CN113297486B - Click rate prediction method and related device - Google Patents
Click rate prediction method and related device Download PDFInfo
- Publication number
- CN113297486B CN113297486B CN202110565423.8A CN202110565423A CN113297486B CN 113297486 B CN113297486 B CN 113297486B CN 202110565423 A CN202110565423 A CN 202110565423A CN 113297486 B CN113297486 B CN 113297486B
- Authority
- CN
- China
- Prior art keywords
- user
- predicted
- behavior
- model
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 230000006399 behavior Effects 0.000 claims abstract description 138
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 239000013598 vector Substances 0.000 claims description 53
- 238000012549 training Methods 0.000 claims description 31
- RPNUMPOLZDHAAY-UHFFFAOYSA-N Diethylenetriamine Chemical compound NCCNCCN RPNUMPOLZDHAAY-UHFFFAOYSA-N 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 7
- 230000007246 mechanism Effects 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 238000012545 processing Methods 0.000 description 13
- 238000010586 diagram Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 6
- 230000014759 maintenance of location Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000007639 printing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Business, Economics & Management (AREA)
- Computing Systems (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Mathematical Physics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a click rate prediction method and a related device, which are characterized in that the click rate prediction method comprises the following steps: collecting behaviors of a user in a historical time period, behavior time points corresponding to the behaviors, target numbers to be predicted and time points to be predicted; the behavior of the user comprises click data generated by clicking a target by the user and state data of the user and the clicked target; preprocessing behaviors and behavior time points to obtain behavior characteristic information of the user; and predicting the click probability of clicking the target to be predicted at the time point to be predicted by the user based on the behavior characteristic information by using the prediction model. By the method, the click rate of the target to be predicted by the user is predicted, so that the target interested by the user is recommended to the user.
Description
Technical Field
The invention relates to the technical field of deep learning, in particular to a click rate prediction method and a related device.
Background
For a live broadcast platform, the retention and growth of users are one of core tasks of platform operation, and the allocation method for optimizing user traffic can improve the viscosity of users to a host/platform, improve the ecology of platform content, and further realize the core tasks of user retention and growth. The recommendation is an important flow distribution scene, and when the recommended content accords with the interests of the user, the user can contribute more flow and remain to the platform; and when the recommended content is contrary to the interest of the user, the user may be lost because the content of interest of the user cannot be found.
The click rate is an important index for measuring the recommendation effect, and when the user is not interested in the recommended content, clicking on the recommended card cannot occur. Therefore, the click probability of the user on the anchor is predicted, so that the anchor interested by the user is recommended to the user, and the method is important for maintaining the operation of the live platform.
Disclosure of Invention
The application mainly solves the technical problem of providing a click rate prediction method and a related device for predicting the click rate of a target to be predicted by a user so as to recommend the target interested by the user to the user.
In order to solve the technical problems, the application provides a click rate prediction method, which comprises the following steps: collecting behaviors of a user in a historical time period, behavior time points corresponding to the behaviors, target numbers to be predicted and time points to be predicted; the behavior of the user comprises click data generated by clicking the target by the user and state data of the user and the clicked target; preprocessing behaviors and behavior time points to obtain behavior characteristic information of a user; and predicting the click probability of clicking the target to be predicted at the time point to be predicted by the user based on the behavior characteristic information by using the prediction model.
The step of preprocessing the behaviors and the behavior time points to obtain behavior characteristic information of the user comprises the following steps: mapping behaviors and behavior time points into sequence features and statistical features; the step of predicting the click probability of the user clicking the target to be predicted at the point in time to be predicted based on the behavior characteristic information by using the prediction model comprises the following steps: extracting feature information of sequence features by using a first model of the prediction model to obtain first feature information; extracting feature information of the statistical features by using a second model of the prediction model to obtain second feature information; splicing the first characteristic information and the second characteristic information to obtain behavior characteristic information; wherein the first model and the second model are cascaded with each other.
The sequence features comprise a target number clicked by a user, a class number clicked by the user, a position number clicked by the user and time/hour clicked by the user.
The statistical features comprise state information of a user, state information of a target and behavior information of the user to the target to be predicted.
