CN107632989B

CN107632989B - Method and device for selecting commodity objects, determining models and determining use heat

Info

Publication number: CN107632989B
Application number: CN201610569600.9A
Authority: CN
Inventors: 叶舟; 王瑜; 陈凡; 杨洋; 董昭萍; 钱倩; 王吉能
Original assignee: Alibaba Group Holding Ltd
Current assignee: Alibaba (Jiangxi) Co.,Ltd.
Priority date: 2016-07-19
Filing date: 2016-07-19
Publication date: 2021-04-13
Anticipated expiration: 2036-07-19
Also published as: WO2018014764A1; CN107632989A; TW201816685A

Abstract

The application discloses a commodity object selecting method, a model determining method and a use heat determining device, at least one commodity object can be automatically selected from a large number of objects to serve as an object meeting the requirements of a user based on commodity object subject terms input by the user and a set object recognition model, so that the commodity object selecting efficiency is greatly improved, the cost of manual goods inventory is reduced, and the operation efficiency is improved.

Description

Method and device for selecting commodity objects, determining models and determining use heat

Technical Field

The application relates to the technical field of internet, in particular to a method and a device for selecting commodity objects, determining models and determining use heat.

Background

In order to improve the transaction performance of the commodity objects in the electronic commerce system, the electronic commerce system often establishes various new channels to increase the exposure of the commodity objects, for example, establishes various special price second activity killing theme channels or theme channels with main commodity printing object adjustability.

These channels, when just established, encounter the problem of inventory, i.e., how to define the merchandise objects that fit the channel to maximize channel engagement. Specifically, at present, in order to solve the above problem, a corresponding commodity object is often selected for each newly-built channel in a manual manner, that is, an operator subjectively selects a commodity object meeting a theme required for each newly-built channel according to manual experience.

However, the selection of the commodity object by the manual method often takes a lot of time, so that the selection efficiency of the commodity object is very low.

Disclosure of Invention

The embodiment of the application provides a method and a device for selecting a commodity object, determining a model and determining the use heat, which are used for solving the problem of low efficiency of the existing commodity object selecting mode.

In one aspect, an embodiment of the present application provides a method for selecting a commodity object, including:

receiving a commodity object subject term sent by a user terminal;

acquiring a primary selection object matched with the commodity object subject term;

determining the object use heat of each primary selection object in a second time period based on the set object recognition model and the use characteristic data of each primary selection object in the first time period; the object identification model is an identification model which is obtained through training and used for representing the incidence relation between the use characteristic data of the commodity object in a first time period and the object use heat of the commodity object in a second time period; the first time period is a specified time period before the second time period;

and selecting at least one commodity object from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period according to the object use heat of each primary selection object in the second time period.

Optionally, selecting at least one commodity object from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period, including:

and selecting at least one commodity object with the use heat not lower than a set heat threshold from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period.

Optionally, before determining the object usage heat of each primary selected object in the second time period based on the set object recognition model and the usage characteristic data of each primary selected object in the first time period, the method further includes:

receiving object identification model training sample data sent by a user terminal, wherein the object identification model training sample data contains basic feature data of each object identification model sample object, and the basic feature data of each object identification model sample object comprises use feature data of the object identification model sample object in each third time period and object use heat of the object identification model sample object in each corresponding fourth time period; the third time period is a specified time period before the corresponding fourth time period;

and training an initial object recognition model which is established in advance and used for predicting the incidence relation between the use characteristic data of the commodity object in the first time period and the object use heat of the commodity object in the second time period according to the basic characteristic data of each object recognition model sample object to obtain the object recognition model.

Optionally, before selecting at least one commodity object from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period, the method further includes:

determining the category use heat of the category corresponding to each primary selection object in one or more historical synchronization time periods corresponding to the second time period based on the set category identification model and the category use heat of the category corresponding to each primary selection object in the second time period; the category identification model is an identification model which is obtained through training and used for representing the association relationship between the category use heat of the commodity object category in a second time period and the category use heat of the commodity object category in one or more historical synchronization time periods corresponding to the second time period;

and updating the object use heat of the primary selection object in the second time period according to the class use heat of the class corresponding to the primary selection object in the second time period aiming at each primary selection object.

Optionally, before determining the category usage heat of the category corresponding to each initially selected object in the second time period based on the set category identification model and the category usage heat of the category corresponding to each initially selected object in one or more historical synchronization time periods corresponding to the second time period, the method further includes:

receiving category identification model training sample data sent by a user terminal, wherein the category identification model training sample data comprises basic characteristic data of each category identification model sample object, and the basic characteristic data of each category identification model sample object comprises category use heat of a category corresponding to the category identification model sample object in each fifth time period and category use heat of the category corresponding to the category identification model sample object in one or more historical contemporaneous time periods corresponding to each fifth time period;

and training an initial category identification model which is established in advance and used for predicting the association relation between the category use heat of the commodity object category in a second time period and the category use heat of the commodity object category in one or more historical synchronization time periods corresponding to the second time period according to the basic characteristic data of the category identification model sample object to obtain the category identification model.

The historical contemporaneous time period of each time period is the historical time period which is under the same Gregorian or lunar calendar day as the time period and corresponds to the time period.

Optionally, for any initially selected object, before updating the object usage heat of the initially selected object in the second time period according to the class usage heat of the class corresponding to the initially selected object in the second time period, the method further includes:

if the second time period is determined to be a specific time period, if the category corresponding to the initially selected object is determined to be a category matched with the specific time period corresponding to the second time period, increasing the category use heat of the category corresponding to the initially selected object in the second time period according to a set coefficient.

determining the similarity between the primary selection objects according to the similarity between the titles corresponding to the primary selection objects;

and for each group of object sets consisting of at least one initially selected object with the similarity between the objects not lower than a first similarity threshold, keeping one object in the object sets, and deleting other objects.

Optionally, before obtaining the primary selection object matched with the subject term of the commodity object, the method further comprises:

aiming at each commodity object subject term sent by the user terminal, determining at least one sample term of which the similarity with the commodity object subject term is not lower than a second similarity threshold value based on the set sample corpus;

and taking each determined sample word as a final required commodity object subject word.

Wherein the object recognition model is a regression model; the category identification model is a linear model.

On the other hand, an embodiment of the present application provides another method for selecting a commodity object, including:

receiving commodity object subject terms input by a user, and sending the commodity object subject terms to a server;

receiving commodity object information returned by the server according to the commodity object subject term, and taking a commodity object corresponding to the commodity object information as a commodity object matched with the commodity object subject term in a second time period;

the commodity object information is related information of a commodity object selected by the server from the primary selection objects matched with the commodity object subject term according to a set object recognition model and the use characteristic data of each primary selection object matched with the commodity object subject term in a first time period; the first time period is a specified time period before the second time period;

the set object identification model is used for representing the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat of the commodity object in the second time period.

Optionally, before receiving the item object subject term input by the user, the method further comprises:

receiving object recognition model training sample data input by a user, wherein the object recognition model training sample data contains basic feature data of each object recognition model sample object, and the basic feature data of each object recognition model sample object comprises use feature data of the object recognition model sample object in each third time period and object use heat of the object recognition model sample object in each corresponding fourth time period; the third time period is a specified time period before the corresponding fourth time period;

and transmitting the object recognition model training sample data to a server, and training an initial object recognition model which is established in advance and used for predicting the association relationship between the use characteristic data of the commodity object in a first time period and the object use heat of the commodity object in a second time period according to the basic characteristic data of each object recognition model sample object contained in the object recognition model training sample data by the server to obtain the set object recognition model.

In another aspect, an embodiment of the present application provides a model determining method, including:

training an initial object recognition model which is established in advance and used for predicting the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat of the commodity object in the second time period according to the basic characteristic data of each object recognition model sample object contained in the object recognition model training sample data to obtain an object recognition model used for representing the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat of the commodity object in the second time period; the first time period is a specified time period before the second time period.

Wherein, the use characteristic data at least comprises any one or more of browsing times, collection times, purchase adding times, transaction times, comment times and search times; the subject use heat includes at least any one or more of a volume of a transaction, and a rate of conversion of a transaction.

In another aspect, an embodiment of the present application provides another model determining method, including:

receiving category identification model training sample data sent by a user terminal, wherein the category identification model training sample data comprises basic characteristic data of various category identification model sample objects, and the basic characteristic data of each category identification model sample object comprises category use heat of a category corresponding to the category identification model sample object in various first time periods and category use heat of the category corresponding to the category identification model sample object in one or more historical same-period time periods corresponding to the various first time periods;

according to basic characteristic data of various category identification model sample objects contained in the category identification model training sample data, training an initial category identification model which is established in advance and used for predicting the association relationship between the category use heat of the commodity object category in a second time period and the category use heat of the commodity object category in one or more historical synchronization time periods corresponding to the second time period, and obtaining a category identification model used for representing the association relationship between the category use heat of the commodity object category in the second time period and the category use heat of the commodity object category in one or more historical synchronization time periods corresponding to the second time period.

Wherein the category usage heat at least comprises any one or more of the volume of the transaction, the volume of the transaction and the conversion rate of the transaction.

In another aspect, an embodiment of the present application further provides a method for determining a usage heat, including:

acquiring use characteristic data of each commodity object in a first time period;

determining the object use heat degree of each commodity object in the second time period based on the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat degree of the commodity object in the second time period and the use characteristic data of each commodity object in the first time period;

the first time period is a specified time period before the second time period; the incidence relation is established according to the use characteristic data of each sample object in each third time period and the object use heat of each sample object in each corresponding fourth time period; the third time period is a specified time period before the corresponding fourth time period.

In another aspect, an embodiment of the present application further provides a commodity object selection device, including:

the subject term receiving unit is used for receiving the subject terms of the commodity objects sent by the user terminal;

the object acquisition unit is used for acquiring a primary selection object matched with the commodity object subject term;

the heat determining unit is used for determining the object use heat of each primary selection object in a second time period based on the set object recognition model and the use characteristic data of each primary selection object in the first time period; the object identification model is an identification model which is obtained through training and used for representing the incidence relation between the use characteristic data of the commodity object in a first time period and the object use heat of the commodity object in a second time period; the first time period is a specified time period before the second time period;

and the object screening unit is used for selecting at least one commodity object from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period according to the object use heat of each primary selection object in the second time period.

On the other hand, the embodiment of the present application further provides another commodity object selection device, including:

the subject term receiving unit is used for receiving the subject terms of the commodity objects input by the user;

the subject term sending unit is used for sending the subject terms of the commodity objects to a server;

the object information receiving unit is used for receiving the commodity object information returned by the server according to the commodity object subject term;

the object determining unit is used for determining the commodity object corresponding to the commodity object information as a commodity object matched with the commodity object subject term in a second time period;

In another aspect, an embodiment of the present application further provides a model determining apparatus, including:

the data receiving unit is used for receiving object identification model training sample data sent by a user terminal, wherein the object identification model training sample data contains basic characteristic data of each object identification model sample object, and the basic characteristic data of each object identification model sample object comprises use characteristic data of the object identification model sample object in each third time period and object use heat of the object identification model sample object in each corresponding fourth time period; the third time period is a specified time period before the corresponding fourth time period;

and the model training unit is used for training an initial object recognition model which is established in advance and used for predicting the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat of the commodity object in the second time period according to the basic characteristic data of each object recognition model sample object contained in the object recognition model training sample data to obtain an object recognition model which is used for representing the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat of the commodity object in the second time period.

On the other hand, an embodiment of the present application further provides another model determining apparatus, including:

the data receiving unit is used for receiving category identification model training sample data sent by a user terminal, wherein the category identification model training sample data comprises basic characteristic data of various category identification model sample objects, and the basic characteristic data of each category identification model sample object comprises category use heat of a category corresponding to the category identification model sample object in various first time periods and category use heat of the category corresponding to the category identification model sample object in one or more historical synchronization time periods corresponding to the various first time periods;

and the model training unit is used for training an initial category identification model which is established in advance and used for predicting the association relationship between the category use heat of the commodity object category in the second time period and the category use heat of the commodity object category in one or more historical contemporaneous time periods corresponding to the second time period according to basic characteristic data of various category identification model sample objects contained in the category identification model training sample data to obtain a category identification model used for representing the association relationship between the category use heat of the commodity object category in the second time period and the category use heat of the commodity object category in one or more historical contemporaneous time periods corresponding to the second time period.

In another aspect, an embodiment of the present application further provides a usage heat determination apparatus, including:

the data acquisition unit is used for acquiring the use characteristic data of each commodity object in a first time period;

the heat determining unit is used for determining the object use heat of each commodity object in the second time period based on the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat of the commodity object in the second time period and the use characteristic data of each commodity object in the first time period;

The beneficial effect of this application is as follows:

the embodiment of the application provides a commodity object selection method, a model determination method and a use heat determination device, and at least one commodity object can be automatically selected from massive objects as a final object meeting user requirements based on commodity object subject terms input by a user and a set object recognition model, so that the commodity object selection efficiency is greatly improved, the cost of manual goods inventory is reduced, and the operation efficiency is improved.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a schematic view of a possible application scenario of a commodity object selection method according to a first embodiment of the present application;

fig. 2 is a schematic flow chart illustrating a possible method for selecting a commodity object according to a first embodiment of the present disclosure;

FIG. 3 is a schematic flow chart illustrating a possible method for determining a model according to a first embodiment of the present disclosure;

FIG. 4 is a schematic flow chart illustrating a possible method for determining a model according to an embodiment of the present disclosure;

FIG. 5 is a schematic flow chart illustrating a possible method for determining heat according to one embodiment of the present disclosure;

fig. 6 is a schematic diagram illustrating a possible structure of a merchandise object selection device according to a second embodiment of the present application;

fig. 7 is a schematic diagram illustrating a possible structure of another merchandise object selection device according to a second embodiment of the present application;

fig. 8 is a schematic diagram illustrating a possible structure of a model determining apparatus according to a second embodiment of the present application;

fig. 9 is a schematic diagram illustrating a possible structure of another model determining apparatus according to the second embodiment of the present application;

fig. 10 is a schematic diagram illustrating a possible structure of a device using heat determination according to a second embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be described in further detail with reference to the accompanying drawings, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

The first embodiment is as follows:

in order to solve the problem of low efficiency of the existing commodity object selection method, an embodiment of the present application provides a commodity object selection method, as shown in fig. 1, which is a schematic view of a possible application scenario of the commodity object selection method, where the scenario may include: a user terminal 11 and a server 12, wherein:

the user terminal 11 can receive the commodity object subject term input by the user 10 and send the commodity object subject term to the server 12; the server 12 may obtain the primary selection objects matched with the commodity object subject term sent by the user terminal 11, determine the object usage heat of each primary selection object in the second time period based on the set object identification model and the usage characteristic data of each primary selection object in the first time period, and select at least one commodity object from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period according to the object usage heat of each primary selection object in the second time period; the user terminal 11 may receive the commodity object information of at least one commodity object returned by the server 12 after selecting the at least one commodity object, and use the commodity object corresponding to the commodity object information as the commodity object matched with the commodity object subject term in the second time period; the set object identification model can be used for representing the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat of the commodity object in the second time period; the first time period is a specified time period before the second time period.

The user terminal 11 and the server 12 may be communicatively connected through a communication network, which may be a local area network, a wide area network, or the like. The user terminal 11 may be a terminal device such as a mobile phone, a tablet computer, a notebook computer, a personal computer, or even a client installed in the terminal device; the server 12 may be any server device capable of supporting processing operations such as screening of commodity objects.

That is to say, in the embodiment of the application, at least one commodity object can be automatically selected from the massive objects as a final object meeting the user requirements based on the commodity object subject term input by the user and the set object identification model, so that the commodity object selection efficiency is greatly improved, the cost of manual goods inventory is reduced, and the operation efficiency is improved.

In the following, with reference to fig. 2, a method for selecting a commodity object in a first embodiment of the present application will be exemplarily described with reference to an application scenario shown in fig. 1. It should be noted that the above application scenarios are only presented to facilitate understanding of the spirit and principles of the present application, and the embodiments of the present application are not limited in this respect. Rather, embodiments of the present application may be applied to any scenario where applicable.

Specifically, as shown in fig. 2, which is a schematic flow chart of a commodity object selection method in an embodiment of the present application, the commodity object selection method may include the following steps:

step 201: the user terminal receives object recognition model training sample data input by a user and sends the object recognition model training sample data to a server.

The object recognition model training sample data comprises basic characteristic data of each object recognition model sample object, and the basic characteristic data of each object recognition model sample object comprises use characteristic data of the object recognition model sample object in each third time period and object use heat of the object recognition model sample object in each corresponding fourth time period; the third time period is a specified time period before the corresponding fourth time period.

Optionally, the usage characteristic data of the commodity objects such as each object recognition model sample object at least may include any one or more of browsing times, collection times, purchase adding (shopping cart adding) times, deal times, comment times and search times; the object use heat of the commodity object such as each object recognition model sample object at least comprises any one or more of volume of bargain, volume of bargain and conversion rate of bargain. It should be noted that the usage characteristic data of the commodity object such as each object identification model sample object may be acquired from the commodity object operation information of each electronic commerce site, and the object usage heat of the commodity object such as each object identification model sample object may be calculated based on the usage characteristic data of the commodity object such as each object identification model sample object, which is not described herein again.

Furthermore, it should be noted that the third and fourth time periods may be historical time periods in general, that is, the basic feature data of each object recognition model sample object may be corresponding historical data in general; and, the size of the third, fourth time period, and the designated time period can be flexibly set according to the actual situation, such as 1 day, 1 week, 1 month, etc. (generally, the minimum is one day). In addition, the lengths of the third and fourth time periods may be the same or different, for example, the fourth time period may be one day, and the third time period corresponding to the fourth time period (i.e. the previous specified time period of the fourth time period) may be one month or one year, etc., or the fourth time period may be one month, and the third time period corresponding to the fourth time period may be one day, etc.; the third time period corresponding to the fourth time period may be a previously specified time period adjacent to the fourth time period, and of course, may not be adjacent to the fourth time period, and is not limited thereto.

Step 202: the server receives object recognition model training sample data sent by the user terminal, and trains a pre-established initial object recognition model according to basic feature data of each object recognition model sample object in the object recognition model training sample data to obtain a required object recognition model.

The initial object identification model is an identification model used for predicting the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat of the commodity object in the second time period; the object identification model is an identification model which is obtained through training and used for representing the association relationship between the use characteristic data of the commodity object in a first time period and the object use heat degree of the commodity object in a second time period, and the first time period is a specified time period before the second time period.

In addition, similar to the description of the third and fourth time periods, the sizes of the first and second time periods may be flexibly set according to actual situations, and the lengths of the first and second time periods may be the same or different (although, the size of the first time period may be generally the same as that of the third time period, and the size of the second time period may be generally the same as that of the fourth time period), which is not limited.

Optionally, taking the object usage heat of the commodity objects such as each object recognition model sample object as the volume of transaction, the corresponding object recognition model may be obtained through the following training steps:

a1: and establishing an initial object identification model related to the volume of the transaction.

Alternatively, it is assumed that the third and fourth time periods can be set to 1 day; and it is assumed that y (t) can represent the volume of a certain commodity object on the date t, x1(t-1) represents the browsing times of the commodity object on the date t-1, x2(t-1) represents the collection times of the commodity object on the date t-1, x3(t-1) represents the purchase times of the commodity object on the date t-1, and the like, and the initial object recognition model is a linear model; the initial object recognition model established may be expressed as y (t) ═ a (1) × 1(t-1) + a (2) × 2(t-1) + a (3) × 3(t-1) + …, where a (1), a (2), a (3), …, etc. are the coefficients that need to be estimated.

A2: calculating the volume of each object identification model sample object in each fourth time period (such as daily sales volume on the date t), and associating the volume with the use characteristic data (such as characteristic data of browsing, collecting, buying, bargaining, commenting and searching on the date t-1) of each object identification model sample object in the corresponding third time period to obtain a plurality of associated data; and training the established initial object recognition model based on the obtained associated data to obtain actual values of coefficients a (1), a (2), a (3) and the like so as to obtain the required object recognition model.

It should be noted that, compared to the linear model, the Regression model can better cross features, improve the prediction capability, prevent the merchant from cheating, and further improve the accuracy of prediction, and therefore, in this embodiment, preferably, the initial object identification model and the object identification model may be Regression models, such as a Gradient Boost Regression Tree model. Of course, if the accuracy requirement for the prediction is relatively low, the initial object recognition model and the object recognition model may also be linear models, which is not limited herein.

In addition, after the object recognition model is obtained through training, the object recognition model can be updated in real time or at regular time according to the latest sample data of the object recognition model, so that the accuracy of the object recognition model is improved. Furthermore, for example, when a commodity object is selected for a certain channel, after the channel is online, the usage characteristic data of each object identification model sample object in each third time period and the object usage heat of each object identification model sample object in each corresponding fourth time period may be respectively replaced with the corresponding usage characteristic data and the object usage heat of each object identification model sample object in the channel, so as to better predict the object usage heat of the commodity object in the channel, which is not described herein again.

In addition, in the present embodiment, the

steps

201 and 202 are steps of creating an object recognition model in advance, and are not steps that need to be executed each time a commodity object is selected, unless the object recognition model training sample data is updated accordingly. That is, after the

steps

201 and 202 are performed, the subsequent steps may be repeatedly performed for a plurality of times, which is not described herein again.

Step 203: the user terminal receives the commodity object subject term input by the user and sends the commodity object subject term to the server.

Optionally, after receiving the commodity object subject term input by the user, the user terminal may directly send the received commodity object subject term to the server, may also expand the received commodity object subject term, and send the expanded commodity object subject term to the server, so as to improve the richness of the commodity object subject term. In addition, the user terminal expands the commodity object subject terms input by the user, and the condition that the server needs to expand a large number of received commodity object subject terms simultaneously when a large number of user terminals simultaneously send the commodity object subject terms to the server can be avoided, so that the processing resources of the server are saved, the working pressure of the server is reduced, and the speed and the efficiency of selecting subsequent commodity objects can be further improved.

Optionally, the user terminal may expand the received subject term of the commodity object by:

aiming at each received commodity object subject term, determining at least one sample term of which the similarity with the commodity object subject term is not lower than a set similarity threshold (the threshold can be flexibly set according to actual conditions) based on set sample linguistic data; and taking each determined sample word as a final required commodity object subject word.

The set sample corpus can be a corpus such as e-commerce news crawled from an external website through a crawler; in addition, when at least one sample word with the similarity between the sample word and each commodity object subject word not lower than a set similarity threshold is determined based on a set sample corpus, a word can be characterized as a language model of a real-valued vector such as a word2vec model and the like based on the set sample corpus, and each commodity object subject word input by a user and each word in the sample corpus can be converted into a vector based on the trained language models such as the word2vec model and the like; then, the similarity between each word in the sample corpus and each commodity object subject word input by the user can be calculated by using a set similarity calculation formula, such as a Cosine formula and the like; finally, the words above (and including) the value are selected as the final required subject words by setting the corresponding similarity threshold.

For example, suppose that a user inputs the following three commodity object subject terms to a user terminal according to actual needs: the user terminal can expand the three words based on the set sample corpus, for example, the three words are expanded to obtain the words such as fashion, trend, shirt, suit, dress, jeans and the like, and the expanded words are used as final subject words of the commodity object.

Step 204: the server receives the commodity object subject term sent by the user terminal and obtains a primary selection object matched with the commodity object subject term.

Alternatively, the server may search for a commodity object whose corresponding commodity object title matches (e.g., partially matches) the commodity object subject term sent by the user terminal from the commodity object information of each e-commerce site based on a text mining method according to the commodity object subject term sent by the user terminal, and use each searched commodity object as an initial selection object matching the commodity object subject term sent by the user terminal.

Each commodity object information in the e-commerce website may include basic information such as an ID (identification), a name (i.e., title), a place of origin, seller user information, and categories of the commodity object, which is not described herein again.

In addition, optionally, before the server obtains the primary selection object matched with the commodity object subject term according to the received commodity object subject term sent by the user terminal, the server may further expand the received commodity object subject term so as to obtain the corresponding primary selection object based on the expanded commodity object subject term.

The specific implementation manner of the server expanding the received subject term of the commodity object is similar to the specific implementation manner of the user terminal expanding the received subject term of the commodity object in step 203, and is not described in detail here.

It should be noted that, the server, not the user terminal, executes the operation of expanding the subject term of the commodity object input by the user, which can reduce the performance requirement on the user terminal, so that the method described in the embodiment of the present application has a wider application range; in addition, for the user terminal, the processing resource of the user terminal can be saved, and the working pressure of the user terminal can be reduced.

Step 205: and the server determines the object use heat of each primary selection object in a second time period based on the trained object recognition model and the use characteristic data of each primary selection object in the first time period.

For example, if the object recognition model is an object recognition model trained by the server based on the volume of each object recognition model sample object at each date and the feature data of each object recognition model sample object such as browsing, collecting, buying, bargaining, commenting, searching at one or more dates before the corresponding date, the volume of each primary selected object at the date t +1 can be predicted based on the object recognition model based on the feature data of each primary selected object such as browsing, collecting, buying, bargaining, commenting, searching at one or more dates before the date t + 1.

It should be noted that, in this embodiment, particularly, when the target use heat of each commodity object is a transaction amount, since the range of the transaction amount is large, it may not be well predicted, and therefore, the transaction amount of each commodity object may not be directly predicted, but the transaction amount may be predicted based on the target recognition model related to the transaction amount, and then multiplied by the corresponding price to obtain the transaction amount, so as to improve the accuracy of prediction.

Step 206: and the server selects at least one commodity object from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period according to the object use heat of each primary selection object in the second time period.

Optionally, in order to improve the accuracy of the selected commodity object, the server may select, according to the object use heat of each primary selected object in the second time period, at least one commodity object whose object use heat is not lower than a set heat threshold (the threshold may be flexibly set according to actual conditions) from the primary selected objects as the commodity object matched with the commodity object subject term in the second time period.

In addition, the server may also sort the primary selection objects in an order from high object use heat to low object use heat, and take the top K (K is any positive integer) primary selection objects as the commodity objects matched with the commodity object subject terms in the second time period.

Further, in order to improve the accuracy of the selected commodity objects, before at least one commodity object is selected from the primary selected objects as a commodity object matched with the subject term of the commodity object in the second time period, each primary selected object can be manually screened according to actual needs, or commodity objects (for example, commodity objects with a low heat of use for short-term objects or prices not meeting the user needs, such as 5 deals within three days, and prices between 10 and 200 yuan, etc.) with a low price are deleted, so that the required commodity objects are selected based on each screened primary selected object; and/or the presence of a gas in the gas,

after at least one commodity object is selected from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period, the selected commodity object can be manually screened according to actual requirements, or the commodity object with low heat of use or the price not meeting the user requirement of the short-term object is deleted, and each screened commodity object is taken as the commodity object matched with the commodity object subject term in the second time period which is finally needed, which is not described in detail herein.

Further, since some of the goods objects have obvious seasonality, such as dress 6, coat 9, etc., and the buyer usually purchases the goods object in advance, the use of the goods object in the advance purchasing time period is not hot enough, so that the goods objects cannot be arranged at the front position by the previous prediction method. Therefore, in order to solve such a problem, after determining the object use heat of each primary selection object in the second time period, the object use heat of each primary selection object and the sequence of each primary selection object may be adjusted according to the time information, so that the final desired commodity object is selected according to each primary selection object after the heat adjustment. The time series model can be used for predicting the heat of the categories in the second time period, so that some commodity objects in due seasons can emerge in advance, and the accuracy of commodity object selection is further improved.

That is, before selecting at least one commodity object from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period, the method may further include:

Optionally, for each initially selected object, a product of the class usage heat of the class corresponding to the initially selected object in the second time period and the class usage heat of the initially selected object in the second time period, or a weighted sum of the two (weights corresponding to the two may be flexibly set according to actual conditions) may be used as the updated object usage heat of the initially selected object in the second time period.

The category usage heat may include at least any one or more of a volume of a transaction, an amount of a transaction, and a rate of conversion of a transaction, and the category usage heat of the commodity object such as each sample object may be calculated based on usage characteristic data of the commodity object such as each sample object, which is not limited thereto. In addition, the historical contemporaneous time period of each time period is the historical time period which is under the same Gregorian or lunar calendar day as the time period and corresponds to the time period; for example, for a period 2016 from 01/05/2016, the historical period may range from 01/2015 to 01/05/2015, from 01/2014 to 01/05/2014, and so on, which is not described herein.

Optionally, in this embodiment, before determining the category usage heat of the category corresponding to each primary selection object in one or more historical synchronization time periods corresponding to the second time period based on the set category identification model and the category usage heat of the category corresponding to each primary selection object in the one or more historical synchronization time periods corresponding to the second time period, the server may obtain the category identification model by:

It should be noted that, the fifth time period may be a historical time period in general; similarly to the above description about the first, second, third and fourth time periods, the size of the fifth time period can also be flexibly set according to the actual situation (however, the size of the fifth time period can be generally the same as the second time period), and this is not limited.

Optionally, taking the class use heat of each class identification model sample object as a volume, the corresponding class identification model may be obtained through the following training steps:

b1: and establishing an initial category identification model related to the volume of the transaction.

For example, it is assumed that each fifth period of time may be set to 1 month; the initial category identification model is a linear model; assuming that the volume of a certain class of target in the t month of the year is z (t), the volume of the target in the same period of the last year is z (t-1), the volume of the target in the same period of the previous year is z (t-2), and so on; the initial category identification model established may be denoted as z (t) ═ b (1) × z (t-1) + b (2) × z (t-2) + …, where b (1), b (2), …, etc. are the coefficients that need to be estimated.

The reason why the linear model is used as the category identification model is that the model parameters are small and the history data of the category is relatively stable. Of course, other models, such as a regression model, may be used as the category identification model to improve the accuracy of the category using the heat prediction, and is not limited herein.

B2: and training the established initial category identification model by using the latest period of category historical data (such as the latest 3-month category historical volume in the current year) and corresponding contemporaneous historical data (such as at least two years of contemporaneous historical data under a gregorian calendar day or a lunar calendar day) to obtain actual values of coefficients b (1), b (2) and the like so as to obtain the finally required category identification model.

In addition, after the category identification model is obtained through training, the category identification model can be updated in real time or at regular time according to the latest sample data of the category identification model, so that the accuracy of the category identification model is improved; furthermore, for example, when a commodity object is selected for a certain channel, after the channel is online, the category usage heat of each category identification model sample object in each fifth time period and the category usage heat of each category identification model sample object in one or more historical synchronization time periods corresponding to each fifth time period may be replaced with the corresponding category usage heat of each category identification model sample object in the channel, so as to better predict the category usage heat of the commodity object category in the channel, and further, the description is omitted here.

Further, after obtaining the category identification model in the above manner, if it is determined that the second time period is the next month, the category usage heat of the category corresponding to each primary selected object in the next month may be determined based on the category identification model obtained through training and the category usage heat of the category corresponding to each primary selected object in one or more historical synchronization time periods corresponding to the next month (e.g., historical synchronization time periods of the previous year or the previous two years, etc.).

In addition, besides predicting the category usage heat of the related commodity object category in the next month in the above manner, a corresponding initial category identification model may be established in units of days, for example, the following initial category identification model z1(t) ═ b1(1) × z1(t-1) + b1(2) × z1(t-2) + … is established, where z1(t) is the usage heat of a certain category on the date t of the present year, z1(t-1) is the usage heat of the category on the same date of the last year, z1(t-2) is the usage heat of the category on the same date of the previous year, and so on; b1(1), b1(2) and the like are coefficients to be estimated; then, the established initial category identification model can be trained by using the category historical data (such as 90 days) of the latest period of time and the corresponding contemporaneous historical data thereof, so as to obtain values of coefficients such as b1(1), b1(2) and the like, and obtain a finally required category identification model; and then, calculating the use heat of various types of purposes for 30 days backwards by using the trained category identification model, and then, adding and averaging to obtain the more stable use heat of the categories one month later, which is not described again here.

Further, in order to make the selected commodity objects more meet the user requirements and improve the transaction performance of the commodity objects, when the second time period is determined to be a specific time period such as a holiday, additional weighting can be performed on the use heat of various types of commodity objects related to the specific time period corresponding to the second time period (the degree of the additional weighting can be determined according to actual requirements), and the commodity objects corresponding to the types can be ensured to be revealed in time. For example, a moon cake may be hot in mid-autumn, and thus, when the second time period is a mid-autumn time period, additional weighting may be applied to the category corresponding to the moon cake.

That is to say, for any initially selected object, before updating the object usage heat of the initially selected object in the second time period according to the class usage heat of the class corresponding to the initially selected object in the second time period, the method may further include:

if it is determined that the second time period is a specific time period (for example, the holiday time period in the middle and autumn festival, the afternoon festival, and the like), if it is determined that the category corresponding to the initially selected object is a category matched with the specific time period corresponding to the second time period, increasing the category use heat of the category corresponding to the initially selected object in the second time period according to a set coefficient (the coefficient can be flexibly adjusted according to the actual situation, for example, if the degree of matching between the category and the specific time period is high, the coefficient can be large, and if the degree of matching is low, the coefficient can be small, and the like).

Further, since directly ordering the commodity objects according to the heat of use of the objects can cause the appearance of some homogeneous commodity objects, for example, two commodity objects of "the Italy imported Feliou chocolate rose DIY gift box containing birthday lovers package stamp" and "the Shunfeng package stamp Italy Feliou chocolate DIY heart-shaped rose gift box containing mid-autumn birthday gift" are very similar, and if the commodity objects are directly displayed together to the foreground, the commodity objects are single. Therefore, before at least one commodity object is selected from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period, the identity removing operation can be further carried out on each primary selection object.

and for each group of object sets consisting of at least one primary selection object with the similarity not lower than a set similarity threshold (the similarity threshold can be the same as or different from the similarity threshold mentioned at the first time), keeping one object in the object sets, and deleting other objects, so that at least one commodity object can be selected from the primary selection objects obtained after the deletion operation is performed as the commodity object matched with the commodity object subject word in the second time period.

That is, the similarity between the product objects can be obtained by calculating the similarity of the product object titles. In addition, the similarity calculation may employ the jaccard similarity formula J (a, B) ═ a crosses B |/| a and B | (i.e., the similarity of two headings is the number of words common to the two headings divided by the number of all words of the two headings). For example, assuming title a is "hazelnut chocolate" and title B is "milk chocolate", the similarity between the two is 1/3, since the title intersection has one word "chocolate" and the union of the two titles has three words. Of course, the similarity between the commodity objects may also be calculated by using any other similarity calculation formula, which is not limited herein.

Furthermore, for each group of object sets composed of at least one initially selected object with the similarity not lower than the set similarity threshold, when an object in the object set is reserved, an object with the highest heat of use of the corresponding object can be usually reserved, so that the transaction performance of the commodity object is improved, and the application experience of the user is improved.

In addition, taking the example of selecting a commodity object for a certain channel, if the channel needs to update the commodity object every day, the commodity object finally displayed by the channel may not be repeated within several days by configuring the relevant logic, and details are not repeated here.

Further, in this embodiment, in addition to performing the operation of going to the same for each primary selection object before at least one product object is selected from the primary selection objects as a product object matching with the product object subject term in the second time period, after at least one product object is selected from the primary selection objects as a product object matching with the product object subject term in the second time period, the operation of going to the same for each selected product object may be performed in a similar manner.

In addition, after selecting at least one commodity object from the primary selection objects as the commodity object matched with the commodity object subject term in the second time period, the server can also store commodity object information (such as the ID of the commodity object, the unique identification information such as the link of the commodity object) of the selected commodity object and/or send the commodity object information to the user terminal.

Step 207: and the user terminal receives the commodity object information of the commodity object matched with the commodity object subject term returned by the server, and takes the commodity object corresponding to the commodity object information as the commodity object matched with the commodity object subject term in a second time period.

Optionally, the user terminal may further display the received commodity object information, and/or display a commodity object corresponding to the commodity object information, so that the user can view the commodity object, which is not described in detail herein.

It should be noted that step 201, step 203, and step 207 independently constitute a product object selection process executed on the user terminal side, and step 202, step 204, to step 206 independently constitute a product object selection process executed on the server side, which will not be described again.

Further, as shown in fig. 3 and 4, the embodiment of the present application further provides two model determination methods. Specifically, as shown in fig. 3, which is a schematic flow chart of a model determining method in a first embodiment of the present application, the model determining method may include the following steps:

step 301: and receiving object recognition model training sample data sent by a user terminal.

Step 302: and training the pre-established initial object recognition model according to the basic characteristic data of each object recognition model sample object contained in the object recognition model training sample data to obtain the required object recognition model.

The initial object recognition model is used for predicting the association relationship between the use characteristic data of the commodity object in a first time period and the object use heat of the commodity object in a second time period; the first time period is a specified time period before the second time period.

The object identification model is used for representing the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat of the commodity object in the second time period.

Further, as shown in fig. 4, which is a schematic flow chart of another model determining method in the first embodiment of the present application, the another model determining method may include the following steps:

step 401: receiving class identification model training sample data sent by a user terminal.

The category identification model training sample data comprises basic feature data of various category identification model sample objects, the basic feature data of each category identification model sample object comprises category usage heat of categories corresponding to the category identification model sample object in various set sample time periods (such as a fifth time period in the commodity object selection method), and the category usage heat of the categories corresponding to the category identification model sample object in one or more historical same-period time periods corresponding to the various set sample time periods can be expressed as a first time period to be distinguished from a second time period mentioned later in the model determination method without being described in detail) if the commodity object selection method is not considered.

Step 402: and training the pre-established initial category identification model according to the basic characteristic data of the various category identification model sample objects contained in the category identification model training sample data to obtain the required category identification model.

The initial category identification model is used for predicting the association relationship between the category use heat of the commodity object category in the second time period and the category use heat of the commodity object category in one or more historical synchronization time periods corresponding to the second time period.

The category identification model is used for representing the association relationship between the category use heat of the commodity object category in a second time period and the category use heat of the commodity object category in one or more historical contemporaneous time periods corresponding to the second time period.

In addition, the execution subjects of the model determination methods shown in fig. 3 and 4 may be servers; the specific implementation of each step of the model determining method shown in fig. 3 and 4 can refer to the related description, which is not repeated.

Further, as shown in fig. 5, an embodiment of the present application also provides a method for determining a usage heat. The use heat determination method may include the steps of:

step 501: acquiring use characteristic data of each commodity object in a first time period;

step 502: determining the object use heat degree of each commodity object in the second time period based on the association relationship between the use characteristic data of the commodity object in the first time period and the object use heat degree of the commodity object in the second time period and the use characteristic data of each commodity object in the first time period;

wherein the first time period is a specified time period before the second time period; the incidence relation is established according to the use characteristic data of each sample object in each third time period and the object use heat of each sample object in each corresponding fourth time period; the third time period is a specified time period before the corresponding fourth time period.

The association relationship is similar to the set object recognition model described above; in addition, the specific implementation of each step of the method for determining the use of the heat degree shown in fig. 5 can refer to the related description, which is not repeated herein.

As can be seen from the content of the first embodiment of the present application, in the first embodiment of the present application, at least one commodity object can be automatically selected from a large number of objects as a final object meeting the user requirements based on the commodity object subject term input by the user and the set object recognition model, so that the selection efficiency of the commodity object is greatly improved, the cost of manual inventory is reduced, and the operation efficiency is improved.

In addition, according to the object use heat of each primary selection object in the second time period, at least one commodity object with the object use heat not lower than the set heat threshold value is selected from the primary selection objects to serve as a final required commodity object, and therefore the commodity object selection accuracy can be improved.

In addition, the use heat of the categories can be calculated based on a time series model, namely a category identification model, and the sequence of each commodity object can be adjusted according to the use heat of the categories, so that the commodity objects can be adjusted in time according to seasons, festivals, holidays and the like, the cost of manual goods taking is further reduced, and the operation efficiency is improved.

Finally, it should be noted that the solution described in the embodiment of the present application has no limitation of language, software, or hardware, and can be implemented based on a general cloud computing platform. However, in order to improve the selection efficiency of the channel object, a programming language with high performance (e.g., C, C + + or Java) and hardware with high performance may be preferably used for implementation, and details thereof are not described in this embodiment of the present application.

Example two:

based on the same inventive concept as the first embodiment of the present application, the second embodiment of the present application provides a commodity object selection device, and the specific implementation of the commodity object selection device may refer to the related description of the server in the first embodiment of the method, and repeated details are not repeated, as shown in fig. 6, the commodity object selection device may mainly include:

a subject word receiving unit 601, configured to receive a commodity object subject word sent by a user terminal;

an object obtaining unit 602, configured to obtain a primary selection object matched with the subject term of the commodity object;

a heat determination unit 603 configured to determine a target usage heat of each of the primary selection targets in the second time period based on the set target recognition model and the usage feature data of each of the primary selection targets in the first time period; the object identification model is an identification model which is obtained through training and used for representing the incidence relation between the use characteristic data of the commodity object in a first time period and the object use heat of the commodity object in a second time period; the first time period is a specified time period before the second time period;

the object screening unit 604 may be configured to select at least one commodity object from the primary selection objects as a commodity object matching the commodity object subject term in the second time period according to the object usage heat of each primary selection object in the second time period.

Optionally, the object filtering unit 604 may be specifically configured to select, from the initially selected objects, at least one commodity object whose usage heat is not lower than a set heat threshold as a commodity object matched with the commodity object topic word in the second time period.

Optionally, the merchandise object selection device may further include an object identification sample data receiving unit 605 and an object identification model determining unit 606:

the object identification sample data receiving unit 605 is configured to receive object identification model training sample data sent by the user terminal before determining the object usage heat of each initially selected object in the second time period based on a set object identification model and the usage feature data of each initially selected object in the first time period, where the object identification model training sample data includes basic feature data of each object identification model sample object, and the basic feature data of each object identification model sample object includes the usage feature data of the object identification model sample object in each third time period and the object usage heat of the object identification model sample object in each corresponding fourth time period; the third time period is a specified time period before the corresponding fourth time period;

the object recognition model determining unit 606 may be configured to train an initial object recognition model, which is pre-established to predict an association relationship between usage feature data of the commodity object in a first time period and object usage heat of the commodity object in a second time period, according to the basic feature data of each object recognition model sample object, to obtain the object recognition model.

Optionally, the merchandise object selection device may further include a category popularity determination unit 607 and an object popularity update unit 608:

the category heat determination unit 607 is configured to, before at least one commodity object is selected from the primary selection objects as a commodity object matching the commodity object subject word in the second time period, determine the category use heat of the category corresponding to each primary selection object in the second time period based on the set category identification model and the category use heat of the category corresponding to each primary selection object in one or more historical synchronization time periods corresponding to the second time period; the category identification model is an identification model which is obtained through training and used for representing the association relationship between the category use heat of the commodity object category in a second time period and the category use heat of the commodity object category in one or more historical synchronization time periods corresponding to the second time period;

the object heat updating unit 608 may be configured to, for each initially selected object, update the object usage heat of the initially selected object in the second time period according to the class usage heat of the class corresponding to the initially selected object in the second time period.

Optionally, the article object selecting apparatus may further include a category identification sample data receiving unit 609 and a category identification model determining unit 610:

the category identification sample data receiving unit 609 may be configured to receive category identification model training sample data sent by the user terminal before determining the category usage heat of the category corresponding to each primary selected object in the second time period based on the set category identification model and the category usage heat of the category corresponding to each primary selected object in one or more historical synchronization time periods corresponding to the second time period, wherein the category identification model training sample data comprises basic characteristic data of various category identification model sample objects, and the basic characteristic data of each category identification model sample object comprises the category use heat of the category corresponding to the category identification model sample object in each fifth time period, and the category usage heat of the category corresponding to the category identification model sample object in one or more historical contemporaneous time periods corresponding to the fifth time periods;

the category identification model determining unit 610 may be configured to train, according to basic feature data of each category identification model sample object, an initial category identification model that is pre-established and used to predict an association relationship between a category usage heat of the commodity object category in a second time period and a category usage heat of the commodity object category in one or more historical synchronization time periods corresponding to the second time period, so as to obtain the category identification model.

Optionally, the commodity object selecting device may further include a category heat updating unit 611:

the category heat degree updating unit 611 is configured to, for any one of the initially selected objects, before updating the object use heat degree of the initially selected object in the second time period according to the category use heat degree of the category corresponding to the initially selected object in the second time period, if it is determined that the second time period is the specific time period, increase the category use heat degree of the category corresponding to the initially selected object in the second time period according to a set coefficient if it is determined that the category corresponding to the initially selected object is the category matched with the specific time period corresponding to the second time period.

Optionally, the merchandise object selection device may further include a merchandise object similarity removing unit 612:

the commodity object identifying unit 612 may be configured to determine, before at least one commodity object is selected from the primary selection objects as a commodity object matched with the commodity object topic word in the second time period, a similarity between the primary selection objects according to a similarity between titles corresponding to the primary selection objects; and for each group of object sets consisting of at least one initially selected object with the similarity between the objects not lower than a first similarity threshold, keeping one object in the object sets, and deleting other objects.

Optionally, the merchandise object selecting device may further include a subject word expansion unit 613:

the subject term expansion unit 613 is configured to determine, based on a set sample corpus, at least one sample term whose similarity to the subject term of the commodity object is not lower than a second similarity threshold for each subject term of the commodity object sent by the user terminal before acquiring the primary selection object matched with the subject term of the commodity object; and taking each determined sample word as a final required commodity object subject word.

In addition, it should be noted that the object recognition model may be a regression model; the category identification model may be a linear model.

Further, based on the same inventive concept as the first embodiment of the present application, the second embodiment of the present application further provides another commodity object selection device, and specific implementation of the another commodity object selection device may refer to related description about the user terminal in the first embodiment of the method, and repeated details are not repeated, as shown in fig. 7, the another commodity object selection device may mainly include:

a subject term receiving unit 701, configured to receive a commodity object subject term input by a user;

a topic word sending unit 702, configured to send the topic words of the commodity objects to a server;

an object information receiving unit 703, configured to receive the commodity object information returned by the server according to the commodity object topic word;

an object determining unit 704, configured to determine a commodity object corresponding to the commodity object information as a commodity object matching the commodity object topic word in a second time period;

Optionally, the another merchandise object selection device may further include an object identification sample data receiving unit 705 and an object identification sample data sending unit 706:

the object identification sample data receiving unit 705 is configured to receive, before receiving a commodity object subject term input by a user, object identification model training sample data input by the user, where the object identification model training sample data includes basic feature data of each object identification model sample object, and the basic feature data of each object identification model sample object includes usage feature data of the object identification model sample object in each third time period and an object usage heat of the object identification model sample object in each corresponding fourth time period; the third time period is a specified time period before the corresponding fourth time period;

the object recognition sample data sending unit 706 may be configured to send the object recognition model training sample data to a server, where the server trains an initial object recognition model, which is pre-established and used to predict association between usage characteristic data of the commodity object in a first time period and object usage heat of the commodity object in a second time period, according to basic characteristic data of each object recognition model sample object included in the object recognition model training sample data, to obtain the set object recognition model.

Optionally, the another merchandise object selection device may further include a category identification sample data receiving unit 707 and a category identification sample data sending unit 708:

the category identification sample data receiving unit 707 may be configured to receive, before receiving the commodity object information returned by the server according to the commodity object subject term, category identification model training sample data input by a user, where the category identification model training sample data includes basic feature data of each category identification model sample object, and the basic feature data of each category identification model sample object includes a category usage heat of a category corresponding to the category identification model sample object in each fifth time period and a category usage heat of a category corresponding to the category identification model sample object in one or more historical contemporaneous time periods corresponding to each fifth time period;

the category identification sample data sending unit 708 may be configured to send the category identification model training sample data to a server, where the server trains, according to basic feature data of each category identification model sample object included in the category identification model training sample data, an initial category identification model that is pre-established and used to predict an association relationship between a category usage heat of the commodity object category in the second time period and category usage heats of one or more historical contemporaneous time periods corresponding to the second time period of the commodity object category, to obtain a category identification model that is used to represent an association relationship between a category usage heat of the commodity object category in the second time period and a category usage heat of the commodity object category in one or more historical contemporaneous time periods corresponding to the second time period.

Further, based on the same inventive concept as that of the first embodiment of the present application, the second embodiment of the present application further provides a model determining apparatus, and specific implementation of the model determining apparatus may refer to the related description about the first model determining method in the first embodiment of the method, and repeated details are not repeated, as shown in fig. 8, the model determining apparatus may mainly include:

a data receiving unit 801, configured to receive object identification model training sample data sent by a user terminal, where the object identification model training sample data includes basic feature data of each object identification model sample object, and the basic feature data of each object identification model sample object includes usage feature data of the object identification model sample object in each third time period and object usage heat of the object identification model sample object in each corresponding fourth time period; the third time period is a specified time period before the corresponding fourth time period;

a model training unit 802, configured to train, according to basic feature data of each object recognition model sample object included in the object recognition model training sample data, an initial object recognition model that is pre-established and used for predicting an association relationship between usage feature data of the commodity object in a first time period and object usage heat of the commodity object in a second time period, so as to obtain an object recognition model that is used for representing an association relationship between usage feature data of the commodity object in the first time period and object usage heat of the commodity object in the second time period; the first time period is a specified time period before the second time period.

Further, based on the same inventive concept as the first embodiment of the present application, the second embodiment of the present application further provides another model determining apparatus, and specific implementation of the another model determining apparatus may refer to related description about another model determining method in the first embodiment of the method, and repeated details are omitted, as shown in fig. 9, the another model determining apparatus may mainly include:

a data receiving unit 901, configured to receive class identification model training sample data sent by a user terminal, wherein the category identification model training sample data comprises basic characteristic data of various category identification model sample objects, and the basic feature data of each category identification model sample object includes the category usage heat of the category corresponding to the category identification model sample object in each set sample time period (e.g. the fifth time period in the aforementioned commodity object selection method; in addition, if the aforementioned commodity object selection method is not considered, the set sample time period can also be expressed as the first time period to be distinguished from the second time period mentioned later by the model determination device, which is not described again), and the category usage heat of the category corresponding to the category identification model sample object in one or more historical synchronization time periods corresponding to each set sample time period;

the model training unit 902 may be configured to train, according to basic feature data of each category identification model sample object included in the category identification model training sample data, an initial category identification model that is pre-established and used to predict an association relationship between a category usage heat of the commodity object category in a second time period and category usage heats of the commodity object category in one or more historical contemporaneous time periods corresponding to the second time period, to obtain a category identification model that is used to represent an association relationship between a category usage heat of the commodity object category in the second time period and category usage heats of the commodity object category in one or more historical contemporaneous time periods corresponding to the second time period.

Further, based on the same inventive concept as that of the first embodiment of the present application, the second embodiment of the present application further provides a device for determining a usage heat, and specific implementation of the device for determining a usage heat may refer to related description about the method for determining a usage heat in the first embodiment of the method, and repeated details are omitted, as shown in fig. 10, the device for determining a usage heat may mainly include:

a data acquisition unit 1001 operable to acquire usage characteristic data of each commodity object in a first time period;

a heat determination unit 1002, configured to determine an object usage heat of each commodity object in a second time period based on an association relationship between usage characteristic data of the commodity object in a first time period and an object usage heat of the commodity object in the second time period, and usage characteristic data of each commodity object in the first time period;

Finally, it should be noted that any designations (e.g., first time period, second time period, etc.) in the description of the embodiments and the drawings of the present application are used for distinction only and are not meant to be limiting in any way.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, apparatus (device), or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (devices) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all alterations and modifications as fall within the scope of the application.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.

Claims

1. A method for selecting a commodity object, comprising:

receiving a commodity object subject term sent by a user terminal;

for each primary selection object, updating the object use heat of the primary selection object in the second time period according to the class use heat of the class corresponding to the primary selection object in the second time period;

2. The method of claim 1, wherein selecting at least one merchandise object from the primary selection objects as the merchandise object matching the merchandise object subject term in the second time period comprises:

3. The method of claim 1, wherein before determining the object usage heat of each of the first selected objects in the second time period based on the set object recognition model and the usage characteristic data of each of the first selected objects in the first time period, the method further comprises:

4. The method according to claim 3, wherein the usage characteristic data includes at least any one or more of a number of brows, a number of collections, a number of buys, a number of deals, a number of reviews, and a number of searches; the subject use heat includes at least any one or more of a volume of a transaction, and a rate of conversion of a transaction.

5. The method of claim 1, wherein before determining the usage heat of the category corresponding to each of the first selected objects in the second time period based on the set category identification model and the usage heat of the category corresponding to each of the first selected objects in one or more historical contemporaneous time periods corresponding to the second time period, the method further comprises:

6. The method of claim 5, wherein the historical contemporaneous time period of each time period is a historical time period corresponding to the time period on the same Gregorian or lunar calendar day as the time period.

7. The method of claim 1, wherein for any of the initially selected objects, before updating the object usage heat of the initially selected object in the second time period according to the class usage heat of the class corresponding to the initially selected object in the second time period, the method further comprises:

8. The method of claim 1, wherein prior to selecting at least one merchandise object from the primary selection objects as a merchandise object matching the merchandise object subject term within the second time period, the method further comprises:

9. The method of claim 1, wherein prior to obtaining the first selected object that matches the item object subject term, the method further comprises:

10. The method of claim 1, wherein the object recognition model is a regression model; the category identification model is a linear model.

11. The method of claim 1, wherein the usage heat is determined by:

12. A method for selecting a commodity object, comprising:

the commodity object information is related information of the commodity object selected by the server from the primary selection objects matched with the commodity object subject words according to a set object identification model and the use characteristic data of each primary selection object matched with the commodity object subject words in a first time period and on the basis of the set category identification model and the category use heat of the categories corresponding to each primary selection object in one or more historical synchronization time periods corresponding to a second time period; the first time period is a specified time period before the second time period;

the set object identification model is used for representing the association relationship between the use characteristic data of the commodity object in a first time period and the object use heat of the commodity object in a second time period;

the set category identification model is used for representing the association relationship between the category use heat of the commodity object category in the second time period and the category use heat of one or more historical synchronization time periods corresponding to the commodity object category in the second time period.

13. The method of claim 12, wherein prior to receiving the user-entered item object subject term, the method further comprises:

14. A method of model determination, comprising:

15. The method of claim 14, wherein the category heat of use includes at least any one or more of volume, and conversion of volume.

16. A merchandise object selection device, comprising:

17. A merchandise object selection device, comprising:

the commodity object information is related information of the commodity object selected by the server from the primary selection objects matched with the commodity object subject words according to a set object identification model and the use characteristic data of each primary selection object matched with the commodity object subject words in a first time period and on the basis of the set category identification model and the category use heat of each primary selection object corresponding to one or more historical synchronization time periods corresponding to a second time period; the first time period is a specified time period before the second time period;

18. A model determination apparatus, comprising: