CN112381607B - Network commodity ordering method, device, equipment and medium - Google Patents
Network commodity ordering method, device, equipment and medium Download PDFInfo
- Publication number
- CN112381607B CN112381607B CN202011264017.XA CN202011264017A CN112381607B CN 112381607 B CN112381607 B CN 112381607B CN 202011264017 A CN202011264017 A CN 202011264017A CN 112381607 B CN112381607 B CN 112381607B
- Authority
- CN
- China
- Prior art keywords
- commodity
- click
- sequence
- conversion rate
- rate
- 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 38
- 238000006243 chemical reaction Methods 0.000 claims abstract description 183
- 230000006399 behavior Effects 0.000 claims abstract description 42
- 238000004590 computer program Methods 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 7
- 238000012163 sequencing technique Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 6
- 230000009471 action Effects 0.000 description 5
- 238000004891 communication Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 238000011156 evaluation Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
Classifications
-
- 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
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- 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
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0641—Shopping interfaces
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a network commodity ordering method, device, equipment and medium. The method comprises the following steps: acquiring user behavior data and commodity exposure data; the user behavior data includes a purchase intent sequence and a historical click sequence; inputting user behavior data and commodity exposure data into a multi-target pre-estimation model constructed based on a multi-task learning framework to obtain corresponding commodity conversion rate; the multi-target pre-estimation model comprises a commodity click rate learning task and a commodity conversion rate learning task; the embedded layer parameters and part of attention parameters are shared between the commodity click rate learning task and the commodity conversion rate learning task; inputting the historical click sequence and commodity exposure data into a click rate estimation model to obtain a corresponding commodity click rate; the commodity conversion rate and the commodity click rate are used for determining the click conversion probability of the commodity, and the commodity is ordered based on the click conversion probability, so that the commodity can be ordered according to the comprehensive quality of the commodity, and further the commodity conversion rate and the total exchange amount are improved.
Description
Technical Field
The present application relates to the field of online shopping, and in particular, to a method, apparatus, device, and medium for ordering online commodities.
Background
Currently, the click Rate (i.e., CTR, click Through Rate) of a web mall commodity is not completely proportional to the Conversion Rate (i.e., CVR, conversion Rate), which represents the probability that an exposed commodity is clicked, and the Conversion Rate represents the probability that a clicked commodity is purchased. For example, the design of the cover of a certain commodity is beautiful, the clicking rate of the commodity is high, but the commodity cannot be purchased after being clicked due to the fact that the actual quality of the commodity does not accord with the commodity description; or the cover of a certain commodity is not attractive enough, the probability of being clicked by a user is low, but the quality, evaluation, historical sales and other indexes of the commodity are good, and once the user clicks and enters a commodity detail page, a series of historical performances of the commodity are observed, the probability of being purchased by the user is high. Therefore, how to prioritize high quality goods, achieving maximum sales is a major problem currently faced.
In the prior art, the conversion rate of the commodity is predicted through the model, but the conversion rate prediction model is used for predicting the conversion rate of the commodity by assuming that the commodity is clicked, namely the traditional conversion rate prediction model usually uses click data as a training set, wherein the clicked but unconverted commodity is a negative example, and the clicked and converted commodity is a positive example, so that the problem of sample deviation exists, and the generalization capability of the model is reduced. In the prior art, click and Conversion probability (namely ctvr) of the commodity is directly predicted Through the ESMM network structure, however, in practical application, when on-line sorting is performed according to the output ctvr score, the Click Rate is greatly reduced, the Click Rate of the commodity is reduced, and further the Conversion Rate and assembly exchange of the commodity are affected.
Disclosure of Invention
Accordingly, the present application is directed to a method, apparatus, device and medium for ordering network commodities, which can order the commodities according to their comprehensive quality, thereby improving commodity conversion rate and assembly exchange. The specific scheme is as follows:
in a first aspect, the application discloses a network commodity ordering method, which comprises the following steps:
acquiring user behavior data and commodity exposure data; the user behavior data comprises a purchase intention sequence and a historical click sequence;
inputting the user behavior data and the commodity exposure data into a multi-target estimated model constructed based on a multi-task learning framework to obtain corresponding commodity conversion rate; the multi-target pre-estimation model comprises a commodity click rate learning task and a commodity conversion rate learning task; the commodity click rate learning task and the commodity conversion rate learning task share an embedded layer parameter and a part of attention parameter;
inputting the historical click sequence and the commodity exposure data into a click rate estimation model to obtain a corresponding commodity click rate;
and determining click conversion probability of the commodity based on the commodity conversion rate and the commodity click rate, and sequencing the commodity according to the click conversion probability.
Optionally, the acquiring the user behavior data includes:
constructing an operation sequence based on the purchasing operation and the collecting operation of the user and through a historical footprint table and/or a secondary clicking operation of the favorites and/or the shopping cart;
sorting the operation sequences according to the sequence from near to far between each operation and the current time to obtain the purchase intention sequence;
the historical click sequence is constructed based on a user's click operation.
Optionally, the inputting the user behavior data and the commodity exposure data into a multi-objective pre-estimation model constructed based on a multi-task learning framework includes:
inputting the commodity exposure data into the commodity click rate learning task and the commodity conversion rate learning task;
configuring non-sharing parameters for the purchase intention sequence, and inputting the purchase intention sequence into the commodity click rate learning task and the commodity conversion rate learning task;
and configuring sharing parameters for the historical click sequence, and inputting the historical click sequence into the commodity click rate learning task and the commodity conversion rate learning task so that the commodity click rate learning task and the commodity conversion rate learning task share the attention parameters of the historical click sequence.
Optionally, the process for constructing the multi-objective pre-estimation model includes:
based on the ESMM network structure, a dual-task learning model sharing the embedded layer parameters and the hidden layer partial parameters is constructed, so that the multi-target pre-estimated model is obtained.
Optionally, the inputting the user behavior data and the commodity exposure data into a multi-objective pre-estimation model constructed based on a multi-task learning framework includes:
acquiring commodity characteristics of commodities; the commodity characteristics comprise any one or more of commodity categories, commodity prices, commodity sales and commodity praise;
corresponding the commodity characteristics to commodities in the historical click sequence to obtain a historical click sequence containing commodity characteristics;
and inputting the historical click sequence containing commodity characteristics, the purchase intention sequence and the commodity exposure data into the multi-target predictive model.
Optionally, the determining the click conversion probability of the commodity based on the commodity conversion rate and the commodity click rate includes:
performing constraint operation on the commodity conversion rate by using a preset value range control function to obtain the constrained commodity conversion rate;
multiplying the constrained commodity conversion rate by the commodity click rate to obtain the click conversion probability.
In a second aspect, the present application discloses a network commodity ordering apparatus, including:
the data acquisition module is used for acquiring user behavior data and commodity exposure data; the user behavior data comprises a purchase intention sequence and a historical click sequence;
the conversion rate determining module is used for inputting the user behavior data and the commodity exposure data into a multi-target estimated model constructed based on a multi-task learning frame to obtain corresponding commodity conversion rate; the multi-target pre-estimation model comprises a commodity click rate learning task and a commodity conversion rate learning task; the commodity click rate learning task and the commodity conversion rate learning task share an embedded layer parameter and a part of attention parameter;
the click rate determining module is used for inputting the historical click sequence and the commodity exposure data into a click rate estimation model to obtain a corresponding commodity click rate;
and the click conversion probability determining module is used for determining the click conversion probability of the commodity based on the commodity conversion rate and the commodity click rate and sequencing the commodity according to the click conversion probability.
Optionally, the conversion rate determining module further includes:
the feature acquisition unit is used for acquiring commodity features of commodities; the commodity characteristics comprise any one or more of commodity categories, commodity prices, commodity sales and commodity praise;
the historical click sequence construction unit is used for corresponding the commodity characteristics to commodities in the historical click sequence to obtain a historical click sequence containing commodity characteristics;
and the data input unit is used for inputting the historical click sequence containing commodity characteristics, the purchase intention sequence and the commodity exposure data into the multi-target estimated model.
In a third aspect, the present application discloses an electronic device, comprising:
a memory for storing a computer program;
and the processor is used for executing the computer program to realize the network commodity ordering method.
In a fourth aspect, the present application discloses a computer-readable storage medium for storing a computer program; wherein the computer program when executed by the processor implements the network commodity ordering method described above.
The application discloses a network commodity ordering method which comprises the following steps: acquiring user behavior data and commodity exposure data; the user behavior data comprises a purchase intention sequence and a historical click sequence; inputting the user behavior data and the commodity exposure data into a multi-target estimated model constructed based on a multi-task learning framework to obtain corresponding commodity conversion rate; the multi-target pre-estimation model comprises a commodity click rate learning task and a commodity conversion rate learning task; the commodity click rate learning task and the commodity conversion rate learning task share an embedded layer parameter and a part of attention parameter; inputting the historical click sequence and the commodity exposure data into a click rate estimation model to obtain a corresponding commodity click rate; and determining click conversion probability of the commodity based on the commodity conversion rate and the commodity click rate, and sequencing the commodity according to the click conversion probability.
It can be seen that, through inputting the purchase intention sequence, the historical click sequence and the commodity exposure data into a multi-target estimation model comprising a commodity click rate learning task and a commodity conversion rate learning task, and because the commodity click rate learning task and the commodity conversion rate learning task share the embedded layer parameters and part of the attention parameters, all samples, namely the commodity exposure data, are utilized for estimation, and through common learning between the commodity click rate learning task and the commodity conversion rate learning task, more accurate commodity conversion rate can be obtained, and finally, the click conversion probability of the commodity is determined according to the commodity conversion rate obtained based on the multi-target estimation model and the commodity click rate obtained based on the click rate estimation model, and the commodity is ordered according to the click conversion probability. Therefore, the optimal ordering sequence of the commodities can be determined, the commodity conversion rate is improved while the commodity click rate is ensured, and the total amount of the commodities is further improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present application, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for ordering network commodities provided by the application;
FIG. 2 is a schematic diagram of a portion of a network architecture for a single task provided by the present application;
FIG. 3 is a flowchart of a specific method for ordering network commodities according to the present application;
FIG. 4 is a schematic diagram of a network commodity sorting apparatus according to the present application;
fig. 5 is a block diagram of an electronic device according to the present application.
Detailed Description
In the prior art, the conversion rate of the commodity is predicted through the model, but the conversion rate prediction model is used for predicting the conversion rate of the commodity by assuming that the commodity is clicked, namely the traditional conversion rate prediction model usually uses click data as a training set, wherein the clicked but unconverted commodity is a negative example, and the clicked and converted commodity is a positive example, so that the problem of sample deviation exists, and the generalization capability of the model is reduced. In order to overcome the technical problems, the application provides a network commodity ordering method utilizing multi-model combination and multi-task learning, which can order according to the comprehensive quality of commodities, thereby improving the commodity conversion rate and the assembly exchange.
The embodiment of the application discloses a network commodity ordering method, which is shown in fig. 1 and can comprise the following steps:
step S11: acquiring user behavior data and commodity exposure data; the user behavior data includes a purchase intent sequence and a historical click sequence.
In this embodiment, first, user behavior data and commodity exposure data are acquired, where the user behavior data is data determined according to user behavior, and includes a purchase intention sequence and a historical click sequence.
In this embodiment, the acquiring the user behavior data may include: constructing an operation sequence based on the purchasing operation and the collecting operation of the user and through a historical footprint table and/or a secondary clicking operation of the favorites and/or the shopping cart; sorting the operation sequences according to the sequence from near to far between each operation and the current time to obtain the purchase intention sequence; the historical click sequence is constructed based on a user's click operation. The purchasing intention sequence is ordered according to the order from near to far between each operation and the current time, the relative position of the operation action which is closer to the current time in the sequence is front, and the relative position of the operation action which is farther from the current time in the sequence is back. It will be appreciated that the user's collection, clicking and purchasing operations for the merchandise, and then clicking again from the historical footprint and/or favorites and/or shopping cart, may indicate that the user's purchase intent for the merchandise is strong. Thus, model learning and prediction of such operational behavior paths as a sequence of buying intents can introduce more effective information related to the deals.
Step S12: inputting the user behavior data and the commodity exposure data into a multi-target estimated model constructed based on a multi-task learning framework to obtain corresponding commodity conversion rate; the multi-target pre-estimation model comprises a commodity click rate learning task and a commodity conversion rate learning task; and sharing the embedded layer parameters and part of attention parameters between the commodity click rate learning task and the commodity conversion rate learning task.
In this embodiment, after the user behavior data and the commodity exposure data are obtained, the data are input into a multi-target estimated model constructed based on a multi-task learning frame, and then the corresponding commodity conversion rate of the commodity is obtained through model prediction; the multi-target prediction model comprises a commodity click rate learning task and a commodity conversion rate learning task, wherein commodity click rate learning is used for predicting the click rate of commodities, commodity conversion rate learning is used for predicting the conversion rate of the commodities, and an embedding layer (namely an embedding layer) parameter and a part of attention parameter are shared between the commodity click rate learning task and the commodity conversion rate learning task.
It will be appreciated that the user's purchase intention sequence is entered into the model simultaneously with the user's historical click sequence for training, where both the user's historical click sequence and purchase intention sequence are delivered to both tasks, which take over these sequences using conventional attention mechanism structures. However, the attention parameter of the history click sequence is shared by two tasks, and the attention parameter of the purchase intention sequence is not shared between the two tasks, and is learned by each of the two tasks. And finally, the attention parameter output results of the historical click sequence and the success intention sequence are subjected to fusion, MLP full-connection layer, dimension reduction, nonlinear activation function and the like, and the final conversion rate is obtained through a commodity conversion rate learning task. It can be understood that the commodity conversion rate learning task can assist training through the parameters learned by the commodity click rate learning task, and the learning capacity of the commodity conversion rate learning task can be improved through sharing the embedded layer parameters and the attention parameters of the historical click sequence between the commodity click rate learning task and the commodity conversion rate learning task, so that the commodity conversion rate more conforming to practical application is obtained.
In this embodiment, the inputting the user behavior data and the commodity exposure data into the multi-objective prediction model constructed based on the multi-task learning framework may include: inputting the commodity exposure data into the commodity click rate learning task and the commodity conversion rate learning task; configuring non-sharing parameters for the purchase intention sequence, and inputting the purchase intention sequence into the commodity click rate learning task and the commodity conversion rate learning task; and configuring sharing parameters for the historical click sequence, and inputting the historical click sequence into the commodity click rate learning task and the commodity conversion rate learning task so that the commodity click rate learning task and the commodity conversion rate learning task share the attention parameters of the historical click sequence. It will be appreciated that the historical click sequence may be configured with shared parameters by configuring non-shared parameters for the purchase intention sequence such that the attention parameters of the historical click sequence are shared between the commodity click rate learning task and the commodity conversion learning task, but the attention parameters of the purchase intention sequence are not shared.
In this embodiment, the process of constructing the multi-objective prediction model may include: based on the ESMM network structure, a dual-task learning model sharing the embedded layer parameters and the hidden layer partial parameters is constructed, so that the multi-target pre-estimated model is obtained. It can be understood that the original embedded layer sharing can be reserved on the basis of the ESMM network structure, and meanwhile, the attention parameter sharing of the history clicking sequence is added to obtain the double-task learning model; the partial network structure of each task is shown in fig. 2, and the parameters of the embedded layer and the partial parameters in the hidden layer, that is, the embedded layer parameters and the attention parameters of the historical click sequence, are shared between the two tasks, and the parameters of the MLP full connection layer of the upper layer are not shared separately.
Step S13: and inputting the historical click sequence and the commodity exposure data into a click rate estimation model to obtain a corresponding commodity click rate.
In this embodiment, the historical click sequence and the commodity exposure data are input into a click rate estimation model at the same time, and the commodity click rate is obtained through the click rate estimation model; the click rate estimation model may be a Wide & deep fm model.
Step S14: and determining click conversion probability of the commodity based on the commodity conversion rate and the commodity click rate, and sequencing the commodity according to the click conversion probability.
In this embodiment, the click conversion probability of the commodity is determined according to the obtained commodity conversion rate and commodity click rate, and then the commodity is ranked according to the click conversion probability, so that the commodity with high click conversion rate is preferentially ranked.
From the above, in this embodiment, the buying intention sequence, the historical click sequence and the commodity exposure data are input into the multi-objective prediction model including the commodity click rate learning task and the commodity conversion rate learning task, and because the embedded layer parameters and part of the attention parameters are shared between the commodity click rate learning task and the commodity conversion rate learning task, all samples, namely the commodity exposure data, are utilized to predict, and through the joint learning between the commodity click rate learning task and the commodity conversion rate learning task, the more accurate commodity conversion rate can be obtained, and finally the click conversion probability of the commodity is determined according to the commodity conversion rate obtained based on the multi-objective prediction model and the commodity click rate obtained based on the click rate prediction model, and the commodity is ordered according to the click conversion probability. Therefore, the optimal ordering sequence of the commodities can be determined, the commodity conversion rate is improved while the commodity click rate is ensured, and the total amount of the commodities is further improved.
The embodiment of the application discloses a specific network commodity ordering method, which is shown in fig. 3 and can comprise the following steps:
step S21: acquiring user behavior data and commodity exposure data; the user behavior data includes a purchase intent sequence and a historical click sequence.
Step S22: acquiring commodity characteristics of commodities; the commodity characteristics comprise any one or more of commodity categories, commodity prices, commodity sales and commodity praise; and corresponding the commodity characteristics to commodities in the historical click sequence to obtain a historical click sequence containing commodity characteristics.
In this embodiment, the commodity features of the commodity can be obtained at the same time; the commodity features include, but are not limited to, commodity category, commodity price, commodity sales, commodity praise, commodity store and brand, and then the commodity features are corresponding to the commodities in the historical click sequence to obtain the historical click sequence containing commodity features. Specifically, the above commodity features may be first classified into 8 barrels, where each feature may be divided into 8 sections, for example, commodity prices are divided into price sections corresponding to different commodity labels; wherein, the barrel dividing boundary can be set manually; after barrel division, each type of characteristics are encoded, and a dedicated ebedding matrix is constructed and used as the supplementary information of the historical click sequence; it will be appreciated that these merchandise features are also presented in a serial fashion and are in one-to-one correspondence with each merchandise in the historical click sequence, i.e., each merchandise in the sequence contains features such as category, price, sales, and praise. Specifically, the encoded feature sequences and the ebedding expression of the historical click sequence can be spliced in the last dimension.
Thus, the commodity characteristics are used as the supplementary information of the historical click sequence, and hidden information of the user, such as category preference, price preference, whether to pay attention to the evaluation star grade or the evaluation quantity of the commodity, whether to pay attention to the commodity sales and the popular preference degree, can be predicted according to the user behavior. Together with the historical click sequence, the information can help the model learn commodity expression, and further enhance the capability of commodity conversion rate task sides.
Step S23: inputting the historical click sequence containing commodity characteristics, the purchase intention sequence and the commodity exposure data into a multi-target pre-estimation model to obtain corresponding commodity conversion rate; the multi-target pre-estimation model comprises a commodity click rate learning task and a commodity conversion rate learning task; and sharing the embedded layer parameters and part of attention parameters between the commodity click rate learning task and the commodity conversion rate learning task.
In this embodiment, the obtained historical click sequence, purchase intention sequence and commodity exposure data containing commodity features are input into a multi-objective pre-estimation model, and then the corresponding commodity conversion rate of the commodity is obtained through model prediction; the multi-target prediction model comprises a commodity click rate learning task and a commodity conversion rate learning task, wherein commodity click rate learning is used for predicting the click rate of commodities, the commodity conversion rate learning task is used for predicting the conversion rate of the commodities, and embedded layer parameters and partial attention parameters are shared between the commodity click rate learning task and the commodity conversion rate learning task.
Step S24: and inputting the historical click sequence and the commodity exposure data into a click rate estimation model to obtain a corresponding commodity click rate.
Step S25: and performing constraint operation on the commodity conversion rate by using a preset value range control function to obtain the constrained commodity conversion rate.
In this embodiment, after the commodity conversion rate is obtained, constraint operation is performed on the commodity conversion rate by using a preset value range control function, so as to obtain the constrained commodity conversion rate; specifically, the value range constraint formula of commodity conversion rate is:
SigmoidCVR＝1/(alpha+e^(-beta*CVR))；
wherein SigmoidCVR is the commodity conversion rate after constraint, alpha can be 0.5, and beta can be 1.1. It can be appreciated that by performing value range constraint on commodity conversion rate output by the multi-target pre-estimation model, the influence of the commodity conversion rate on the whole value range can be reduced.
Step S26: multiplying the constrained commodity conversion rate by the commodity click rate to obtain the click conversion probability, and sequencing the commodities according to the click conversion probability.
In this embodiment, after the constrained commodity conversion rate is obtained, multiplying the constrained commodity conversion rate by the commodity click rate to obtain a click conversion probability ctvr of the commodity, where ctvr=ctr×sigmoidcvr, and then sorting the commodity according to the click conversion probability, where the commodity with high click conversion rate is preferentially sorted.
For the specific process of step S21 and step S24, reference may be made to the corresponding content disclosed in the foregoing embodiment, and no further description is given here.
From the above, in this embodiment, the historical click sequence including the commodity feature is obtained by acquiring the commodity feature of the commodity and corresponding the commodity feature to the commodity in the historical click sequence; then inputting the historical click sequence, the purchase intention sequence and the commodity exposure data containing commodity characteristics into a multi-target pre-estimation model to obtain corresponding commodity conversion rate; and performing constraint operation on the commodity conversion rate by using a preset value range control function to obtain a constrained commodity conversion rate, multiplying the constrained commodity conversion rate by the commodity click rate to obtain click conversion probability, and sequencing the commodities according to the click conversion probability. By utilizing commodity characteristics as the supplementary information of the historical click sequence, the model can be helped to learn commodity expression, and the capability of the commodity conversion rate task side is further enhanced. And the influence of the commodity conversion rate on the whole value range can be reduced by carrying out constraint operation on the obtained commodity conversion rate, so that the determined click conversion probability is more in line with the actual application scene.
Correspondingly, the embodiment of the application also discloses a network commodity sorting device, which is shown in fig. 4 and comprises the following components:
a data acquisition module 11 for acquiring user behavior data and commodity exposure data; the user behavior data comprises a purchase intention sequence and a historical click sequence;
the conversion rate determining module 12 is configured to input the user behavior data and the commodity exposure data into a multi-objective pre-estimation model constructed based on a multi-task learning framework, so as to obtain a corresponding commodity conversion rate; the multi-target pre-estimation model comprises a commodity click rate learning task and a commodity conversion rate learning task; the commodity click rate learning task and the commodity conversion rate learning task share an embedded layer parameter and a part of attention parameter;
the click rate determining module 13 is configured to input the historical click sequence and the commodity exposure data into a click rate estimation model to obtain a corresponding commodity click rate;
the click conversion probability determining module 14 is configured to determine click conversion probabilities of the commodities based on the commodity conversion rate and the commodity click rate, and order the commodities according to the click conversion probabilities.
From the above, in this embodiment, the buying intention sequence, the historical click sequence and the commodity exposure data are input into the multi-objective prediction model including the commodity click rate learning task and the commodity conversion rate learning task, and because the embedded layer parameters and part of the attention parameters are shared between the commodity click rate learning task and the commodity conversion rate learning task, all samples, namely the commodity exposure data, are utilized to predict, and through the joint learning between the commodity click rate learning task and the commodity conversion rate learning task, the more accurate commodity conversion rate can be obtained, and finally the click conversion probability of the commodity is determined according to the commodity conversion rate obtained based on the multi-objective prediction model and the commodity click rate obtained based on the click rate prediction model, and the commodity is ordered according to the click conversion probability. Therefore, the optimal ordering sequence of the commodities can be determined, the commodity conversion rate is improved while the commodity click rate is ensured, and the total amount of the commodities is further improved.
In some specific embodiments, the data acquisition module 11 may specifically include:
the purchasing intention sequence construction unit is used for constructing an operation sequence based on the purchasing operation and the collecting operation of the user and through a historical footprint table and/or a favorites and/or a clicking operation of the shopping cart again; sorting the operation sequences according to the sequence from near to far between each operation and the current time to obtain the purchase intention sequence;
and the historical click sequence construction unit is used for constructing the historical click sequence based on the click operation of the user.
In some embodiments, the conversion determination module 12 may specifically include:
the feature acquisition unit is used for acquiring commodity features of commodities; the commodity characteristics comprise any one or more of commodity categories, commodity prices, commodity sales and commodity praise;
the historical click sequence construction unit is used for corresponding the commodity characteristics to commodities in the historical click sequence to obtain a historical click sequence containing commodity characteristics;
and the data input unit is used for inputting the historical click sequence containing commodity characteristics, the purchase intention sequence and the commodity exposure data into the multi-target estimated model.
In some specific embodiments, the click rate determination module 13 may specifically further include:
the constrained commodity conversion rate determining unit is used for performing constraint operation on the commodity conversion rate by utilizing a preset value range control function to obtain the constrained commodity conversion rate;
and the click conversion probability determining unit is used for multiplying the constrained commodity conversion rate by the commodity click rate so as to obtain the click conversion probability.
Further, the embodiment of the application also discloses an electronic device, and referring to fig. 5, the content in the drawing should not be considered as any limitation on the application scope of the application.
Fig. 5 is a schematic structural diagram of an electronic device 20 according to an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input output interface 25, and a communication bus 26. Wherein the memory 22 is configured to store a computer program that is loaded and executed by the processor 21 to implement the relevant steps in the network commodity ordering method disclosed in any of the foregoing embodiments.
In this embodiment, the power supply 23 is configured to provide an operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and an external device, and the communication protocol to be followed is any communication protocol applicable to the technical solution of the present application, which is not specifically limited herein; the input/output interface 25 is used for acquiring external input data or outputting external output data, and the specific interface type thereof may be selected according to the specific application requirement, which is not limited herein.
The memory 22 may be a carrier for storing resources, such as a read-only memory, a random access memory, a magnetic disk, or an optical disk, and the resources stored therein include an operating system 221, a computer program 222, and data 223 including user behavior data and commodity exposure data, and the storage may be temporary storage or permanent storage.
The operating system 221 is used for managing and controlling various hardware devices on the electronic device 20 and the computer program 222, so as to implement the operation and processing of the processor 21 on the mass data 223 in the memory 22, which may be Windows Server, netware, unix, linux, etc. The computer program 222 may further comprise a computer program capable of performing other specific tasks in addition to the computer program capable of performing the network commodity ordering method performed by the electronic device 20 as disclosed in any of the preceding embodiments.
Further, the embodiment of the application also discloses a computer storage medium, wherein the computer storage medium stores computer executable instructions, and when the computer executable instructions are loaded and executed by a processor, the steps of the network commodity ordering method disclosed in any embodiment are realized.
In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, so that the same or similar parts between the embodiments are referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software modules may be disposed in Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The above detailed description of the method, the device, the equipment and the medium for ordering network commodities provided by the application applies specific examples to illustrate the principle and the implementation of the application, and the above examples are only used for helping to understand the method and the core idea of the application; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present application, the present description should not be construed as limiting the present application in view of the above.
Claims (6)
1. A method for ordering network goods, comprising:
acquiring user behavior data and commodity exposure data; the user behavior data comprises a purchase intention sequence and a historical click sequence;
inputting the user behavior data and the commodity exposure data into a multi-target estimated model constructed based on a multi-task learning framework to obtain corresponding commodity conversion rate; the multi-target pre-estimation model comprises a commodity click rate learning task and a commodity conversion rate learning task; the commodity click rate learning task and the commodity conversion rate learning task share an embedded layer parameter and a part of attention parameter;
inputting the historical click sequence and the commodity exposure data into a click rate estimation model to obtain a corresponding commodity click rate;
determining click conversion probability of the commodity based on the commodity conversion rate and the commodity click rate, and sorting the commodity according to the click conversion probability;
the inputting the user behavior data and the commodity exposure data into a multi-target pre-estimation model constructed based on a multi-task learning framework comprises the following steps:
inputting the commodity exposure data into the commodity click rate learning task and the commodity conversion rate learning task;
configuring non-sharing parameters for the purchase intention sequence, and inputting the purchase intention sequence into the commodity click rate learning task and the commodity conversion rate learning task;
configuring sharing parameters for the historical click sequence, and inputting the historical click sequence into the commodity click rate learning task and the commodity conversion rate learning task so that the commodity click rate learning task and the commodity conversion rate learning task share attention parameters of the historical click sequence;
the construction process of the multi-target pre-estimation model comprises the following steps:
based on an ESMM network structure, constructing a dual-task learning model sharing embedded layer parameters and hidden layer partial parameters so as to obtain the multi-target pre-estimated model;
the inputting the user behavior data and the commodity exposure data into a multi-target pre-estimation model constructed based on a multi-task learning framework comprises the following steps:
acquiring commodity characteristics of commodities; the commodity characteristics comprise any one or more of commodity categories, commodity prices, commodity sales and commodity praise;
corresponding the commodity characteristics to commodities in the historical click sequence to obtain a historical click sequence containing commodity characteristics;
and inputting the historical click sequence containing commodity characteristics, the purchase intention sequence and the commodity exposure data into the multi-target predictive model.
2. The method of claim 1, wherein the obtaining user behavior data comprises:
constructing an operation sequence based on the purchasing operation and the collecting operation of the user and through a historical footprint table and/or a secondary clicking operation of the favorites and/or the shopping cart;
sorting the operation sequences according to the sequence from near to far between each operation and the current time to obtain the purchase intention sequence;
the historical click sequence is constructed based on a user's click operation.
3. The network commodity ordering method according to claim 1 or 2, wherein said determining a click-through conversion probability of a commodity based on said commodity conversion rate and said commodity click-through rate comprises:
performing constraint operation on the commodity conversion rate by using a preset value range control function to obtain the constrained commodity conversion rate;
multiplying the constrained commodity conversion rate by the commodity click rate to obtain the click conversion probability.
4. A network commodity ordering apparatus, comprising:
the data acquisition module is used for acquiring user behavior data and commodity exposure data; the user behavior data comprises a purchase intention sequence and a historical click sequence;
the conversion rate determining module is used for inputting the user behavior data and the commodity exposure data into a multi-target estimated model constructed based on a multi-task learning frame to obtain corresponding commodity conversion rate; the multi-target pre-estimation model comprises a commodity click rate learning task and a commodity conversion rate learning task; the commodity click rate learning task and the commodity conversion rate learning task share an embedded layer parameter and a part of attention parameter;
the click rate determining module is used for inputting the historical click sequence and the commodity exposure data into a click rate estimation model to obtain a corresponding commodity click rate;
the click conversion probability determining module is used for determining the click conversion probability of the commodity based on the commodity conversion rate and the commodity click rate and sequencing the commodity according to the click conversion probability;
the inputting the user behavior data and the commodity exposure data into a multi-target pre-estimation model constructed based on a multi-task learning framework comprises the following steps: inputting the commodity exposure data into the commodity click rate learning task and the commodity conversion rate learning task; configuring non-sharing parameters for the purchase intention sequence, and inputting the purchase intention sequence into the commodity click rate learning task and the commodity conversion rate learning task; configuring sharing parameters for the historical click sequence, and inputting the historical click sequence into the commodity click rate learning task and the commodity conversion rate learning task so that the commodity click rate learning task and the commodity conversion rate learning task share attention parameters of the historical click sequence;
the multi-target pre-estimation model is based on an ESMM network structure, and is obtained by constructing a dual-task learning model sharing embedded layer parameters and hidden layer partial parameters;
wherein, conversion rate determination module still includes:
the feature acquisition unit is used for acquiring commodity features of commodities; the commodity characteristics comprise any one or more of commodity categories, commodity prices, commodity sales and commodity praise;
the historical click sequence construction unit is used for corresponding the commodity characteristics to commodities in the historical click sequence to obtain a historical click sequence containing commodity characteristics;
and the data input unit is used for inputting the historical click sequence containing commodity characteristics, the purchase intention sequence and the commodity exposure data into the multi-target estimated model.
5. An electronic device, comprising:
a memory for storing a computer program;
a processor for executing the computer program to implement the network commodity ordering method according to claim 1 or 2.
6. A computer-readable storage medium storing a computer program; wherein the computer program when executed by a processor implements the network commodity ordering method according to claim 1 or 2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011264017.XA CN112381607B (en) | 2020-11-12 | 2020-11-12 | Network commodity ordering method, device, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011264017.XA CN112381607B (en) | 2020-11-12 | 2020-11-12 | Network commodity ordering method, device, equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112381607A CN112381607A (en) | 2021-02-19 |
| CN112381607B true CN112381607B (en) | 2023-11-24 |
Family
ID=74583480
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011264017.XA Active CN112381607B (en) | 2020-11-12 | 2020-11-12 | Network commodity ordering method, device, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112381607B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112200627B (en) * | 2020-10-09 | 2022-03-25 | 四川长虹电器股份有限公司 | Method for improving shopping cart order conversion rate based on improved LFU strategy |
| CN113159905A (en) * | 2021-05-20 | 2021-07-23 | 深圳马六甲网络科技有限公司 | Commodity recommendation method, commodity recommendation device, commodity recommendation equipment and storage medium for new user |
| CN113312555B (en) * | 2021-06-15 | 2023-10-03 | 北京百度网讯科技有限公司 | Conversion rate prediction method, conversion rate prediction device, electronic equipment and storage medium |
| CN113506154A (en) * | 2021-07-16 | 2021-10-15 | 杭州时趣信息技术有限公司 | Commodity recommendation method and system, electronic equipment and related components |
| CN115860870A (en) * | 2022-12-16 | 2023-03-28 | 深圳市云积分科技有限公司 | Commodity recommendation method, system and device and readable medium |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779985A (en) * | 2017-02-24 | 2017-05-31 | 武汉奇米网络科技有限公司 | A kind of method and system of personalized commercial sequence |
| CN110008399A (en) * | 2019-01-30 | 2019-07-12 | 阿里巴巴集团控股有限公司 | A kind of training method and device, a kind of recommended method and device of recommended models |
| CN110033314A (en) * | 2019-03-18 | 2019-07-19 | 北京品友互动信息技术股份公司 | Advertisement data processing method and device |
| WO2019169977A1 (en) * | 2018-03-07 | 2019-09-12 | 阿里巴巴集团控股有限公司 | Information conversion rate prediction method and apparatus, and information recommendation method and apparatus |
| CN110336944A (en) * | 2019-07-05 | 2019-10-15 | 杭州时趣信息技术有限公司 | Automatically snap and analyze the method, apparatus, equipment and system of image of clothing |
| CN110415032A (en) * | 2019-07-24 | 2019-11-05 | 深圳乐信软件技术有限公司 | A kind of exposure conversion ratio predictor method, device, equipment and storage medium |
| CN110569427A (en) * | 2019-08-07 | 2019-12-13 | 智者四海（北京）技术有限公司 | Multi-target sequencing model training and user behavior prediction method and device |
| CN110796513A (en) * | 2019-09-25 | 2020-02-14 | 北京三快在线科技有限公司 | Multitask learning method and device, electronic equipment and storage medium |
| CN110838043A (en) * | 2019-11-05 | 2020-02-25 | 智者四海（北京）技术有限公司 | Commodity recommendation method and device |
| CN110880124A (en) * | 2019-09-29 | 2020-03-13 | 清华大学 | Conversion rate evaluation method and device |
| CN111027895A (en) * | 2019-05-16 | 2020-04-17 | 珠海随变科技有限公司 | Stock prediction and behavior data collection method, apparatus, device and medium for commodity |
| CN111079015A (en) * | 2019-12-17 | 2020-04-28 | 腾讯科技（深圳）有限公司 | Recommendation method and device, computer equipment and storage medium |
| CN111159241A (en) * | 2019-12-20 | 2020-05-15 | 深圳前海微众银行股份有限公司 | Click conversion estimation method and device |
| WO2020107762A1 (en) * | 2018-11-27 | 2020-06-04 | 深圳前海微众银行股份有限公司 | Ctr estimation method and device, and computer readable storage medium |
-
2020
- 2020-11-12 CN CN202011264017.XA patent/CN112381607B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779985A (en) * | 2017-02-24 | 2017-05-31 | 武汉奇米网络科技有限公司 | A kind of method and system of personalized commercial sequence |
| WO2019169977A1 (en) * | 2018-03-07 | 2019-09-12 | 阿里巴巴集团控股有限公司 | Information conversion rate prediction method and apparatus, and information recommendation method and apparatus |
| WO2020107762A1 (en) * | 2018-11-27 | 2020-06-04 | 深圳前海微众银行股份有限公司 | Ctr estimation method and device, and computer readable storage medium |
| CN110008399A (en) * | 2019-01-30 | 2019-07-12 | 阿里巴巴集团控股有限公司 | A kind of training method and device, a kind of recommended method and device of recommended models |
| CN110033314A (en) * | 2019-03-18 | 2019-07-19 | 北京品友互动信息技术股份公司 | Advertisement data processing method and device |
| CN111027895A (en) * | 2019-05-16 | 2020-04-17 | 珠海随变科技有限公司 | Stock prediction and behavior data collection method, apparatus, device and medium for commodity |
| CN110336944A (en) * | 2019-07-05 | 2019-10-15 | 杭州时趣信息技术有限公司 | Automatically snap and analyze the method, apparatus, equipment and system of image of clothing |
| CN110415032A (en) * | 2019-07-24 | 2019-11-05 | 深圳乐信软件技术有限公司 | A kind of exposure conversion ratio predictor method, device, equipment and storage medium |
| CN110569427A (en) * | 2019-08-07 | 2019-12-13 | 智者四海（北京）技术有限公司 | Multi-target sequencing model training and user behavior prediction method and device |
| CN110796513A (en) * | 2019-09-25 | 2020-02-14 | 北京三快在线科技有限公司 | Multitask learning method and device, electronic equipment and storage medium |
| CN110880124A (en) * | 2019-09-29 | 2020-03-13 | 清华大学 | Conversion rate evaluation method and device |
| CN110838043A (en) * | 2019-11-05 | 2020-02-25 | 智者四海（北京）技术有限公司 | Commodity recommendation method and device |
| CN111079015A (en) * | 2019-12-17 | 2020-04-28 | 腾讯科技（深圳）有限公司 | Recommendation method and device, computer equipment and storage medium |
| CN111159241A (en) * | 2019-12-20 | 2020-05-15 | 深圳前海微众银行股份有限公司 | Click conversion estimation method and device |
Non-Patent Citations (5)
| Title |
|---|
| Entire Space Multi-Task Model: An Effective Approach for Estimating Post-Click Conversion Rate;Xiao Ma，等;Short Research Papers II;1137-1140 * |
| 刘宝锤.《大数据分类模型和算法研究》.云南大学出版社,2019,第154页. * |
| 基于增强型FNN的广告点击率预测模型;杨妍婷;韩斌;;南京理工大学学报(第01期);37-43 * |
| 基于特征工程的广告点击转化率预测模型;邓秀勤;谢伟欢;刘富春;张翼飞;樊娟;;数据采集与处理(第05期);56-63 * |
| 基于知识图谱的推荐系统研究综述;秦川;祝恒书;庄福振;郭庆宇;张琦;张乐;王超;陈恩红;熊辉;;中国科学:信息科学(第07期);937-956 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112381607A (en) | 2021-02-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112381607B (en) | Network commodity ordering method, device, equipment and medium | |
| US11281969B1 (en) | Artificial intelligence system combining state space models and neural networks for time series forecasting | |
| CN111080413A (en) | E-commerce platform commodity recommendation method and device, server and storage medium | |
| CN109829116B (en) | Content recommendation method and device, server and computer readable storage medium | |
| CN110083334A (en) | The online method and device of model | |
| CN112365283B (en) | Coupon issuing method and device, terminal equipment and storage medium | |
| JP6994872B2 (en) | Correction device, correction method and correction program | |
| CN114117216A (en) | Recommendation probability prediction method and device, computer storage medium and electronic equipment | |
| CN114463091A (en) | Information push model training and information push method, device, equipment and medium | |
| CN110766513A (en) | Information sorting method and device, electronic equipment and readable storage medium | |
| Lakshmanan et al. | Sales demand forecasting using LSTM network | |
| CN111680213B (en) | Information recommendation method, data processing method and device | |
| CN112348460A (en) | Resource limit adjusting method and device of cooperative task and electronic equipment | |
| WO2015030929A1 (en) | Keyword bid optimization for text ads | |
| WO2024051707A1 (en) | Recommendation model training method and apparatus, and resource recommendation method and apparatus | |
| CA2918915A1 (en) | Automatic computation of keyword bids for pay-per-click advertising campaigns and methods and systems incorporating the same | |
| CN112256998B (en) | Popularization information management method and device and electronic equipment | |
| JP6866069B2 (en) | Information processing equipment, information processing methods and information processing programs | |
| CN108537570B (en) | Information exposure method and device | |
| CN113360816A (en) | Click rate prediction method and device | |
| CN111178987B (en) | Method and device for training user behavior prediction model | |
| KR102372432B1 (en) | Method, device and system for providing recommended content using click and exposure information | |
| CN113761352A (en) | Information pushing method, device, equipment and storage medium | |
| US20100010891A1 (en) | Methods for advertisement slate selection | |
| CN113094602A (en) | Hotel recommendation method, system, equipment and medium |
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 |