Wherein the first model comprises a DIEN model; the second model includes an FM model.
The step of extracting feature information of the sequence features by using a first model of the prediction model to obtain the first feature information comprises the following steps: mapping the sequence features into sequence feature vectors by utilizing a feature extraction network; the sequence feature vector comprises a behavior sequence vector, a behavior sequence time vector, a target vector to be predicted and a time point vector to be predicted of a user; training the sequence feature vector by using a door control loop network and an attention mechanism to obtain a behavior interest vector of a user; and splicing the behavior interest vector, the target vector to be predicted and the time point vector to be predicted, and training by using the full-connection layer network to obtain first characteristic information.
The step of splicing the first characteristic information and the second characteristic information to obtain behavior characteristic information further comprises the following steps: and predicting the behavior characteristic information by using the full-connection layer network and the normalized exponential function to obtain the click probability of the target to be predicted.
The method comprises the following steps of predicting the click probability of a user clicking a target to be predicted at a point in time to be predicted based on behavior characteristic information by using a prediction model: collecting behavior data of a user in a historical time period and a real target number clicked at a time point to be predicted; the behavior data comprise behaviors of the user in a historical time period and behavior time points corresponding to the behaviors; the behavior data of the historical time period is input into an initial model, and the initial model is utilized to predict a click target of a user at a time point to be predicted based on the behavior data of the historical time period, so that a predicted target number is obtained; training the initial model by using the real target number and the predicted target number, and determining the model after training as a predicted model.
The step of training the initial model by using the real target number and the predicted target number and determining the trained model as the predicted model further comprises the following steps: detecting the trained model through a detection sample; and determining that model training is completed in response to the prediction precision of the trained model reaching the preset precision.
In order to solve the above problems, the present application further provides an intelligent terminal, where the intelligent terminal includes a processor and a memory coupled to each other, the memory is configured to store program instructions, and the processor is configured to execute the program instructions stored in the memory to implement the click rate prediction method of any one of the above embodiments.
In order to solve the above-mentioned problems, the present application also provides a computer readable storage medium, which includes a processor and a memory, the memory storing computer program instructions, the processor being configured to execute the program instructions to implement the click rate prediction method according to any one of the above-mentioned embodiments.
The beneficial effects of the application are as follows: the behavior of the user and the behavior time point of the user are preprocessed by acquiring the behavior of the user in the historical time period, the behavior time point corresponding to the behavior, the target number to be predicted and the time point to be predicted, behavior characteristic information of the user is obtained, and the click probability of the user clicking the target to be predicted (anchor) at the time point to be predicted is predicted by using a prediction model. The click probability of the user on different anchor at specific time and under the page is predicted by the prediction model, so that the anchor interested by the user is recommended to the user, the user experience of the user for watching live broadcast is ensured, and the retention rate of the user in the live broadcast platform is improved. In addition, compared with other shopping prediction models, the prediction model provided by the application considers some simple characteristics of the user, the clicking time period of the user and the clicking position of the user, models the characteristics of the user, the clicking time period and the clicking position, and effectively improves the accuracy of prediction of the prediction model.
Drawings
FIG. 1 is a flowchart illustrating an embodiment of a click rate prediction method according to the present application;
FIG. 2 is a flow chart illustrating the step S13 of FIG. 1 according to an embodiment;
FIG. 3 is a schematic diagram of a prediction model according to an embodiment of the present application;
FIG. 4 is a flowchart illustrating a click rate prediction model according to an embodiment of the present application;
FIG. 5 is a flowchart illustrating the step S43 of FIG. 4 according to an embodiment;
FIG. 6 is a schematic structural diagram of an embodiment of the intelligent terminal of the present application;
fig. 7 is a schematic structural diagram of an embodiment of a computer readable storage medium according to the present application.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be noted that the following differences exist between the user in the live scene and the user in the shopping scene: 1) The anchor of the live scene is relatively more concentrated, with the vast majority of users being gathered in the few head anchor's living room. According to the data display of a certain platform, the first 1% of the anchor attracts more than 80% of flow, which means that most of the interests of users in a live scene are concentrated on the anchor, and the mining requirement on long-tail data is weaker than that of an electronic commerce platform. As the number of anchor to be described is not large, some simple features of users are constructed, and a certain effect can be achieved by using machine learning models such as XG-boost and the like. 2) In a live scene, the interests of users in different time periods vary greatly. The purposes of users opening a live platform at different time periods may vary from one host to another, due to the effects of the anchor opening and the ecology of the platform. 3) In a live scene, the interests of users in different pages also have great difference, and the users in specific pages indicate that the users are only interested in specific contents, such as the users in the specific category hero alliance pages only are interested in the anchor for playing the hero alliance contents.
The application provides a click rate prediction method which is applied to a live broadcast scene according to a specific scene of the live broadcast platform. Referring to fig. 1, fig. 1 is a flowchart illustrating an embodiment of a click rate prediction method according to the present application. As shown in fig. 1, includes:
Step S11: and acquiring the behavior of the user in the historical time period, a behavior time point corresponding to the behavior, a target number to be predicted and a time point to be predicted.
The behavior of the user comprises click data generated by clicking the target by the user and state data of the clicked target by the user. In this embodiment, the target refers to the anchor of the live platform, and the target number to be predicted refers to the anchor number to be predicted. The point in time to be predicted is the click time to point to click on the anchor to be predicted.
Specifically, click data generated by a user click anchor includes: a user clicks on the anchor ID; a category ID clicked by the user, wherein the category ID includes a category in which the anchor is located, for example: the prince glows, hero alliance, etc.; the location ID that the user clicks on, the location ID including the click on which page.
It should be noted that, the behavior of the user further includes behavior data statistics of the user, specifically including: viewing days, total viewing time, maximum single day viewing time, average day viewing time, number of viewing anchor, number of viewing categories, number of access pages, seven historical day access, previous 10 category access, anchor subscription number, etc., and anchor behavior data statistics related to user behavior data statistics, such as: the method comprises the steps of hosting level, live time length, live time number, live time hour number, total user watching time length, maximum user watching time length, average user watching time length and the like, and counting behavior data of a user on the hosting, and comprises the following steps: viewing days, viewing duration, single day maximum viewing duration, average viewing duration, access volume, number of starts to view at different times, total viewing duration for the category in which the anchor is located, access volume to the anchor for historical time periods, and the like.
In this embodiment, the historical time period includes a custom preset time period, for example, a time period of seven days in the past or other preset time periods are selected as the historical time period, and the behavior data of the user in the past seven days are specifically collected, which is not limited herein.
Step S12: and preprocessing the behaviors and the behavior time points to obtain behavior characteristic information of the user.
The method specifically comprises the following steps: and carrying out association processing on the acquired behaviors of the user and the behavior time points to obtain behavior characteristic information of behaviors made by the user at the behavior time points. Wherein the behavior time point is within a preset time period. The behavior feature information includes sequence features and statistical features.
It should be noted that: the sequence feature refers to a sequence formed by the change of the user behavior along with the time, and in this embodiment, the sequence feature includes: and in a preset time period, the content clicked by the user for the first fifty times. The statistical features are statistical data of historical behavior of the user, and in this embodiment, the statistical features include: and in a preset time period, the state information of the user, the state information of the anchor to be predicted and the statistical data of the behavior of the user to the anchor to be predicted.
Specifically, in the present embodiment, the sequence features include: a) The number of the object clicked by the user, such as the anchor ID clicked by the user; b) The category number clicked by the user, such as the category ID clicked by the user (the category in which the anchor is located); c) A location number clicked by the user, such as a location ID clicked by the user; d) The time/hour of the user click.
The statistical features include: a) Status data of the user, for example: subscription anchor number, user level, etc.; b) Statistics of behavior data of users and characterization of recent activity of users, for example: viewing days, total viewing time, maximum single day viewing time, average day viewing time, number of viewing anchor, number of viewing categories, number of pages accessed, seven-day history of access, 10 previous categories of access, etc.; c) Status data of the anchor, for example: category, anchor level, anchor subscription number, etc.; d) Recent behavior data statistics of the anchor, such as: live time length, live days, live hours, total user viewing time length, maximum user viewing time length, average user viewing time length and the like; e) The behavior data statistics of the user on the anchor comprise: viewing days, viewing duration, single day maximum viewing duration, average viewing duration, access volume, number of times of starting viewing at different times, total viewing duration of the category in which the anchor is located, access volume of the history period to the anchor.
Step S13: and predicting the click probability of clicking the target to be predicted at the time point to be predicted by the user based on the behavior characteristic information by using the prediction model.
The method specifically comprises the following steps: and respectively carrying out network training on the sequence features of the user by using a first model of the prediction model to obtain first feature information, carrying out network training on the statistical features of the user by using a second model of the prediction model to obtain second feature information, carrying out splicing processing on the first feature information and the second feature information by using a splicing layer network to obtain a hidden interest vector of the user, and carrying out prediction processing on the hidden interest vector by using a full-connection layer and a normalized index function to obtain click probability of a target to be predicted. Wherein the first model and the second model are cascaded with each other. The first model comprises a DIEN model and the second model comprises an FM model. The DIEN model captures the interests of the user from the user's rich historical behavior and gives recommended items that meet the user's interests. The FM model combines the first-order features two by two to generate a second-order feature, and predicts by using the first-order and second-order features at the same time.
Specifically, referring to fig. 2, fig. 2 is a flow chart illustrating an embodiment of step S13 in fig. 1. As shown in fig. 2, includes:
Step S21: and extracting the characteristic information of the sequence characteristic by using a first model of the prediction model to obtain the first characteristic information.
Wherein the first model comprises a DIEN model. Specifically, extracting feature information of the sequence feature using the first model includes: step 1: the method comprises the steps of mapping sequence features into sequence feature vectors by using a feature extraction network, wherein a feature extraction network layer comprises an embedded layer network (DIEN Embedding Layer) and a word vector correlation model (word 2 vec), specifically, training a anchor ID by using the word vector correlation model to obtain a representation vector of the anchor ID, and mapping the input anchor ID, class ID, position ID, time and the like into the sequence feature vectors by using the embedded layer network, wherein the sequence feature vectors comprise feature information of the sequence features. Step 2: training the sequence feature vector by using a gate control loop network (GRU) and an Attention mechanism (Attention) to obtain a behavior interest vector of the user. Specifically, firstly, the door control loop network is utilized to correlate the behavior and the behavior time point of the user to form a behavior time sequence of the user, and then the door control loop network and the attention mechanism are utilized to train the behavior time sequence of the user, the target vector to be predicted and the time point vector to be predicted to obtain the behavior interest vector of the user. Step 3: and then splicing the behavior interest vector, the target vector to be predicted and the time point vector to be predicted, and training by using a full-connection layer network (Dense) to obtain first characteristic information. The embedded layer network (DIEN Embedding Layer), the word vector correlation model (word 2 vec), the gate control loop network (GRU), the Attention mechanism (Attention), the splicing layer network (Concat) and the full connection layer network (Dense) are all part of the network layers of the prediction model.
Step S22: and extracting the feature information of the statistical features by using a second model of the prediction model to obtain second feature information.
Wherein the second model comprises an FM model.
It should be noted that, compared with the conventional recommendation algorithm (such as a collaborative filtering algorithm), the factorization model (FM Layer) combines the first-order features two by two to generate the second-order features, and predicts using the first-order and second-order features at the same time. The second-order features extracted from the FM model can obviously increase the representation capability of the model, and a more accurate prediction result is obtained; on the other hand, the computational effort required for the approximate solution obtained using matrix decomposition is also not great. Therefore, FM has a wide range of applications in recommendation systems.
Step S23: and performing splicing processing on the first characteristic information and the second characteristic information to obtain behavior characteristic information of the user.
Specifically, the first characteristic information and the second characteristic information are spliced by using a splicing layer network (Concat).
In this embodiment, the first model and the second model are cascaded with each other.
Step S24: and predicting the behavior characteristic information of the user by using the full-connection layer network and the normalized exponential function to obtain the click probability of the target to be predicted.
For a more clear explanation of the prediction model in the present application, please further refer to fig. 3, fig. 3 is a schematic structural diagram of an embodiment of the prediction model in the present application. As shown in fig. 3, a behavior sequence, a behavior time point, a target to be predicted, and a time point to be predicted of a user are obtained, where the time point to be predicted is a click time of the user in a next period, for example, the user opens a live platform at 12 points, and predicts a click probability of the user clicking a host to be predicted at 12 points, and the time point to be predicted also includes a time period to be predicted, for example, a time period from 12 points to 1 point, which is not limited herein.
And converting the behavior sequence, the behavior time point, the target to be predicted and the time point to be predicted of the user into vectors by using an embedded network (DIEN Embedding Layer), and carrying out association processing on the behavior sequence and the behavior time point to form a behavior sequence vector. And then, processing the behavior sequence of the user by using a gate control loop network (GRU) to obtain a hidden interest vector h (T) of the user. And then inputting the hidden interest vector, the target vector to be predicted and the time point vector to be predicted into a gate control loop network (GRU) and an Attention mechanism (Attention), and performing prediction training on a behavior sequence of the user by using the gate control loop network (GRU) and the Attention mechanism (Attention) to obtain a behavior interest vector h' (T) of the user. Finally, the behavior interest vector h' (T), the target vector to be predicted and the time point vector to be predicted are spliced by using a splicing layer network (concat), and then input into a full-connection layer network (Dense) for preliminary processing, so that first characteristic information of a user is obtained.
On the other hand, a statistical feature vector of the user is obtained, wherein the statistical feature is statistical data of historical behaviors of the user, and the statistical feature can be obtained by a statistical system of the live platform. And preprocessing the statistical feature vector of the user by utilizing an factorization model (FM Layer) to obtain second feature information of the user.
And performing splicing processing on the first characteristic information and the second characteristic information by using a splicing layer network (concat) to obtain behavior characteristic information of the user, and performing prediction processing on the behavior characteristic information of the user by using a full-connection layer network (Dense) and a normalized exponential function (Softmax) to obtain the click probability of clicking the target to be predicted by the user at the time point to be predicted.
The beneficial effects of the application are as follows: the behavior of the user and the behavior time point of the user are preprocessed by acquiring the behavior of the user in the historical time period, the behavior time point corresponding to the behavior, the target number to be predicted and the time point to be predicted, behavior characteristic information of the user is obtained, and the click probability of the user clicking the target to be predicted (anchor) at the time point to be predicted is predicted by using a prediction model. The click probability of the user on different anchor at specific time and under the page is predicted by the prediction model, so that the anchor interested by the user is recommended to the user, the user experience of the user for watching live broadcast is ensured, and the retention rate of the user in the live broadcast platform is improved. In addition, compared with other shopping prediction models, the prediction model provided by the application considers some simple characteristics of the user, the clicking time period of the user and the clicking position of the user, models the characteristics of the user, the clicking time period and the clicking position, and effectively improves the accuracy of prediction of the prediction model.
The application further provides a training method of the click rate prediction model, please further refer to fig. 4, fig. 4 is a flow chart of an embodiment of the click rate prediction model of the application. As shown in fig. 4, includes:
step S41: and collecting behavior data of the user in a historical time period and the real target number clicked at the time point to be predicted.
Wherein the behavior data includes behaviors of the user in the history period and behavior time points corresponding to the behaviors. The behavior comprises click data generated by clicking the target by a user and state data of the clicked target by the user. The method specifically comprises the following steps: click data such as a host ID clicked by a user, a category ID clicked by the user, a location ID clicked by the user, and a time clicked by the user, as well as state information of the user, state information of the host to be predicted, behavior information of the user to be predicted, and the like.
Step S42: and inputting the behavior data of the historical time period into an initial model, and predicting the click target of the user at the time point to be predicted based on the behavior data of the historical time period by using the initial model to obtain a predicted target number.
The method specifically comprises the following steps: the behavior data of the historical time period is mapped into a sequence feature and a statistical feature. Wherein, the sequence feature and the statistical feature are both vector representations. The initial model performs predictive training based on the vector representations of the sequence features and the statistical features. Specifically, the method comprises the steps of performing feature extraction on sequence features by using a DIEN model to obtain first feature information, performing feature extraction on statistical features by using an FM model to obtain second feature information, performing splicing processing on the first feature information and the second feature information, and predicting the spliced first feature information and second feature information by using a full-connection layer and a normalized index function to obtain a predicted target (anchor) number.
Step S43: training the initial model by using the real target number and the predicted target number, and determining the model after training as a predicted model.
Specifically, until the predicted target number predicted by the initial model is the same as the target number actually clicked by the user, the training is determined to be completed, and the model after the training is determined to be the predicted model.
In order to ensure the accuracy of prediction of the prediction model in the present application, the present application trains the model by detecting samples, further referring to fig. 5, fig. 5 is a flowchart of a specific embodiment of step S43 in fig. 4. As shown in fig. 5, includes:
Step S51: and detecting the trained model through detection samples.
The detection sample comprises behavior data of multiple groups of users in a historical time period and real target numbers clicked at a time point to be predicted. And respectively training the trained models by using each group of detection samples, and recording the prediction accuracy of the models.
Step S52: and determining that model training is completed in response to the prediction precision of the trained model reaching the preset precision.
The preset precision comprises the following steps: the accuracy of the prediction reaches more than 70%.
The beneficial effects of this embodiment are: the method comprises the steps of collecting behavior data of a user in a historical time period and real target numbers clicked at a time point to be predicted, inputting the behavior data of the historical time period into an initial model, predicting a clicked target of the user at the time point to be predicted based on the behavior data of the historical time period by using the initial model to obtain a predicted target number, training the initial model by using the real target number and the predicted target number, determining the trained model as a predicted model, and accordingly establishing the predicted model to be applied to prediction of click rate of a live broadcast platform on a target anchor.
The application also utilizes an experimental data set comprising 600w constructed by a certain live broadcast platform to detect the effect of the FM model, the DIEN model and the prediction model in the application. Specifically, the experimental data set was prepared according to 7:3, dividing the model into a training sample data set and a detection sample data set, respectively training an FM model, a DIEN model and a prediction model in the application by using the training sample data set to obtain the FM prediction model, the DIEN prediction model and the prediction model in the application, detecting the prediction model by using the detection sample data set, and measuring an AUC index and a precision index. The results are shown in the following table:
from the table, the AUC index and the precision index of the prediction model in the application can reach more than 90%, which shows that the prediction model in the application has better prediction effect on the target anchor when applied to the live broadcast scene.
The present application further provides an intelligent terminal, please refer to fig. 6, fig. 6 is a schematic structural diagram of an embodiment of the intelligent terminal of the present application, as shown in fig. 6, the intelligent terminal 60 includes a processor 601 and a memory 602 coupled to each other, and the processor 601 is configured to execute program instructions stored in the memory 602 to implement steps in any of the above method embodiments or steps correspondingly executed by a client in any of the above method embodiments. The terminal may include, in addition to the above-described processor 601 and memory 602, a touch screen, a printing component, a communication circuit, etc., as required, which is not limited herein.
In particular, the processor 601 is arranged to control itself and the memory 602 to implement the steps of any of the method embodiments described above. The processor 601 may also be referred to as a CPU (Central Processing Unit ). The processor 601 may be an integrated circuit chip with signal processing capabilities. The Processor 601 may also be a general purpose Processor, a digital signal Processor (DIGITAL SIGNAL Processor, DSP), an Application SPECIFIC INTEGRATED Circuit (ASIC), a Field-Programmable gate array (Field-Programmable GATE ARRAY, FPGA) or other Programmable logic device, a discrete gate or transistor logic device, a discrete hardware component. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. In addition, the processor 601 may be commonly implemented by a plurality of integrated circuit chips.
The present application also provides a computer readable storage medium, as shown in fig. 7, and fig. 7 is a schematic structural diagram of an embodiment of the computer readable storage medium of the present application.
The computer readable storage medium 70 includes a computer program 701 stored on the computer readable storage medium 70, which when executed by the processor, implements the steps of any of the method embodiments described above or the steps of the click rate prediction method of network time service in the method embodiments described above.
In particular, the integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium 70. Based on such understanding, the technical solution of the present application, or a part or all of the technical solution contributing to the prior art, may be embodied in the form of a software product stored in a computer-readable storage medium 70, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (processor) to perform all or part of the steps of the methods of the embodiments of the present application. And the aforementioned computer-readable storage medium 70 includes: a usb disk, a removable hard disk, a read-only memory (ROM), a random access memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In the several embodiments provided in the present application, it should be understood that the disclosed method and apparatus may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of modules or units is merely a logical functional division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical, or other forms.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiment.
In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (processor) to execute all or part of the steps of the methods of the embodiments of the present application. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a read-only memory (ROM), a random access memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The foregoing is only the embodiments of the present application, and therefore, the patent scope of the application is not limited thereto, and all equivalent structures or equivalent processes using the descriptions of the present application and the accompanying drawings, or direct or indirect application in other related technical fields, are included in the scope of the application.
Claims (6)
1. The click rate prediction method is characterized by comprising the following steps of:
Collecting behaviors of a user in a historical time period, a behavior time point corresponding to the behaviors, a target number to be predicted and a time point to be predicted; the behavior of the user comprises click data generated by clicking a target by the user and state data of the user and the clicked target;
Preprocessing the behaviors and the behavior time points to obtain behavior characteristic information of the user; the method specifically comprises the following steps: mapping the behavior and the behavior time points into sequence features and statistical features;
Collecting behavior data of a user in a historical time period and a real target number clicked at the time point to be predicted; wherein the behavior data comprises behaviors of a user in a historical time period and behavior time points corresponding to the behaviors; the behavior data of the historical time period is input into an initial model, and a click target of a user at the time point to be predicted is predicted by utilizing the initial model based on the behavior data of the historical time period, so that a predicted target number is obtained; training the initial model by utilizing the real target number and the predicted target number, and determining the model after training as a predicted model;
Extracting feature information of the sequence features by using a first model of the prediction model to obtain first feature information; the first model includes a DIEN model; the method specifically comprises the following steps: mapping the sequence features into sequence feature vectors by using a feature extraction network; the sequence feature vector comprises a behavior sequence vector, a behavior sequence time vector, a target vector to be predicted and a time point vector to be predicted of a user; training the sequence feature vector by using a door control loop network and an attention mechanism to obtain a behavior interest vector of a user; splicing the behavior interest vector, the target vector to be predicted and the time point vector to be predicted, and training by using a full-connection layer network to obtain the first characteristic information;
extracting feature information of the statistical features by using a second model of the prediction model to obtain second feature information; the second model comprises an FM model; wherein the first model and the second model are cascaded with each other;
splicing the first characteristic information and the second characteristic information to obtain behavior characteristic information;
predicting the behavior characteristic information by using a full-connection layer network and a normalized exponential function to obtain the click probability of the user clicking the target to be predicted at the time point to be predicted; the time point to be predicted is a click time for clicking the target to be predicted, and the click time comprises a click time point or a click time period in one day.
2. The click-through rate prediction method of claim 1, wherein the sequence feature comprises a target number clicked by a user, a category number clicked by a user, a position number clicked by a user, or a time/hour clicked by a user.
3. The click-through rate prediction method of claim 1, wherein the statistical features include state information of a user, state information of a target, and behavior information of the user to the target to be predicted.
4. The click-through rate prediction method according to claim 1, wherein the training of the initial model using the real object number and the predicted object number and determining the trained model as a predicted model comprises:
Detecting the model after training is completed through a detection sample;
And determining that the model training is completed in response to the prediction precision of the trained model reaching a preset precision.
5. An intelligent terminal, characterized in that the intelligent terminal comprises a processor and a memory, which are coupled to each other, the memory is used for storing program instructions, and the processor is used for executing the program instructions stored in the memory to implement the click rate prediction method according to any one of claims 1-4.
6. A computer readable storage medium comprising a processor and a memory, the memory storing computer program instructions, the processor configured to execute the program instructions to implement the click rate prediction method of any one of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110565423.8A CN113297486B (en) | 2021-05-24 | 2021-05-24 | Click rate prediction method and related device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110565423.8A CN113297486B (en) | 2021-05-24 | 2021-05-24 | Click rate prediction method and related device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113297486A CN113297486A (en) | 2021-08-24 |
| CN113297486B true CN113297486B (en) | 2024-04-19 |
Family
ID=77324243
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110565423.8A Active CN113297486B (en) | 2021-05-24 | 2021-05-24 | Click rate prediction method and related device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113297486B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113965805A (en) * | 2021-10-22 | 2022-01-21 | 北京达佳互联信息技术有限公司 | Prediction model training method and device and target video editing method and device |
| CN114710686A (en) * | 2022-03-11 | 2022-07-05 | 武汉斗鱼鱼乐网络科技有限公司 | Exposure method, device, medium and equipment for live broadcast room |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110020877A (en) * | 2018-01-09 | 2019-07-16 | 腾讯科技（深圳）有限公司 | The prediction technique of clicking rate, the determination method of clicking rate and server |
| CN110858912A (en) * | 2018-08-22 | 2020-03-03 | 中国电信股份有限公司 | Streaming media caching method and system, caching policy server and streaming service node |
| CN110929206A (en) * | 2019-11-20 | 2020-03-27 | 腾讯科技（深圳）有限公司 | Click rate estimation method and device, computer readable storage medium and equipment |
| CN111539512A (en) * | 2020-03-31 | 2020-08-14 | 北京三快在线科技有限公司 | Behavior prediction method and device, electronic equipment and readable storage medium |
-
2021
- 2021-05-24 CN CN202110565423.8A patent/CN113297486B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110020877A (en) * | 2018-01-09 | 2019-07-16 | 腾讯科技（深圳）有限公司 | The prediction technique of clicking rate, the determination method of clicking rate and server |
| CN110858912A (en) * | 2018-08-22 | 2020-03-03 | 中国电信股份有限公司 | Streaming media caching method and system, caching policy server and streaming service node |
| CN110929206A (en) * | 2019-11-20 | 2020-03-27 | 腾讯科技（深圳）有限公司 | Click rate estimation method and device, computer readable storage medium and equipment |
| CN111539512A (en) * | 2020-03-31 | 2020-08-14 | 北京三快在线科技有限公司 | Behavior prediction method and device, electronic equipment and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113297486A (en) | 2021-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Karimi et al. | News recommender systems–Survey and roads ahead | |
| CN110941740B (en) | Video recommendation method and computer-readable storage medium | |
| CN110781321B (en) | Multimedia content recommendation method and device | |
| CN111159564A (en) | Information recommendation method and device, storage medium and computer equipment | |
| CN105023165A (en) | Method, device and system for controlling release tasks in social networking platform | |
| CN113297486B (en) | Click rate prediction method and related device | |
| EP4083857A1 (en) | Information prediction model training method and apparatus, information prediction method and apparatus, storage medium, and device | |
| CN110472154A (en) | A kind of resource supplying method, apparatus, electronic equipment and readable storage medium storing program for executing | |
| CN108475381A (en) | The method and apparatus of performance for media content directly predicted | |
| CN111400586A (en) | Group display method, terminal, server, system and storage medium | |
| CN112100221A (en) | Information recommendation method and device, recommendation server and storage medium | |
| CN115618024A (en) | Multimedia recommendation method and device and electronic equipment | |
| CN107944026A (en) | A kind of method, apparatus, server and the storage medium of atlas personalized recommendation | |
| CN114245185B (en) | Video recommendation method, model training method, device, electronic equipment and medium | |
| CN112269943B (en) | Information recommendation system and method | |
| CN113722588A (en) | Resource recommendation method and device and electronic equipment | |
| CN113204699B (en) | Information recommendation method and device, electronic equipment and storage medium | |
| CN116821475B (en) | Video recommendation method and device based on client data and computer equipment | |
| CN115374348A (en) | Information recommendation method, information recommendation device and readable storage medium | |
| CN113297417B (en) | Video pushing method, device, electronic equipment and storage medium | |
| CN111143688B (en) | Evaluation method and system based on mobile news client | |
| CN115344774A (en) | User account screening method and device and server | |
| CN114329055A (en) | Search recommendation method and recommendation device, electronic device and storage medium | |
| CN111127057B (en) | Multi-dimensional user portrait recovery method | |
| CN114461919A (en) | Information recommendation model training method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |