to receive, from the client, a delivery index and a content attribute dimension to be analyzed, where the delivery index and the content attribute dimension are configured by the client in response to a user operating the control in the first page;

an influence determining unit 603 configured to:

An influence sending unit 604, configured to provide the influence degree of each information item to the client, so that the client presents the influence degree of each information item in the first page or the second page.

In some examples, the specified time includes a point in time, the influence level determination unit 603 is configured to:

acquiring historical release data corresponding to each information item at the time point;

calculating the value of the release index at the time point according to the historical release data corresponding to each information item at the time point;

determining an index value of each information item at the time point for contribution to the release index according to the historical release data corresponding to each information item at the time point;

and determining the influence degree of each information item on the release index according to the index value of each information item on the release index contribution at the time point and the value of the release index at the time point.

In some examples, the specified time includes: a first time point and a second time point; the influence determining unit 603 is configured to:

obtaining historical release data corresponding to each information item at the first time point and historical release data corresponding to each information item at the second time point;

determining a value of a release index at the first time point according to historical release data corresponding to each information item at the first time point, and determining a value of a release index at the second time point according to historical release data corresponding to each information item at the second time point;

determining a first index value of each information item at the first time point contributing to the release index according to the historical release data corresponding to each information item at the first time point; determining a second index value of contribution of each information item on the second time point to the release index according to the historical release data corresponding to each information item on the second time point;

the determining the influence degree of each information item on the change of the release index comprises: and determining the influence degree of each information item on the change of the release index according to the first index value contributed by each information item, the second index value contributed by each information item, the value of the release index at the first time point and the value of the release index at the second time point.

In some examples, the specified time includes: a time period; the influence degree determining unit 603 is configured to:

acquiring historical release data corresponding to each information item in the time period;

determining the value of a release index in the time period according to historical release data corresponding to each information item in the time period;

and determining the influence degree of each information item on the change of the release index in the time period according to the historical release data corresponding to each information item in the time period and the value of the release index in the time period.

In some examples, the influence degree determination unit 603 is further configured to:

determining a first time point and a second time point in the time period;

determining an index value of each information item in the time period for the contribution of the release index according to historical release data corresponding to each information item in the time period;

performing linear fitting on the index values contributed by the information items in the time period, and determining a first index value contributed by each information item to the release index at the first time point and a second index value contributed by each information item to the release index at the second time point according to the obtained fitting straight line;

performing linear fitting on the value of the release index in the time period, and determining the value of the release index at the first time point and the value of the release index at the second time point according to the obtained fitting straight line;

the determining the influence degree of each information item on the change of the release index in the time period comprises: and determining the influence degree of each information item on the change of the release index in the time period according to the first index value contributed by each information item, the second index value contributed by each information item, the value of the release index at the first time point and the value of the release index at the second time point.

Fig. 7 is a block diagram showing a configuration of a computing device in which the content delivery data analysis device 600 is located. As shown in fig. 7, the computing device includes one or more processors (CPUs) 702, a communication module 704, a memory 706, a user interface 710, and a communication bus 708 for interconnecting these components.

The processor 702 may receive and transmit data via the communication module 704 to enable network communications and/or local communications.

User interface 710 includes one or more output devices 712, including one or more speakers and/or one or more visual displays. The user interface 710 also includes one or more input devices 714, including, for example, a keyboard, a mouse, a voice command input unit or microphone, a touch screen display, a touch sensitive tablet, a gesture capture camera or other input buttons or controls, and the like.

The memory 706 may be high-speed random access memory, such as DRAM, SRAM, DDR RAM, or other random access solid state memory devices; or non-volatile memory, such as one or more magnetic disk storage devices, optical disk storage devices, flash memory devices, or other non-volatile solid-state storage devices.

The memory 706 stores a set of instructions executable by the processor 702, including:

an operating system 716, including programs for handling various basic system services and for performing hardware related tasks;

the applications 718 include various applications for analysis of content delivery data, which can implement the process flow in the above examples, such as some or all of the units or modules in the analysis apparatus 600 that may include content delivery data. At least one of the units in the analysis means 600 of content delivery data may store machine executable instructions. The processor 702 may be capable of performing the functions of at least one of the units or modules described above by executing machine-executable instructions in at least one of the units in the memory 706.

It should be noted that not all steps and modules in the above flows and structures are necessary, and some steps or modules may be omitted according to actual needs. The execution sequence of the steps is not fixed and can be adjusted according to the needs. The division of each module is only for convenience of describing adopted functional division, and in actual implementation, one module may be divided into multiple modules, and the functions of multiple modules may also be implemented by the same module, and these modules may be located in the same device or in different devices.

The hardware modules in the embodiments may be implemented in hardware or a hardware platform plus software. The software includes machine-readable instructions stored on a non-volatile storage medium. Thus, embodiments may also be embodied as a software product.

In various examples, the hardware may be implemented by specialized hardware or hardware executing machine-readable instructions. For example, the hardware may be specially designed permanent circuits or logic devices (e.g., special purpose processors, such as FPGAs or ASICs) for performing the specified operations. The hardware may also include programmable logic devices or circuits temporarily configured by software (e.g., including a general purpose processor or other programmable processor) to perform certain operations.

In addition, each example of the present application can be realized by a data processing program executed by a data processing apparatus such as a computer. It is clear that a data processing program constitutes the present application. Further, the data processing program, which is generally stored in one storage medium, is executed by directly reading the program out of the storage medium or by installing or copying the program into a storage device (such as a hard disk and/or a memory) of the data processing device. Such a storage medium therefore also constitutes the present application, which also provides a non-volatile storage medium in which a data processing program is stored, which data processing program can be used to carry out any one of the above-mentioned method examples of the present application.

The corresponding machine-readable instructions of the modules of fig. 7 may cause an operating system or the like operating on the computer to perform some or all of the operations described herein. The nonvolatile computer-readable storage medium may be a memory provided in an expansion board inserted into the computer or written to a memory provided in an expansion unit connected to the computer. A CPU or the like mounted on the expansion board or the expansion unit may perform part or all of the actual operations according to the instructions.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and should not be taken as limiting the scope of the present invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims

1. A method for analyzing content delivery data, comprising:

providing data of a first page to a client to enable the client to display the first page, wherein a control used for configuring a release index and a content attribute dimension to be analyzed is displayed in the first page;

when the first page is an abnormal operation analysis page, executing the following processing:

obtaining historical release data corresponding to each information item at a first time point under the content attribute dimension and historical release data corresponding to each information item at a second time point;

determining the value of the release index at the first time point according to historical release data corresponding to each information item at the first time point; determining the value of the release index at the second time point according to historical release data corresponding to each information item at the second time point;

determining a first index value of each information item at the first time point contributing to the release index according to the historical release data corresponding to each information item at the first time point; determining a second index value of each information item on the second time point for contribution of the release index according to historical release data corresponding to each information item on the second time point;

determining the influence degree of each information item on the change of the release index according to the first index value, the second index value, the value of the release index at the first time point and the value of the release index at the second time point;

2. The method of claim 1, further comprising:

when the first page is an influence analysis page, executing the following processing:

acquiring historical release data corresponding to each information item at a third time point;

calculating the value of the release index at the third time point according to the historical release data corresponding to each information item at the third time point;

determining a third index value of each information item at the third time point contributing to the release index according to the historical release data corresponding to each information item at the third time point;

and determining the influence degree of each information item on the putting index according to the third index value and the value of the putting index at the third time point.

3. The method according to claim 2, wherein, when the placement index is a single index, the determining the influence of each information item on the placement index according to the third index value and the value of the placement index at the third time point comprises:

and taking the ratio of the third index value to the value of the release index at the third time point as the influence degree of each information item on the release index.

4. The method according to claim 2, wherein when the placement index comprises a single index a and a single index B, the third index value comprises index values that each information item contributes to the single index a and the single index B, respectively; the value of the release index at the third time point comprises the value of a single index A and the value of a single index B;

determining the influence degree of each information item by the following formula (1):

wherein, imp _i The influence degree of the ith information item on the placement index in the content attribute dimension, index being the value of the placement index at the third time point,

a is the value of the single index A, B is the value of the single index B, index _i A third index value contributed by the ith information item to the placement index,

A _i for the index value, B, of the contribution of the i-th information item to the single index A _i The index value of the contribution of the ith information item to the single index B.

5. The method of claim 1, wherein an analysis dimension control and an analysis indicator control are exposed in the transaction analysis page, the analysis dimension control being used for user input of the content property dimension.

6. The method of claim 1, further comprising:

when the first page is a trend analysis page, executing the following processing:

acquiring historical release data corresponding to each information item in a time period;

determining the value of the release index in the time period according to historical release data corresponding to each information item in the time period;

7. The method of claim 6, further comprising:

determining a fourth time point and a fifth time point in the time period;

determining an index value of each information item in the time period for the contribution of the release index according to the historical release data corresponding to each information item in the time period;

performing linear fitting on the index values of the information items in the time period, which contribute to the release index, to obtain a first fitted straight line, and determining a fourth index value of the information items at the fourth time point, which contribute to the release index, and a fifth index value of the information items at the fifth time point, which contribute to the release index, according to the first fitted straight line;

performing linear fitting on the value of the release index in the time period to obtain a second fitted straight line, and determining the value of the release index at the fourth time point and the value of the release index at the fifth time point according to the second fitted straight line;

the determining the influence degree of each information item on the change of the release index in the time period according to the historical release data corresponding to each information item in the time period and the value of the release index in the time period comprises: and determining the influence degree of each information item on the change of the release index in the time period according to the fourth index value, the fifth index value, the value of the release index at the fourth time point and the value of the release index at the fifth time point.

8. The method according to claim 1, wherein, when the release indicator is a single indicator, said determining the influence of each information item on the change of the release indicator based on the first indicator value, the second indicator value, the value of the release indicator at the first point in time, and the value of the release indicator at the second point in time comprises:

determining a first difference between the first index value and the second index value;

determining a second difference between the value of the placement indicator at the first point in time and the value of the placement indicator at the second point in time;

and taking the ratio of the first difference value to the second difference value as the influence degree of each information item on the change of the putting index.

9. The method according to claim 1, wherein when the placement index includes a single index a and a single index B, the first index value includes an index value contributed to the single index a and an index value contributed to the single index B by each information item at the first time point; the second index value comprises an index value which is contributed to the single index A and an index value which is contributed to the single index B by each information item at the second time point; the value of the release index at the first time point comprises the value of a single index a and the value of a single index B at the first time point, and the value of the release index at the second time point comprises the value of the single index a and the value of the single index B at the second time point;

the influence degree of each information item on the change of the release index is determined by the following formula (2):

wherein, imp _i For the influence of the i-th information item in the content property dimension,

index is the value of the delivery index at the first time point,

a is the first timeThe value of a single index A at a point, B being the value of a single index B at said first point in time,

a 'is the value of a single indicator A at said second point in time, B' is the value of a single indicator B at said second point in time,

index _i the first indicator value contributed for the ith information item at the first point in time,

A _i for the index value, B, of the i-th information item contributing to the single index A at said first point in time _i For the index value of the i-th information item that contributes to the single index B at the first point in time,

index _i ' the second index value contributed by the i-th information item at the second point in time,

A _i ' index value, B, of the i-th information item contributing to the single index A at the second point in time _i ' is an index value of the i-th information item that contributes to the single index B at the second point in time.

10. An apparatus for analyzing content delivery data, comprising:

an influence degree determination unit configured to: when the first page is an abnormal operation analysis page, executing the following processing:

determining the value of the release index at the first time point according to historical release data corresponding to each information item at the first time point; determining the value of the release index at the second time point according to the historical release data corresponding to each information item at the second time point;

11. The apparatus of claim 10, wherein the influence level determination unit is further configured to:

12. The apparatus of claim 10, wherein an analysis dimension control and an analysis indicator control are exposed in the transaction analysis page, the analysis dimension control for user input of the content property dimension.

13. The apparatus of claim 10, wherein the influence level determination unit is further configured to:

14. The apparatus of claim 13, wherein the influence level determination unit is further configured to:

determining a fourth time point and a fifth time point in the time period;

determining an index value of each information item in the time period for contribution to the release index according to historical release data corresponding to each information item in the time period;

and determining the influence degree of each information item on the change of the release index in the time period according to the fourth index value, the fifth index value, the value of the release index at the fourth time point and the value of the release index at the fifth time point.

15. A computing device comprising a memory and a processor, the memory having stored therein computer-readable instructions that, when executed by the processor, implement the method of any of claims 1 to 9.

16. A computer-readable storage medium having computer-readable instructions stored thereon which, when executed by at least one processor, implement the method of any one of claims 1 to 9.