CN114707075A

CN114707075A - Cold start recommendation method and device

Info

Publication number: CN114707075A
Application number: CN202210631258.6A
Authority: CN
Inventors: 徐晓宇; 冯晓东
Original assignee: Honor Device Co Ltd
Current assignee: Honor Device Co Ltd
Priority date: 2022-06-06
Filing date: 2022-06-06
Publication date: 2022-07-05
Anticipated expiration: 2042-06-06
Also published as: CN114707075B

Abstract

The embodiment of the application provides a cold start recommendation method and equipment, relates to the technical field of terminals, and can solve the problem of cold start of a recommendation system. The method is applied to the server and specifically comprises the following steps: receiving a recommendation list acquisition request from the terminal equipment, wherein the recommendation list acquisition request comprises the model of the terminal equipment used by the new user; determining a first commodity access record and a first version block access record corresponding to a first model, wherein the first application is a shopping application, and the second application is a community application; obtaining a first recommendation list according to the first commodity access record; obtaining a second recommendation list according to the first version access record; determining a similar model of the first model and a second commodity access record corresponding to the similar model, obtaining a third recommendation list according to the second commodity access record, and performing duplicate removal and sorting on commodity information in the first recommendation list, the second recommendation list and the third recommendation list to obtain a final recommendation list; and sending the final recommendation list to the terminal equipment.

Description

Cold start recommendation method and device

Technical Field

The application relates to the technical field of terminals, in particular to a cold start recommendation method and device.

Background

Recommendation technology is widely applied in internet products of today, such as search result recommendation of browser websites, topic recommendation of community websites, commodity recommendation of shopping websites, friend making recommendation of social websites, and the like. However, most of the existing recommendations are obtained by a recommendation system based on a large amount of user data, and when a new user uses an internet product for the first time, the recommendation system cannot accurately recommend the new user because the new user does not have user data.

The problem that the new user cannot be accurately recommended is the cold start of the recommendation system. In contrast, most internet products design related guide interfaces to guide new users to log in by using social network accounts, so that user data can be acquired according to the social network accounts of the users, and personalized recommendation can be performed for the users. However, in order to protect privacy, a new user may avoid logging in by using a social network account, so that internet products cannot acquire user data, and a recommendation system cannot make accurate recommendations for the new user.

Disclosure of Invention

In view of this, the present application provides a cold start recommendation method and device, which can solve the cold start problem of a recommendation system.

In a first aspect, the present application provides a cold start recommendation method, which is applied to a server, and includes: the server receives a recommendation list acquisition request from the terminal equipment, wherein the recommendation list acquisition request comprises the model of the terminal equipment used by the new user, and the model of the terminal equipment is a first model; determining a first commodity access record and a first version access record corresponding to a first model, wherein the first commodity access record is obtained by screening from a commodity access record of a first application, the first version access record is obtained by screening from a version access record of a second application, the first application is a shopping application, and the second application is a community application; obtaining a first recommendation list according to the first commodity access record; obtaining a second recommendation list according to the first version access record; determining a similar model of the first model and a second commodity access record corresponding to the similar model, and obtaining a third recommendation list according to the second commodity access record, wherein the first recommendation list, the second recommendation list and the third recommendation list all comprise commodity information and a scoring result corresponding to the commodity; removing duplication and sorting commodity information in the first recommendation list, the second recommendation list and the third recommendation list to obtain a final recommendation list, wherein the final recommendation list is used for recommending commodities for a new user; and the server sends the final recommendation list to the terminal equipment.

Generally, when a user wants to purchase a commodity, the user may select to query the function introduction of the commodity through the community-type application, and then determine whether to purchase the commodity according to the function introduction of the commodity, or select another commodity having a function equivalent to that of the commodity. After the commodity which is wanted to be purchased is determined, the shopping application is opened to complete the commodity purchase.

Based on the user purchasing habits, the cold start problem of the recommendation system is solved by utilizing the commodity access records of the shopping applications and the block access records of the community applications. As the commodity access records of the shopping application and the plate access records of the community application are large in quantity, the server conducts preliminary screening on the commodity access records and the plate access records according to the model of the terminal equipment used by the new user under the condition that other effective data of the new user do not exist, and the first commodity access records and the first plate access records are obtained. And then, the first commodity access record and the first version block access record are respectively analyzed to obtain a first recommendation list and a second recommendation list.

Since the first recommendation list is obtained according to the first commodity access record, the first recommendation list is used for determining the commodity preference of the user in the aspect of historical shopping. Since the second recommendation list is obtained according to the first edition access record, the second recommendation list is used for determining the commodity preference of the user in the aspect of community editions. Meanwhile, the third recommendation list is determined according to the similar model of the first model. With similar models, the resulting third recommendation list may also express user preferences, as the user audience of the similar model has an overlapping portion with the user audience of the first model.

And finally, processing the first recommendation list, the second recommendation list and the third recommendation list to obtain a final recommendation list, wherein the final recommendation list is judged from multiple aspects, such as historical shopping, community block and similar models, and is more in line with the preference of a new user. Therefore, the problem of cold start is solved, and the recommended commodities are more in line with the preference of a new user, so that the user viscosity can be further increased, and the user satisfaction is improved.

In one possible implementation, after the third recommendation list is derived, the method further includes: determining a first type word corresponding to the post related to each block access record in the first block access records based on the post related to each block access record in the first block access records, and creating a mapping relation between the post and the first type word, wherein the first type word is a word related to a product name in the post; determining a post preferred by the user based on the first version block access record; determining a first type word corresponding to a post preferred by a user according to the mapping relation; determining commodity information corresponding to the first type of words by using a fuzzy matching algorithm, and generating a fourth recommendation list according to the commodity information corresponding to the first type of words; removing duplication and sorting commodity information in the first recommendation list, the second recommendation list and the third recommendation list to obtain a final recommendation list, wherein the steps of: and performing duplicate removal and sorting on the commodity information in the first recommendation list, the second recommendation list, the third recommendation list and the fourth recommendation list to obtain a final recommendation list.

In the scheme, the posts in the first version block access records are scored, and the posts preferred by the user are determined. And then, based on the mapping relation between the posts and the words (first type words) related to the product names in the posts, obtaining the first type words corresponding to the posts preferred by the user. The user preferred posts can determine the attention points of the user, and then the attention points of the user are further processed, namely, the first type words corresponding to the user preferred posts are determined, so that the words which are concerned by the user and are related to the product name are obtained. Because the words related to the product name cannot be directly corresponding to the product, a fuzzy matching algorithm is also needed to be used for matching the words related to the product name with the product, and a fourth recommendation list is obtained after matching. According to the method and the device, the preference of the user is determined through the posts, which is equivalent to a fourth recommendation list more fitting the user requirements from the user attention point. And finally, processing the first recommendation list, the second recommendation list, the third recommendation list and the fourth recommendation list to obtain a final recommendation list. Therefore, the commodities in the final recommendation list are more in line with the preference of the user, and the accuracy is higher.

In one possible implementation, after the third recommendation list is derived, the method further includes: determining a second type word corresponding to the post related to each block access record in the first block access record based on the post related to each block access record in the first block access record, wherein the second type word is a word related to the product characteristic in the post; sorting according to the frequency of all the second type words to generate a first sorting result; associating commodity selling points based on the second type words in the first sequencing result to generate a fifth recommendation list, wherein the fifth recommendation list comprises commodity information corresponding to the commodity selling points; removing duplication and sorting commodity information in the first recommendation list, the second recommendation list and the third recommendation list to obtain a final recommendation list, wherein the steps of: and removing the weight of the commodity information in the first recommendation list, the second recommendation list, the third recommendation list and the fifth recommendation list, and sequencing to obtain a final recommendation list. Or removing duplication and sorting commodity information in the first recommendation list, the second recommendation list, the third recommendation list, the fourth recommendation list and the fifth recommendation list to obtain a final recommendation list.

In the scheme of the application, words (second type words) related to product characteristics are determined from each layout access record, and then the second type words are sequenced to obtain a first sequencing result. The first sequencing result can indicate the product characteristics concerned by the user, then commodity selling points are related according to the product characteristics concerned by the user, namely, commodities are related according to the product characteristics concerned by the user, and finally a fifth recommendation list is obtained according to the related commodities. The method and the device for determining the product characteristics concerned by the user according to the plate access records accessed by the user, and further obtain the commodity concerned by the user. The determined commodities are obtained according to the user layout access record, and the layout access record is the real expression of the user intention, so that the commodities in the fifth recommendation list are more in line with the user requirements. And finally, the first recommendation list, the second recommendation list, the third recommendation list, the fourth recommendation list and the fifth recommendation list are arranged and combined, and after arrangement and combination, duplicate removal and sorting are carried out to obtain a final recommendation list. The items in the final recommendation list are more in line with the user's true intention.

In one possible implementation manner, deriving the first recommendation list according to the first article access record includes: scoring the commodities related to each commodity access record in the first commodity access record to obtain a scoring result of the commodities related to each commodity access record; accumulating the scoring results of the same commodity to obtain the scoring results of all commodities; and sorting the commodities from high to low according to the scoring results of all commodities to obtain a first recommendation list.

In the scheme of the application, the first recommendation list is obtained by scoring the commodities related to the first commodity access record. Therefore, the commodities in the first recommendation list are more objective and more accurate.

In a possible implementation manner, scoring the commodities related to each commodity access record in the first commodity access record to obtain a scoring result of the commodities related to each commodity access record includes: based on the first interaction behavior of the user and the commodities in each commodity access record, scoring the commodities related to each commodity access record to obtain a first scoring result; based on the occurrence time of the first interaction behavior in each commodity access record, scoring the commodities related to each commodity access record to obtain a second scoring result; and multiplying the first scoring result and the second scoring result to obtain the scoring result of the commodity related to each commodity access record.

The different first interaction behavior can account for different purchasing intentions of the user, for example, the purchasing intention expressed by the behavior that the user joins the shopping cart is far larger than the purchasing intention expressed by the behavior that the user browses. And the purchasing intention of the user can be changed along with the time, so that the commodity is scored based on the first interactive behavior and the occurrence time of the first interactive behavior, the obtained scoring result is more objective, and the purchasing intention of the user can be expressed better.

In a possible implementation manner, scoring the commodities related to each commodity access record based on the first interaction behavior of the user and the commodities in each commodity access record to obtain a first scoring result includes: and comparing the first interaction behaviors of the user and the commodity with the behaviors in the behavior scoring rule based on the first interaction behaviors of the user and the commodity in each commodity access record to obtain a first scoring result, wherein the behavior scoring rule comprises behaviors and scoring results corresponding to the behaviors. The scheme of the application discloses that the scoring result of the first interactive behavior is obtained by comparing the first interactive behavior with the behavior in the behavior scoring rule.

In a possible implementation manner, scoring the commodities related to each commodity access record based on the occurrence time of the first interaction behavior in each commodity access record to obtain a second scoring result, includes: and inputting the occurrence time of the first interaction behavior into a time attenuation function based on the occurrence time of the first interaction behavior in each commodity access record to obtain a second scoring result. The scheme of the application discloses that the occurrence time of the first interactive behavior is scored by using a time attenuation function.

In one possible implementation, the deriving the second recommendation list according to the first version access record includes: determining a user preference block and a user group corresponding to the user preference block based on the first block access record; and determining the commodities preferred by the users in the user group from the commodity access records based on the names of the users in the user group, and generating a second recommendation list based on the commodities preferred by the users.

Since many tiles are involved in community-like applications, each tile is also very different, for example, a cell phone tile and a game tile. Therefore, the user preference block is determined from the plurality of blocks based on the first block access record, so that the data processing difficulty can be greatly reduced, and the user preference block can be accurately positioned. And then, determining a user group according to the user preference plate, and then determining the commodities preferred by the users in the user group from the commodity access record according to the user group, which is equivalent to linking community applications and shopping applications, so that the users firstly determine the performance of the commodities, and then select the commodities of the mood finder according to the performance of the commodities. Therefore, the shopping habit of the user is better fitted, and the obtained commodities are more in line with the requirements of the user.

In a possible implementation manner, determining a user preference section and a user group corresponding to the user preference section based on the first section access record includes: and based on the second interaction behavior between the user and the layout in each layout access record in the first layout access record and the occurrence time of the second interaction behavior, scoring the layout related to each layout access record to obtain the user preference layout and a user group corresponding to the user preference layout.

The different second interactive behavior can account for different degrees of user interest in the posts in the section, e.g., the user comment post behavior expresses a degree of interest that is much greater than the user browse behavior. And the attention point of the user can change along with the time, so the method and the device determine the layout blocks which are more concerned by the user based on the second interactive behaviors and the occurrence time of the second interactive behaviors, and the obtained user preference layout blocks can be more objective and can express the attention point of the user.

In one possible implementation, before determining the similar model of the first model, the method further includes: distinguishing the commodity access records according to the equipment models, and obtaining a user behavior sequence corresponding to each equipment model based on the distinguished commodity access records, wherein the user behavior sequence comprises products accessed by the user; determining similar models of the first model, comprising: determining the similarity between the user behavior sequence of the first model and the user behavior sequence of the second model, wherein when the similarity between the user behavior sequence of the first model and the user behavior sequence of the second model meets a threshold value, the second model is the similar model of the first model, and the second model is all models except the first model; determining a second commodity access record corresponding to the similar model, and obtaining a third recommendation list according to the second commodity access record, wherein the third recommendation list comprises: distinguishing the commodity access records according to the equipment models to obtain second commodity access records corresponding to similar models; based on the second commodity access record, scoring the commodities related to the second commodity access record to obtain commodity scoring results corresponding to similar equipment models; and generating a third recommendation list based on the commodity scoring result.

The method also considers commodity recommendation for the new user from the similar model of the first model. The similar model is determined according to the similarity of the user behavior sequence corresponding to each model in the commodity access record. And then obtaining a third recommendation list according to the second access records with similar models. Because the similar model is determined according to the similarity of the user behavior sequences, the user behavior sequences comprise commodities visited by the user, the similarity of the user behavior sequences can also be the similarity of the commodities, if the similarity of the user behavior sequences of the two models is higher, the situation that the demands of the commodity of the user using the model and the demands of the commodity using the user using the first model overlap is shown, and the model is taken as the similar model. Therefore, the third recommendation list obtained by utilizing the second access record with the similar model is more in line with the user requirements.

In a possible implementation manner, the model of the terminal device used by the new user is determined to be the first model through an application program interface API disclosed by an operating system of the terminal device.

In a possible implementation manner, the commodity access record of the first application and the block access record of the second application are obtained through page buried points, the page buried points include handwriting buried points and non-buried points, the handwriting buried points are located at preset detection positions, if data are detected, the data are reported, and the non-buried points are all data reported. The application discloses that a commodity access record of a first application and a layout access record of a second application are obtained by utilizing two modes, namely a handwriting point burying mode and a non-point burying mode.

In one possible implementation, the first interaction behavior of the user with the commodity includes clicking to enter a commodity detail page, collecting the commodity, adding a shopping cart, clicking to immediately purchase the commodity, submitting a commodity purchase order, successfully paying the commodity order, canceling the commodity collection, sharing the commodity, searching for the commodity, successfully confirming the commodity arrival notice, consulting the commodity, commenting the commodity, and clicking to browse the commodity in the footprint.

In one possible implementation manner, the second interaction behavior of the user and the layout block includes browsing of posts under the layout block by the user, comments under the posts by the user, replies to the comments under the posts by the user, sharing of the posts by the user, favorite posts by the user, praise of the posts by the user, detailed information of the user clicking the comment posts by the user, posts issued by the user clicking the comment posts by the user, and appreciation of the posts by the user.

In a second aspect, the present application further provides another cold start recommendation method, which is applied to a terminal device, and the method includes: the method comprises the steps that the terminal equipment responds to the operation of logging in a first application by a new user, and sends a recommendation list acquisition request to a server, wherein the recommendation list acquisition request comprises the model of the terminal equipment used by the new user, and the model of the terminal equipment is a first model; the terminal equipment receives a final recommendation list sent by the server; and the terminal equipment recommends commodities for the new user according to the final recommendation list. According to the scheme, when a user triggers commodity recommendation, the terminal equipment acquires a final recommendation list through the server, and then recommends commodities for a new user according to the final recommendation list.

In a third aspect, a server is provided, which has the function of implementing the method of the first aspect. The function can be realized by hardware, and can also be realized by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the functions described above.

In a fourth aspect, a terminal device is provided, which has the function of implementing the method of the second aspect. The function can be realized by hardware, and can also be realized by executing corresponding software by hardware. The hardware or software includes one or more modules corresponding to the functions described above.

In a fifth aspect, a server is provided, which includes: a wireless communication module, memory, and one or more processors; the wireless communication module and the memory are coupled with the processor;

wherein the memory is configured to store computer program code, the computer program code comprising computer instructions; the computer instructions, when executed by the processor, cause the server to perform the method of any one of the first aspects as described above.

In a sixth aspect, a terminal device is provided, the terminal device comprising: a wireless communication module, memory, and one or more processors; the wireless communication module and the memory are coupled with the processor;

wherein the memory is configured to store computer program code, the computer program code comprising computer instructions; the computer instructions, when executed by the processor, cause the electronic device to perform the method of the second aspect as described above.

In a seventh aspect, a computer-readable storage medium is provided, which stores instructions that, when executed on a computer, enable the computer to perform the cold start recommendation method of any one of the first aspect and the cold start recommendation method of any one of the second aspect.

In an eighth aspect, there is provided a computer program product comprising instructions which, when run on a computer, cause the computer to perform the cold start recommendation method of any one of the first aspects above and the cold start recommendation method of any one of the second aspects above.

In a ninth aspect, there is provided an apparatus (which may be a system-on-a-chip, for example) comprising a processor configured to enable a server to implement the functions referred to in the first aspect above, and to enable a terminal device to implement the functions referred to in the second aspect above. In one possible design, the apparatus further includes a memory for storing program instructions and data necessary for the server and program instructions and data necessary for the terminal device. When the device is a chip system, the device may be composed of a chip, or may include a chip and other discrete devices.

The technical effects brought by any one of the design manners in the third aspect to the ninth aspect may be referred to the technical effects brought by the different design manners in the first aspect and the second aspect, and are not described herein again.

Drawings

Fig. 1 is a schematic view of an interface display for logging in a first application according to an embodiment of the present disclosure;

fig. 2 is a schematic structural diagram of a system architecture according to an embodiment of the present application;

fig. 3 is a schematic diagram of a hardware structure of a terminal device according to an embodiment of the present disclosure;

fig. 4 is a schematic diagram of a software structure of a terminal device according to an embodiment of the present application;

fig. 5 is a schematic hardware structure diagram of a server according to an embodiment of the present disclosure;

fig. 6 is a flowchart illustrating a cold start recommendation method according to an embodiment of the present application;

fig. 7 is a second flowchart illustrating a cold start recommendation method according to a second embodiment of the present application;

fig. 8 is a third schematic flowchart of a cold start recommendation method according to an embodiment of the present application;

fig. 9 is a fourth flowchart illustrating a cold start recommendation method according to an embodiment of the present application;

fig. 10 is a fifth flowchart illustrating a cold start recommendation method according to an embodiment of the present application;

fig. 11 is a sixth schematic flowchart of a cold start recommendation method according to an embodiment of the present application;

fig. 12 is a schematic application diagram of a cold start recommendation method according to an embodiment of the present application;

fig. 13 is a schematic structural diagram of a chip system according to an embodiment of the present disclosure.

Detailed Description

The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. In the description of the present application, unless otherwise specified, "at least one" means one or more, "a plurality" means two or more. In addition, in order to facilitate clear description of technical solutions of the embodiments of the present application, in the embodiments of the present application, terms such as "first" and "second" are used to distinguish the same items or similar items having substantially the same functions and actions. Those skilled in the art will appreciate that the terms "first," "second," etc. do not denote any order or quantity, nor do the terms "first," "second," etc. denote any order or importance.

With the diversification of internet products, the number of information items, the number of commodities and the types of commodities are rapidly increasing, and users usually spend a lot of time to find the information or the commodities required by the users. And with the browsing of a large amount of irrelevant information and products, the browsing desire or purchasing desire of the user can be gradually reduced, thereby causing the user to lose.

In order to solve the above problems, a recommendation system is developed, and the recommendation system can be understood as follows: the simulation customer service staff helps the user to quickly find the information or the commodity required by the user on the Internet platform. And the personalized recommendation in the recommendation system is to recommend information and commodities which are interested by the user to the user according to the interest characteristic data and the purchasing behavior data of the user, so that the stickiness between the user and an internet platform can be improved, and the loss of customers is avoided.

However, the recommendation system can accurately recommend information or goods to the user only when the user data such as the interest characteristic data and the purchasing behavior data of the user are acquired. If a new user logs in the internet platform for the first time, the interest of the new user cannot be predicted according to the historical behavior of the new user because the internet platform does not have user data of the new user, so that personalized recommendation cannot be made for the new user, which is also a problem of cold start of the user in the recommendation system.

In practical application, for the problem of cold start of a user, most internet platforms guide a new user to log in through a social account through a carefully set interactive interface when the new user logs in for the first time, so as to obtain social account information of the new user, for example: concern relationships, friend relationships, release content, and the like. And then, generating an interest model for the new user by using the social account information of the new user, and further performing accurate recommendation for the new user. But to protect privacy, new users typically refuse to log in using social accounts.

Referring to fig. 1 (a), the mobile phone may receive a click operation of a user on an icon 102 of a mall APP in a main interface (i.e., a desktop) 101 of the mobile phone. If the mall APP does not run, the mobile phone can respond to the click operation and start the mall APP. After the mall APP is started, if the user is a new user of the mall APP, the mall APP displays an interface 103 shown in fig. 1 (b). A WeChat-account quick-login button, a QQ-account quick-login button, and an alternate-login-mode button 104 (e.g., login via a cell-phone number) may be included in interface 103. In response to the user operating the switch login mode button 104 at the interface 103, the user logs in the mall APP by other login modes. In this way, the mall APP cannot acquire social contact account information of the user, and further cannot acquire more user information.

Alternatively, when a new user logs in for the first time, the application platform provides coarse-grained interest options for the user to capture the user interests. For example, music-based applications may allow users to select types of music that they like to listen to (rock, classical, ballad, etc.), and fitness-based applications may allow users to select types of sports that they are interested in (yoga, rope jump, boxing, etc.). But such interest selection interfaces require elaborate interactive design that would otherwise affect the user experience. Also, the interest of the new user at registration and the interest of the new user at real access may change, and the captured user interest may not be used for recommendation.

Therefore, the cold start recommendation method and the cold start recommendation device are provided, and other user behavior data corresponding to the model are captured according to the model of the terminal device used by the new user. And based on other user behavior data corresponding to the model, accurate recommendation is made for the new user.

Fig. 2 is a schematic diagram of a system architecture according to an embodiment of the present application. The system may include a terminal device and a server. Wherein, the terminal device can be installed with an Application (APP), and the application can be a shopping application (e.g., naobao APP, glory mall APP, kyoto APP) and a community application (e.g., glory club APP, bean APP), etc. The server may be a management server of the application or a cloud server.

In the embodiment of the application, the terminal responds to the operation of logging in the first application by the new user, and sends a request for obtaining the recommendation list to the server, wherein the recommendation list request comprises the model (the first model) of the terminal device used by the new user. The server responds to the request for obtaining the recommendation list, and determines a first commodity access record and a first version block access record corresponding to the first model. And then obtaining a first recommendation list according to the first commodity access record, and obtaining a second recommendation list according to the first version access record. And then the server determines the similar model of the first model and a second commodity access record corresponding to the similar model, and obtains a third recommendation list according to the second access record. And finally, de-duplicating and sequencing the first recommendation list, the second recommendation list and the third recommendation list to obtain a final recommendation list, and sending the final recommendation list to the terminal equipment, wherein the terminal equipment recommends the user according to the commodities in the final recommendation list.

In one possible design, the terminal device first sends a data acquisition request to the server in response to an operation of logging in the first application by a new user, wherein the data acquisition request is used for requesting to acquire the commodity access record of the first application and the block access record of the second application. And the server responds to the data acquisition request and sends the commodity access record of the first application and the plate access record of the second application to the terminal equipment. After the terminal device receives the commodity access record of the first application and the version access record of the second application, the commodity access record is distinguished according to the type (first type) of the terminal device used by a new user, and a first commodity access record and a first version access record corresponding to the first type are obtained. And then obtaining a first recommendation list according to the first commodity access record, and obtaining a second recommendation list according to the first version access record. And then determining a similar model of the first model and a second commodity access record corresponding to the similar model, and obtaining a third recommendation list according to the second access record. And finally, de-duplicating and sequencing the first recommendation list, the second recommendation list and the third recommendation list to obtain a final recommendation list, and recommending the commodities for the user by the terminal equipment according to the final recommendation list.

For example, the terminal device in the embodiment of the present application may be a mobile phone, a tablet computer, a desktop computer (desktop computer), a handheld computer, a notebook computer (laptop), an ultra-mobile personal computer (UMPC), a netbook, a Personal Digital Assistant (PDA), an Augmented Reality (AR) \ Virtual Reality (VR) device, and the like, which may be installed with the target application, and the embodiment of the present application does not particularly limit the specific form of the first device.

The operating system of the terminal device may be an Android (Android), a damming system, an IOS system, or another operating system, and the type of the operating system of the first device is not limited in the embodiment of the present application.

In this embodiment, taking the example that the terminal device shown in fig. 3 is an electronic device 300 (such as a mobile phone), a structure of the terminal device provided in this embodiment is illustrated. As shown in fig. 3, the electronic device 300 may include a processor 310, an external memory interface 320, an internal memory 321, a Universal Serial Bus (USB) interface 330, a charging management module 340, a power management module 341, a battery 342, an antenna 1, an antenna 2, a mobile communication module 350, a wireless communication module 360, an audio module 370, a speaker 370A, a receiver 370B, a microphone 370C, an earphone interface 370D, a sensor module 380, a button 390, a motor 391, an indicator 392, a camera 393, a display 394, and a Subscriber Identification Module (SIM) card interface 395, a pupil detector 396, and an iris detector 397, and the like.

The sensor module 380 may include a pressure sensor, a gyroscope sensor, an air pressure sensor, a magnetic sensor, an acceleration sensor, a distance sensor, a proximity light sensor, a fingerprint sensor, a temperature sensor, a touch sensor, an ambient light sensor, a bone conduction sensor, and the like.

It is to be understood that the illustrated structure of the present embodiment does not constitute a specific limitation to the electronic device 300. In other embodiments, electronic device 300 may include more or fewer components than illustrated, or combine certain components, or split certain components, or a different arrangement of components. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

Processor 310 may include one or more processing units, such as: the processor 310 may include an Application Processor (AP), a modem processor, a Graphics Processor (GPU), an Image Signal Processor (ISP), a controller, a memory, a video codec, a Digital Signal Processor (DSP), a baseband processor, and/or a neural-Network Processing Unit (NPU), among others. The different processing units may be separate devices or may be integrated into one or more processors.

The controller may be a neural center and a command center of the electronic device 300. The controller can generate an operation control signal according to the instruction operation code and the timing signal to complete the control of instruction fetching and instruction execution.

A memory may also be provided in the processor 310 for storing instructions and data. In some embodiments, the memory in the processor 310 is a cache memory. The memory may hold instructions or data that have just been used or recycled by the processor 310. If the processor 310 needs to reuse the instruction or data, it can be called directly from memory. Avoiding repeated accesses reduces the latency of the processor 310, thereby increasing the efficiency of the system.

In some embodiments, processor 310 may include one or more interfaces. The interface may include an integrated circuit (I3C) interface, an integrated circuit built-in audio (I3S) interface, a Pulse Code Modulation (PCM) interface, a universal asynchronous receiver/transmitter (UART) interface, a Mobile Industry Processor Interface (MIPI), a general-purpose input/output (GPIO) interface, a Subscriber Identity Module (SIM) interface, and/or a Universal Serial Bus (USB) interface, etc.

It should be understood that the connection relationship between the modules illustrated in the present embodiment is only an exemplary illustration, and does not limit the structure of the electronic device 300. In other embodiments, the electronic device 300 may also adopt different interface connection manners or a combination of multiple interface connection manners in the above embodiments.

The charging management module 340 is configured to receive charging input from a charger. The charging management module 340 may also supply power to the electronic device 300 through the power management module 341 while charging the battery 342.

The power management module 341 is configured to connect the battery 342, the charging management module 340 and the processor 310. The power management module 341 receives input from the battery 342 and/or the charge management module 340 and provides power to the processor 310, the internal memory 321, the external memory, the display 394, the camera 393, and the wireless communication module 360. In other embodiments, the power management module 341 may also be disposed in the processor 310. In other embodiments, the power management module 341 and the charging management module 340 may be disposed in the same device.

The wireless communication function of the electronic device 300 may be implemented by the antenna 1, the antenna 2, the mobile communication module 350, the wireless communication module 360, a modem processor, a baseband processor, and the like.

The antennas 1 and 2 are used for transmitting and receiving electromagnetic wave signals. Each antenna in the electronic device 300 may be used to cover a single or multiple communication bands. Different antennas can also be multiplexed to improve the utilization of the antennas. For example: the antenna 1 may be multiplexed as a diversity antenna of a wireless local area network.

The mobile communication module 350 may provide a solution including 2G/3G/4G/5G wireless communication applied to the electronic device 300. The mobile communication module 350 may include at least one filter, a switch, a power amplifier, a Low Noise Amplifier (LNA), and the like. The mobile communication module 350 may receive the electromagnetic wave from the antenna 1, filter, amplify, etc. the received electromagnetic wave, and transmit the filtered electromagnetic wave to the modem processor for demodulation. The mobile communication module 350 may also amplify the signal modulated by the modem processor, and convert the signal into electromagnetic wave through the antenna 1 to radiate the electromagnetic wave.

The modem processor may include a modulator and a demodulator. The modulator is used for modulating a low-frequency baseband signal to be transmitted into a medium-high frequency signal. The demodulator is used for demodulating the received electromagnetic wave signal into a low-frequency baseband signal. The demodulator then passes the demodulated low frequency baseband signal to a baseband processor for processing. The low frequency baseband signal is processed by the baseband processor and then transferred to the application processor. The application processor outputs sound signals through an audio device (not limited to the speaker 370A, the receiver 370B, etc.) or displays images or video through the display 394.

The wireless communication module 360 may provide solutions for wireless communication applied to the electronic device 300, including Wireless Local Area Networks (WLANs) (e.g., wireless fidelity (Wi-Fi) networks), bluetooth (bluetooth, BT), Global Navigation Satellite System (GNSS), Frequency Modulation (FM), Near Field Communication (NFC), Infrared (IR), and the like. The wireless communication module 360 may be one or more devices integrating at least one communication processing module. The wireless communication module 360 receives electromagnetic waves via the antenna 2, performs frequency modulation and filtering processing on electromagnetic wave signals, and transmits the processed signals to the processor 310. The wireless communication module 360 may also receive a signal to be transmitted from the processor 310, frequency-modulate and amplify the signal, and convert the signal into electromagnetic waves via the antenna 2 to radiate the electromagnetic waves.

In some embodiments, antenna 1 of electronic device 300 is coupled to mobile communication module 350 and antenna 2 is coupled to wireless communication module 360 such that electronic device 300 may communicate with networks and other devices via wireless communication techniques. The wireless communication technology may include global system for mobile communications (GSM), General Packet Radio Service (GPRS), Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA), time-division code division multiple access (TD-SCDMA), long term evolution (long term evolution, LTE), BT, GNSS, WLAN, NFC, FM, and/or IR technologies, among others. GNSS may include Global Positioning System (GPS), global navigation satellite system (GLONASS), beidou satellite navigation system (BDS), quasi-zenith satellite system (QZSS), and/or Satellite Based Augmentation System (SBAS).

The electronic device 300 implements display functions via the GPU, the display 394, and the application processor, among other things. The GPU is an image processing microprocessor coupled to a display 394 and an application processor. The GPU is used to perform mathematical and geometric calculations for graphics rendering. The processor 310 may include one or more GPUs that execute program instructions to generate or alter display information.

The display screen 394 is used to display images, video, and the like. The display screen 394 includes a display panel. The display panel may employ a Liquid Crystal Display (LCD), a light-emitting diode (LED), an organic light-emitting diode (OLED), an active-matrix organic light-emitting diode (active-matrix organic light-emitting diode, AMOLED), a flexible light-emitting diode (FLED), a miniature, a Micro-oeled, a quantum dot light-emitting diode (QLED), or the like.

Here, if the display screen 394 in the embodiment of the present application is integrated with a touch sensor, the display screen 394 may be referred to as a touch screen. The touch sensor may also be referred to as a "touch panel". That is, the display screen 394 may include a display panel and a touch panel. The touch sensor is used to detect a touch operation applied thereto or nearby. After the touch sensor detects a touch operation, a drive (e.g., a TP drive) of the core layer may be triggered to periodically scan touch parameters generated by the touch operation. Then, the driving of the kernel layer transmits the touch parameters to a related module of the upper layer, so that the related module determines a touch event corresponding to the touch parameters.

Additionally, the display screen 394 may provide visual output related to touch operations. In other embodiments, the touch sensor may be disposed on a surface of the electronic device 300 instead of being integrated into the display screen 394. At this time, the touch sensor and the display screen 394 may be located at different positions. In the embodiment of the present application, a specific process of a display method of a playing interface is described by taking an example in which a screen is a screen integrated with a touch sensor.

The electronic device 300 may implement a shooting function through the ISP, the camera 393, the video codec, the GPU, the display 394, the application processor, and the like. The ISP is used to process the data fed back by the camera 393. The camera 393 is used to capture still images or video. The digital signal processor is used for processing digital signals, and can process digital image signals and other digital signals. Video codecs are used to compress or decompress digital video. The electronic device 300 may support one or more video codecs. In this way, the electronic device 300 may play or record video in a variety of encoding formats, such as: moving Picture Experts Group (MPEG) 3, MPEG3, MPEG3, MPEG4, and the like.

The NPU is a neural-network (NN) computing processor that processes input information quickly by using a biological neural network structure, for example, by using a transfer mode between neurons of a human brain, and can also learn by itself continuously. The NPU can realize applications such as intelligent recognition of the electronic device 300, for example: image recognition, face recognition, speech recognition, text understanding, and the like.

The external memory interface 320 may be used to connect an external memory card, such as a Micro SD card, to extend the memory capability of the electronic device 300. The external memory card communicates with the processor 310 through the external memory interface 320 to implement a data storage function. For example, files such as music, video, etc. are saved in an external memory card. The internal memory 321 may be used to store computer-executable program code, which includes instructions. The processor 310 executes various functional applications of the electronic device 300 and data processing by executing instructions stored in the internal memory 321. For example, in the embodiment of the present application, the processor 310 may execute instructions stored in the internal memory 321, and the internal memory 321 may include a program storage area and a data storage area. The storage program area may store an operating system, an application program (such as a sound playing function, an image playing function, etc.) required by at least one function, and the like. The data storage area may store data (e.g., audio data, phone book, etc.) created during use of the electronic device 300, and the like. In addition, the internal memory 321 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, a Universal Flash Storage (UFS), and the like.

The electronic device 300 may implement audio functions via the audio module 370, the speaker 370A, the receiver 370B, the microphone 370C, the headphone interface 370D, and the application processor. Such as music playing, recording, etc.

The audio module 370 is used to convert digital audio information into an analog audio signal output and also to convert an analog audio input into a digital audio signal. The audio module 370 may also be used to encode and decode audio signals. The speaker 370A, also called a "horn", is used to convert the audio electrical signal into an acoustic signal. The receiver 370B, also called "earpiece", is used to convert the electrical audio signal into an acoustic signal. Microphone 370C, also known as a "microphone," is used to convert sound signals into electrical signals. The headphone interface 370D is used to connect wired headphones.

Keys 390 include a power-on key, a volume key, etc. The keys 390 may be mechanical keys. Or may be touch keys. The electronic device 300 may receive a key input, and generate a key signal input related to user setting and function control of the electronic device 300. The motor 391 may generate a vibration cue. The motor 391 may be used for both incoming call vibration prompting and touch vibration feedback. Indicator 392 may be an indicator light that may be used to indicate a state of charge, a change in charge, or a message, missed call, notification, etc. The SIM card interface 395 is for connecting a SIM card. The SIM card can be attached to and detached from the electronic device 300 by being inserted into and removed from the SIM card interface 395. The electronic device 300 may support 3 or N SIM card interfaces, N being a positive integer greater than 3. The SIM card interface 395 may support a Nano SIM card, a Micro SIM card, a SIM card, etc.

The software system of the mobile phone can adopt a layered architecture, an event-driven architecture, a micro-core architecture, a micro-service architecture or a cloud architecture. The embodiment of the invention exemplifies the software structure of the mobile phone by taking the Android system with a layered architecture as an example.

The layered architecture divides the software into several layers, each layer having a clear role and division of labor. And the layers communicate with each other through an interface. In some embodiments, the Android system is divided into four layers, namely an application layer, an application framework layer, an Android runtime and system library, and a kernel layer from top to bottom.

The application layer may include a series of application packages.

As shown in fig. 4, the application packages may include camera, mail, calendar, phone call, map, navigation, WLAN, bluetooth, music, mall, community club, etc. applications.

The application framework layer provides an Application Programming Interface (API) and a programming framework for the application program of the application layer. The application framework layer includes a number of predefined functions.

The application framework layer may include a window manager, a content provider, an activity manager, a resource manager, a notification manager, a view system, and the like, which is not limited in this embodiment.

Activity Manager (Activity Manager): for managing the lifecycle of each application. Applications typically run in the form of Activity in an operating system. For each Activity, there is an application record (activetyrecord) in the Activity manager corresponding to it, which records the state of the Activity of the application. The Activity manager can schedule Activity processes for the application using this Activity record as an identification.

Window manager (windowmanager service): graphical User Interface (GUI) resources for managing GUI resources used on a screen may specifically be used to: acquiring the size of a screen, creating and destroying a window, displaying and hiding the window, arranging the window, managing a focus, managing an input method, managing wallpaper and the like.

The system library and the kernel layer below the application framework layer may be referred to as an underlying system, and the underlying system includes an underlying display system for providing display services, for example, the underlying display system includes a display driver in the kernel layer and a surface manager in the system library.

The Android Runtime (Android Runtime) includes a core library and a virtual machine. The Android runtime is responsible for scheduling and managing an Android system. The core library comprises two parts: one part is a function which needs to be called by java language, and the other part is a core library of android. The application layer and the application framework layer run in a virtual machine. And executing java files of the application program layer and the application program framework layer into a binary file by the virtual machine. The virtual machine is used for performing the functions of object life cycle management, stack management, thread management, safety and exception management, garbage collection and the like.

The system library may include a plurality of functional modules. For example: surface managers (surface managers), Media Libraries (Media Libraries), three-dimensional graphics processing Libraries (e.g., OpenGL ES), 2D graphics engines (e.g., SGL), and the like.

The surface manager is used to manage the display subsystem and provide fusion of 2D and 3D layers for multiple applications.

The media library supports a variety of commonly used audio, video format playback and recording, and still image files, among others. The media library may support a variety of audio-video encoding formats, such as: MPEG4, H.264, MP3, AAC, AMR, JPG, PNG, etc.

OpenGL ES is used to implement three-dimensional graphics drawing, image rendering, compositing, and layer processing, among others.

SGL is a drawing engine for 2D drawing.

The kernel layer is a layer between hardware and software. The inner core layer at least comprises a display driver, a camera driver, an audio driver and a sensor driver.

In the embodiment of the present application, a server 500 shown in fig. 5 is taken as an example, and a structure of the server provided in the embodiment of the present application is described as an example. Fig. 5 is a schematic diagram of a hardware structure of a server according to an embodiment of the present application. The server 500 includes at least one processor 501, communication lines 502, memory 503, and at least one communication interface 504.

The processor 501 may be a general-purpose Central Processing Unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more ics for controlling the execution of programs in accordance with the present disclosure.

The communication link 502 may include a path for transmitting information between the aforementioned components.

Communication interface 504, using any transceiver or the like, may be used to communicate with other devices or communication networks, such as ethernet, Radio Access Network (RAN), Wireless Local Area Networks (WLAN), etc.

The memory 503 may be, but is not limited to, a read-only memory (ROM) or other type of static storage device that may store static information and instructions, a Random Access Memory (RAM) or other type of dynamic storage device that may store information and instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disk storage, optical disk storage (including compact disc, laser disc, optical disc, digital versatile disc, blu-ray disc, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. The memory may be self-contained and coupled to the processor via communication link 502. The memory may also be integral to the processor.

The memory 503 is used for storing computer-executable instructions for executing the present application, and is controlled by the processor 501 to execute. The processor 501 is configured to execute computer-executable instructions stored in the memory 503, so as to implement the communication method of multicast/broadcast service provided by the following embodiments of the present application.

Optionally, the computer-executable instructions in the embodiments of the present application may also be referred to as application program codes, which are not specifically limited in the embodiments of the present application.

In particular implementations, processor 501 may include one or more CPUs such as CPU0 and CPU1 in fig. 5 as one embodiment.

In particular implementations, server 500 may include multiple processors, such as processor 501 and processor 507 in FIG. 5, for example, as an embodiment. Each of these processors may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. A processor herein may refer to one or more devices, circuits, and/or processing cores that process data (e.g., computer program instructions).

Optionally, the server 500 may also include an output device 505 and an input device 506. An output device 505, which is in communication with the processor 501, may display information in a variety of ways. For example, the output device 505 may be a Liquid Crystal Display (LCD), a Light Emitting Diode (LED) display device, a Cathode Ray Tube (CRT) display device, a projector (projector), or the like. The input device 506 is in communication with the processor 501 and may receive user input in a variety of ways. For example, the input device 506 may be a mouse, a keyboard, a touch screen device, or a sensing device, among others.

The server 500 may be a general-purpose device or a dedicated device. In a specific implementation, the server 500 may be a desktop, a laptop, a web server, a Personal Digital Assistant (PDA), a mobile phone, a tablet, a wireless terminal device, an embedded device, or a device with a similar structure as in fig. 5. The embodiment of the present application does not limit the type of the server 500.

A cold start recommendation method provided in an embodiment of the present application is described below with reference to fig. 6 as an example, which specifically includes:

601. and determining the model of the terminal equipment used by the new user to access the first application.

The model of the terminal equipment used by the new user is the first model.

After the new user enters the first application, the first application guides the new user to log in the first application through the interactive interface, and after the new user logs in the first application, the first application displays a recommendation interface for the new user. Before the first application presents the recommendation interface for the new user, the first application needs to determine the model of the terminal device used by the new user.

Optionally, the first application may be an application program in a mobile phone, and may also be a website platform in a computer. The recommendation interface presented for the user may be used for recommending books, recommending social friends, recommending videos, recommending music, recommending restaurants, recommending goods, and so forth.

The first application determines the model of the terminal device used by the new user, and specifically may determine the terminal device information of the new user through an operating system of the terminal device used by the new user. Currently, most operating systems of terminal devices have an Application Program Interface (API) for a third party Application (Aapplication) to obtain some basic information of the terminal device. After the authorization of the user is obtained, the third party App can obtain the access authority of the API disclosed by the operating system of the terminal device, and then obtain the required information through the API.

In an android operating system, a third party App (i.e., a first application) can obtain a terminal device model and all App information (installation package names) of installation. In the iOS system, the third party App (namely the first application) can not directly acquire the App list installed on the user terminal equipment, but can indirectly acquire the information by inquiring the mode whether the terminal equipment installs a certain App. This approach requires the third party App (i.e., the first application) to have built in the id list of the commonly used iOS apps, which can be crawled from the web page version of App Store.

602. And acquiring a commodity access record of the first application.

The commodity access record of the first application can be acquired by the terminal equipment or the server. If the terminal device obtains the commodity access record, the specific obtaining mode can be that the terminal device sends a commodity access record obtaining instruction to a cloud server of the first application, and the cloud server of the first application responds to the commodity access record obtaining instruction and sends the commodity access record of the first application to the terminal device. The cloud server of the first application may also periodically send the commodity access record of the first application to all terminal devices installed with the first application. Or after the cloud server of the first application detects the new commodity access record of the first application, automatically sending the new commodity access record to all terminal devices provided with the first application. Or when the terminal device downloads the installation package of the first application, the installation package of the first application is automatically carried, and all commodity access records of the first application at the current moment are cut off. If the commodity access record is acquired by the server, the specific acquisition mode is that the server sends a commodity access record acquisition instruction to the cloud server of the first application, and the cloud server of the first application responds to the commodity access record acquisition instruction and sends the commodity access record of the first application to the server.

The first application may be a shopping-type application, such as a mall APP, a shopping APP, or the like.

The commodity access record can be obtained by a page point-burying technology. The page point burying technology can specifically acquire data in two modes of handwriting point burying and no point burying. The handwriting points are: after determining the data to be monitored (e.g., the user's clicking behavior, browsing behavior, etc.), the code is then written in the corresponding location. And when the third party App detects the preset monitoring data, directly reporting the detected data. The method of burying points by handwriting has great flexibility, but the workload is also great, and corresponding codes are inserted into each place needing to be monitored. The non-buried points refer to: the data is acquired by counting all events and reporting at regular time without needing a developer to write points by hand. Although this method of no buried point is less cumbersome than handwriting the buried point, it is necessary to filter out the required data at a later stage because all events are counted.

Generally, a Page burying point monitors data such as Page View (PV), number of independent access Users (UV), dwell time, and user interaction behavior. The PV refers to the page view volume or click volume of the website. The UV refers to the number of visitors distinguished according to IP addresses, and is also a UV if one IP address is repeatedly accessed for a period of time. The stay time specifically refers to the stay time of a user on a certain page of the third party App or one access (session). The user interaction behavior refers to all behaviors triggered by the user on the first application, such as browsing, clicking, collecting, sharing and the like.

The commodity access record refers to a record of the access condition of the user to commodities in the shopping application stored in the first application within a preset time length. The merchandise access records may include user behavior records and user device records.

The user behavior record refers to a record of the first interaction behavior of the user with the commodity. For example, the user behavior record may include a name of the user, a name of the item visited by the user, a first interaction behavior of the user with the item, and an occurrence time of the first interaction behavior. The first interaction behavior of the user with the commodity may include clicking into a commodity detail page, collecting the commodity, joining a shopping cart, clicking to purchase the commodity immediately, submitting a commodity purchase order, paying for the commodity order successfully, canceling the commodity collection, sharing the commodity, searching for the commodity, notifying the commodity arrival to confirm the success, consulting the commodity, commenting on the commodity, and clicking on the commodity in the browsing footprint. The user behavior records can directly reflect the real requirements of the user, and the commodities which the user is interested in can be determined by analyzing the first interaction behaviors of the user and the commodities. And then personalized recommendation is carried out for the user based on the interested commodities, so that the real requirements of the user can be better met.

The user device record may be a record of the terminal device used by the user when the user accesses the first application. For example, the user device record may include the name of the user, the model of the terminal device used by the user, and the time the device was running the first application.

If the user accessing the first application is a new user, the first application cannot provide data for reference, so that the commodity access records of other users of the same terminal equipment type are inquired according to the terminal equipment type used by the new user, and the recommendation data are provided for the new user according to the commodity access records of the other users.

Illustratively, the merchandise visit record is shown in Table 1 below.

TABLE 1

603. And determining a first commodity access record corresponding to the first model in the commodity access records.

The commodity access record of the first application may be multiple, and the number of the commodity access records is increased as the online time of the first application is longer. And selecting the commodity access record using the terminal equipment of the first model from the plurality of commodity access records to obtain a first commodity access record.

Illustratively, in conjunction with table 1, a first merchandise visit record is obtained, which is shown in table 2 below.

TABLE 2

604. And obtaining a first recommendation list according to the first commodity access record.

And scoring the commodities related to each commodity access record based on each commodity access record in the first commodity access record to obtain a scoring result of the commodities related to each commodity access record. And then accumulating the scoring results of the same commodity based on the commodity names to obtain the final scoring results of all commodities. And then sorting the commodities according to the scoring results of all the commodities, and generating a first recommendation list by combining the sorting results. And finally, recommending commodities for the new user according to the commodity information in the first recommendation list.

And scoring the commodities related to each commodity access record from two aspects, namely scoring according to the first interaction behavior of the user and the commodities in each commodity access record on one hand, and scoring according to the occurrence time of the first interaction behavior in each commodity access record on the other hand. And finally, multiplying the scoring results of the two aspects to obtain a final scoring result of the commodities related to each commodity access record.

Scoring according to the first interaction behavior of the user and the commodity in each commodity access record, specifically, comparing the first interaction behavior of the user and the commodity in each commodity access record with the behavior in the behavior scoring rule, and when it is determined that the first interaction behavior of the user and the commodity in the commodity access record is consistent with one behavior in the behavior scoring rule, taking the scoring result of the behavior specified in the behavior scoring rule as the scoring result of the first interaction behavior of the user and the commodity in the current commodity access record. And if the first interactive behavior of the user and the commodity in the commodity access record does not have the corresponding behavior in the behavior scoring rule, scoring the first interactive behavior of the user and the commodity as 0. Exemplary, behavior scoring rules are as shown in table 3:

TABLE 3

Based on tables 2 and 3, the first interaction behavior of the user with the commodity is scored, and the scoring results are shown in table 4 below:

TABLE 4

And scoring according to the occurrence time of the first interactive behavior in each commodity access record, wherein the scoring result is used for determining the occurrence time of the first interactive behavior in the commodity access record based on a time decay function. Illustratively, the time decay function is shown by the following equation:

wherein t0 is the initial temperature, t1 is the cooling coefficient, t0 and t1 are all fixed coefficient values, and the interval time is the interval between the occurrence time of the first interaction and the time to be recommended. Exemplarily, t0=1, t1= 0.2.

And substituting the first interaction behavior occurrence time in each commodity access record into the time attenuation function to calculate F (t), and then normalizing the result of the F (t) to obtain a numerical value between [0 and 1 ].

It is well known that the purchasing needs of a user may change over time. The closer to the recommendation time, the more referential is the first interaction of the user with the item that occurs. If the time for recommending the goods for the new user is 10 months and 19. Referring to table 2, the king is No. 3/10, the detail page of the article a is entered, and the nay is No. 10/18, and the article C is collected. By comparison, 10 month 18 is closer to the recommended time 10 month 19, and the score for 10 month 18 is higher than the score for 3 month 10.

After the scoring result of the first interaction behavior between the user and the commodity and the scoring result of the occurrence time of the first interaction behavior are determined, the scoring result of the commodity related to each commodity access record meets the following formula:

A=B*C

wherein, A is the scoring result of the commodities related in each commodity access record, B is the scoring result of the first interaction behavior in each commodity access record, and C is the scoring result of the occurrence time of the first interaction behavior in each commodity access record.

And finally, multiplying the first interactive behavior scoring result in the table 4 by the first interactive behavior occurrence time scoring result calculated based on the time attenuation function to obtain the scoring value of the corresponding commodity in each commodity access record. The score of the corresponding commodity in each commodity access record is specifically shown in table 5 below:

TABLE 5

And after the scoring result of the commodity corresponding to each commodity access record is obtained, accumulating the scoring results of the same commodities to obtain the addition scoring result value corresponding to each commodity. With reference to table 6, the commodity corresponding to the first commodity access record is commodity a, the commodity corresponding to the third commodity access record is also commodity a, the commodity corresponding to the nth commodity access record is commodity C, and then the scoring results that are all commodity a are accumulated, that is, the scoring result obtained from the first commodity access record is 0.025 and the scoring result obtained from the third commodity access record is 0.04, so that the scoring result corresponding to commodity a is 0.065. The scoring results for the other goods involved are then calculated in turn.

And finally, generating a recommendation list according to the scoring results of all the commodities, wherein the recommendation list comprises commodity information and the scoring results corresponding to the commodities, and the commodity information can be commodity names and recommendation serial numbers. Illustratively, the recommendation list is shown in table 6:

TABLE 6

After the recommendation list is obtained, the first N commodities in the recommendation list may be selected according to the recommendation requirement to form a first recommendation list. For example, the recommendation requirement may be to select the top 10 items in the recommendation list as recommended items. Or, the commodity with the scoring result exceeding the threshold value is taken as the recommended commodity. The specific recommended requirements can be adjusted according to actual requirements, and the application is not limited to this. After the first recommendation list is obtained, the new user can be recommended according to the commodity information in the first recommendation list and the scoring result corresponding to the commodity, or the first recommendation list and other recommendation lists are combined to be recommended by the new user.

Specifically, 603-604 may be executed by the terminal device or may be executed by the server. If 603-604 are performed by the server, before 603, the method further comprises: the server receives a recommendation list acquisition request from the terminal equipment, wherein the recommendation list acquisition request comprises the model of the terminal equipment used by the new user, and the model of the terminal equipment is a first model. After the first recommendation list is obtained through 604, the server sends the first recommendation list to the terminal device, and the terminal device recommends commodities for the new user based on the commodity information in the first recommendation list.

Another cold start recommendation method provided in the embodiment of the present application is described below with reference to fig. 7 as an example, where the method further includes:

701. the model of the terminal device used by the new user to access the first application is determined.

See 601 for details.

702. And acquiring a block access record of the second application.

The chunk access record of the second application may be a terminal device acquisition or a server acquisition. If the terminal device obtains the version information, the specific obtaining mode can be that the terminal device sends a version access record obtaining instruction to a cloud server of the second application, and the cloud server of the second application responds to the version access record obtaining instruction and sends the version access record of the second application to the terminal device; the cloud server of the second application can also periodically send the layout access records of the second application to all terminal devices installed with the second application; or after the cloud server of the second application detects a new layout access record of the second application, automatically sending the new layout access record to all terminal devices installed with the second application; or when the terminal device downloads the installation package of the second application, the installation package of the second application is automatically carried, and all the layout block access records on the second application are recorded at the current moment. If the server is used for obtaining, the specific obtaining mode is that the server sends a request for obtaining the version block access record instruction to a cloud server of the second application, and the cloud server of the second application responds to the request for obtaining the version block access record instruction and sends the version block access record of the second application to the server.

The second application may be a community-like application, such as a community club APP.

The community club APP can provide a plurality of interest blocks for users to interact with. The plurality of interest blocks may include: the system comprises a mobile phone layout, a smart life layout, a User Interface Design (UI) layout, a cloud service layout, an interest block layout, a game layout and an offline club layout. The posts below the mobile phone edition block are function introductions of different mobile phone models; posts below the smart life section are introduced to functions of intelligent equipment, and the intelligent equipment can be a notebook, a tablet, a headset sound box, a smart screen, a router, intelligent wearable equipment and the like; posts below the UI layout are functional introductions related to the UI; posts below the cloud service block are introduction of functions related to the cloud service; the posts below the interest block are introduced for different interests, such as food, travel, sports, digital, theme, games, photography, and slow life. Posts below the game pieces are game introduction; the posts below the offline club section are introduction to the offline activity.

The block access record can also be obtained by a page burial point technology. For the page buried point, the description at 402 can be referred to specifically, and will not be described herein again.

The layout access record refers to a record of access conditions of users to each layout in community-class applications, which are stored by the second application, within a preset time length. The chunk access records may include chunk behavior records and chunk device records.

The section behavior record refers to a record of second interaction behavior of the user with the section. For example, the slab behavior record may include a name of the user, a name of the slab browsed by the user, a name of a post under the name of the slab browsed by the user, a second interaction behavior of the user with the slab, and a record of a time at which the second interaction behavior occurs. The second interaction behavior of the user with the layout block may include browsing of a post under the layout block by the user, comment of the user under the post, reply to the comment under the post by the user, sharing of the post by the user, favorite post by the user, approval of the post by the user, detailed information of the user who clicks the comment post by the user, post issued by the user who clicks the comment post, and appreciation of the post by the user.

The block device record may be a record of a terminal device used by the user when the user accesses the second application. For example, the chunk device record may include a name of the user, a model of the terminal device used by the user, and a time that the device runs the second application. The specific content of the block access record may refer to the content in table 1, which is not described in detail herein.

And, the chunk access record may include only the chunk access record of the login user of the second application, and may further include the chunk access record of the non-login user of the second application. The login user refers to a user who has registered an account in accessing the second application and has logged in the account for access. The non-login user refers to a user who accesses the second application without a login account. When the user accesses the second application, the second application reminds the user to log in the second application through a popup window, if the user refuses to log in the second application, a non-login user is marked and numbered, and then all second interactive behaviors of the non-login user are recorded until the non-login user logs out of the second application.

703. And determining a first version access record corresponding to the first type in the version access records.

The second application may have a plurality of the section access records, and the section access records of the second application are more numerous as the online time of the second application is longer. And selecting the block access record using the equipment of the first model from the plurality of block access records to obtain the first block access record. The specific content of the first version access record may refer to the content in table 2, which is not described herein again.

704. And determining the user preference layout and a user group corresponding to the user preference layout based on the first layout access record.

And scoring the layout blocks related in each layout block access record based on the second interaction behavior between the user of each layout block access record in the first layout block access record and the layout blocks and the occurrence time of the second interaction behavior, so as to obtain the layout blocks preferred by the user.

Illustratively, referring to 604, the rating of the layout pieces is divided into two aspects, namely, the rating according to the second interaction behavior of the user and the layout pieces in each layout piece access record on one hand, and the rating according to the occurrence time of the second interaction behavior in each layout piece access record on the other hand. And finally, multiplying the scoring results of the two aspects to obtain the final scoring result of the layout block related in each layout block access record.

The behavior that the user browses the posts in the layout block in the second interactive behavior is that the corresponding scoring result is 0.25; the behavior of the comment of the user under the post in the block corresponds to a scoring result of 0.3; the user shares the behavior of the posts in the block, and the corresponding scoring result is 0.4; and the behavior of collecting the posts in the layout block by the user is 0.4 in the corresponding rating result, the behavior of praise of the posts in the layout block by the user is 0.6 in the corresponding rating result. The scoring result of the second interaction behavior between other users and the layout can be set by combining with the actual requirement, which is not limited in the present application.

The occurrence time score of the second interactive action may be calculated according to the time decay function in 604.

After the scoring result of the second interactive behavior between the user and the layout and the scoring result of the occurrence time of the second interactive behavior are determined, the scoring result of the layout related to each layout access record meets the following formula:

A1=B1*C1

the A1 is a scoring result of the layout involved in each layout access record, the B1 is a scoring result of the second interaction behavior in each layout access record, and the C1 is a scoring result of the occurrence time of the second interaction behavior in each layout access record.

And sequentially calculating the scoring results of the plates related in all the plate access records, and accumulating the scoring results of the same plate based on the plate names so as to obtain the final scoring of all the related plates. And then sorting the layout blocks according to the grading results of all the layout blocks, and taking the layout block 3 at the top of the sorting as a user preference layout block of the same model according to the sorting result.

And screening out a second edition access record only containing the user preference edition from the first edition access record based on the name of the user preference edition. In connection with 702, each of the chunk access records includes a name of the user. And then recording the name of the user in each layout access record in the second layout access, and deleting the repeated names to obtain a user group corresponding to the user preference layout.

Firstly, the same model is used for screening for the first time, and the version access records of the equipment with the model consistent with that of the terminal equipment used by the new user are screened out. And then, scoring is carried out on the layout blocks in the screened layout block access records, and then the user preference layout blocks are screened out. And finally, based on the preferred layout blocks of the user, personalized recommendation is made for the user, so that the user requirements can be better met.

705. And determining the goods preferred by each user in the user group from the goods access records based on the name of each user in the user group, and generating a second recommendation list based on the goods preferred by the user.

According to the name of each user in the user group, the commodity access record of each user in the user group is indexed from the commodity access record, then according to the commodity access record of each user, the commodities related to each commodity access record are scored, the specific scoring process can refer to 604, and then the commodities preferred by the users in the user group are obtained by sequencing based on the scoring results. And finally, generating a second recommendation list according to the commodity information of the commodities preferred by the user and the scoring result corresponding to the commodities, wherein the obtained second recommendation list can be directly used for recommending a new user, or the second recommendation list and other recommendation lists are combined to be recommended by the new user.

The 703 processing 705 may be executed by the terminal device or may be executed by the server. If 703-705 is performed by the server, before 703, the method further comprises: the server receives a recommendation list acquisition request from the terminal equipment, wherein the recommendation list acquisition request comprises the model of the terminal equipment used by the new user, and the model of the terminal equipment is a first model. After the second recommendation list is obtained through 705, the server sends the second recommendation list to the terminal device, and the terminal device may recommend a commodity for the new user based on the commodity information in the second recommendation list.

Another cold start recommendation method provided in the embodiment of the present application is described below with reference to fig. 8 as an example, where the method further includes:

801. the model of the terminal device used by the new user to access the first application is determined.

See 601 for details.

802. And acquiring a block access record of the second application.

See 702 for details.

803. And determining a first version access record corresponding to the first model in the version access records.

See 703 for details.

804. And determining a first type word corresponding to the post related to each block access record based on the posts related to each block access record in the first block access records, and creating a mapping relation for the posts and the first type word.

Wherein the first type of word is a word related to the product name involved in the post.

The first version access record is a version access record using a terminal device of a first model. By viewing the first block access record, the posts involved in each block access record in the first block access record can be known. And looking up the posts to obtain words related to the product name in each post, wherein the words related to the product name are the first type words. And establishing a mapping relation between the post and the first type word corresponding to the post. For example, if the post involved in the first version of the access record is a post under the mobile phone version, the word related to the product name (i.e., the first type word) in the post may be the mobile phone name.

805. And determining the posts preferred by the user based on the first version block access records.

In combination with 702, the first layout block access record includes a second interaction behavior of the user with the layout block and occurrence time of the second interaction behavior, where the second interaction behavior of the user with the layout block is generally an interaction behavior of the user with a post under the layout block, that is, the second interaction behavior of the user with the layout block and the occurrence time of the second interaction behavior may be considered as the second interaction behavior of the user with the post and the occurrence time of the second interaction behavior. And then scoring the occurrence time of the second interaction behavior and the occurrence time of the second interaction behavior of the user and the post in each block access record in the first block access record to determine the scoring result of the posts related in each block access record, and finally accumulating the scoring results of the same post to obtain the final scoring value of each post. The scoring 704 may be referred to for the second interaction behavior of the user with the post and the occurrence time of the second interaction behavior, which is not described herein again.

And then sorting the posts according to the scoring result of the posts, and screening out the user preference posts of the same model according to the sorting result. Illustratively, the top 10 posts in the ranking result may be selected as user preference posts.

806. And determining a first type word corresponding to the post preferred by the user according to the mapping relation.

And combining 804, after the posts preferred by the user are determined, obtaining the first type words corresponding to the posts preferred by the user according to the mapping relation.

807. And determining the commodity information corresponding to the first type of words by using a fuzzy matching algorithm, and generating a fourth recommendation list according to the commodity information corresponding to the first type of words.

Illustratively, fuzzy matching may be achieved through fuzzy Wuzzy. FuzzyWuzzy is a simple easy-to-use fuzzy string matching toolkit. It calculates the difference between two sequences according to the edit Distance (Levenshtein Distance) algorithm. The edit distance refers to the minimum number of edit operations required to change from one character string to another. Editing operations may include replacing one character with another, inserting one character, and deleting one character. Generally, the smaller the edit distance, the greater the similarity of the two sequences.

Fuzzy matching can be carried out on the FuzzyWuzzy through any one of four functions of simple matching (Ratio), incomplete matching (Partial Ratio), neglected sequential matching (Token Sort Ratio) and deduplication subset matching (Token Set Ratio), and therefore the commodities can be matched according to the first type of words.

For example, the user preferred posts are post 1, post 2, and post 3 described below. The title of post 1 is: [ photography tutorial ] how a nice-looking night scene of INS green orange wind is captured, trade name A. The title of post 2 is: the flagship machine is exactly what, and the commodity name A tells you the answer. The title of post 3 is: name of May product B night scene street racket.

The first type word corresponding to post 1 is "commodity name a", the first type word corresponding to post 2 is "commodity name a", and the first type word corresponding to post 3 is "commodity name B". And then, the commodity name A corresponding to the post 1, the commodity name A corresponding to the post 2 and the commodity name B corresponding to the post 3 are sequentially associated to a vertical domain product directory tree through the fuzzy Wuzzy. Or after the first type words of all posts are determined, the first type words of all posts are related to the names of the product directory trees of the vertical domains through the fuzzy Wuzzy. For example, if "product name a" is "blazed Magic4 Pro" and product B is "V30", then "product name a" can be matched to product "blazed Magic4 Pro" and "product name B" can be matched to "blazed V30" based on FuzzyWuzzy.

The vertical domain is a small domain vertically subdivided under a large domain. Vertical refers to longitudinal extension, not lateral extension, subdivision is in vertical industry plates, and then main service depth development is selected. Taking a mobile phone as an example, the mobile phone is a vertical field and can be subdivided into a mobile phone, a shoal mobile phone, a rice mobile phone, a fruit mobile phone, and the like.

The product catalog tree is used for representing the hierarchical relation of all products under a certain brand, so that the development of all products under the brand can be clearly described, and the advantages and the disadvantages of the products are further shown. The hierarchical relationship of the product can be determined according to the performance of the parts of the product.

After the commodity information is matched through the fuzzy Wuzzy, the matched commodity information is sorted according to the scoring result of the posts preferred by the user, and then a fourth recommendation list is generated. Subsequently, recommending the new user according to the commodity information in the fourth recommendation list and the scoring result corresponding to the commodity; or combine the fourth recommendation list with other recommendation lists as new user recommendations.

803, 807 can be executed by the terminal device or the server. If 803-807 are performed by a server, before 803, the method further comprises: the server receives a recommendation list acquisition request from the terminal equipment, wherein the recommendation list acquisition request comprises the model of the terminal equipment used by the new user, and the model of the terminal equipment is a first model. After the fourth recommendation list is obtained through 807, the server sends the fourth recommendation list to the terminal device, and the terminal device recommends commodities for the new user based on the commodity information in the fourth recommendation list.

Another cold start recommendation method provided in the embodiment of the present application is described below with reference to fig. 9 as an example, where the method further includes:

901. the model of the terminal device used by the new user to access the first application is determined.

See 601 for details.

902. And acquiring a block access record of the second application.

See 702 for details.

903. And determining a first version access record corresponding to the first type in the version access records.

See 703 for details.

904. And determining a second type word corresponding to the post related to each block access record based on the posts related to each block access record in the first block access record.

Wherein the second type of word is a word related to the product characteristic involved in the post.

The first version access record is a version access record using a terminal device of a first model. By viewing the first block access record, the posts involved in each block access record in the first block access record can be known. And looking up the posts to obtain words related to the product characteristics in each post, wherein the words related to the product characteristics are the second type words. For example, if the post involved in the first version of the access record is a post under the mobile phone version, the word (i.e., the second type word) related to the product characteristic in the post may be a chip, a camera, a memory, a battery, or the like.

905. And sequencing according to the frequency of all the second type words to generate a first sequencing result, and associating the commodity selling points based on the second type words in the first sequencing result to generate a fifth recommendation list.

And the fifth recommendation list comprises commodity information corresponding to the commodity selling points.

And sequencing all the second type words according to the occurrence frequency of the second type words to obtain a first sequencing result. Obtaining a first ranking result by counting the frequency of the second type word in the posts, where the first ranking result may be: camera > chip > battery.

The second type of word may be determined according to a product dictionary or a trained model. Typically, the product dictionary includes product and product characteristics, and the second type of word is derived after the product introduced by the post is determined. Or inputting the post into the trained model, and outputting the second type word corresponding to the post by the model. The training process of the model is as follows: and acquiring a plurality of posts and second type words corresponding to the posts, wherein the second type words of each post are marked artificially. And finally, correcting the initial model through a loss function, so that the deviation between the output second type words and the artificially labeled second type words meets a threshold value, and thus the trained model is obtained.

The terminal device can store a commodity selling point database, and the commodity selling point database comprises all commodities in a shopping mall and commodity selling points of each commodity. And carrying out fuzzy matching on the commodity selling points in the commodity selling point database and each second type word in the first sequencing result to obtain the commodity selling point corresponding to each second type word. A second type of word may be matched to the point of sale of the plurality of items. And then, according to the commodity selling point performance of each commodity in the plurality of commodities, sequencing the commodities to obtain an initial commodity sequence corresponding to the second type word.

In combination 904, the first ranking result has a plurality of second type words, each second type word is matched with a commodity selling point, and one second type word can be matched with a plurality of commodity selling points. And then sorting the matched commodity selling points according to the performance of the commodity selling points from high to low. Because the commodity selling points can be directly corresponding to the commodities, the sequencing result of the commodity selling points can be converted into the sequencing result of the commodities. A second type of word may correspond to a product ordering result.

Since there are multiple words of the second type in the first ranking result, multiple product ranking results can be obtained. After the plurality of commodity sorting results are obtained, if the plurality of commodity sorting results are inconsistent, the plurality of commodity sorting results are adjusted by taking the sorting result of the second type word in the first sorting result as the main result to obtain a final commodity sorting result, and the final commodity sorting result is the fifth recommendation list.

Illustratively, the item selling points for item one are: the camera comprises a four-curved-surface screen, 5000 ten thousand main cameras, 5000 ten thousand ultra wide-angle lenses, 6400 ten thousand pixel periscopic telephoto lenses, a 4800mAh battery, a 66W super fast charger and a brand new generation CellTour 8 mobile platform chip. The commodity selling points of the second commodity are as follows: front and back symmetric double curved screens, 1 hundred million pixel camera +5000 million ultra wide-angle lens, 4800mAh battery, 66W super fast charging, multi-lens video space-isolated lens change, high-pass celling 778G chip and the like. The selling points of the third commodity are as follows: the device comprises a hyperbolic screen, 5000 ten thousand main cameras, 800 ten thousand ultra wide-angle lenses, a 4000mAh battery, 66W ultra fast charge and a breguet 1000+ chip.

The aforementioned first ordering results in camera > chip > battery. And associating the second type words in the first sequencing result with the commodity selling points in the first commodity, the second commodity and the third commodity. For the cameras in the first ranking result, the item selling points associated with item one are: 5000 ten thousand main shots, 5000 ten thousand super wide-angle lenses and 6400 ten thousand pixel periscopic telephoto lenses; the item selling points associated with item two are: 1 hundred million pixel cameras +5000 million ultra wide angle lenses; the item selling points associated with item three are: 5000 ten thousand main shots and 800 ten thousand ultra wide-angle lenses. By combining the commodity selling point performance of the commodities, the commodity sequence corresponding to the camera is as follows: commodity two > commodity one > commodity three.

For chips in the first ordering result, the commodity selling point associated with commodity one is: a brand new generation of Xiaolong 8 mobile platform chips; the item selling points associated with item two are: a high-pass Xiaolong chamber 778G chip; the item selling points associated with item three are: breguet 1000+ chip. By combining the commodity selling point performance of the commodities, the commodity sequence corresponding to the chip is as follows: the first commodity, the second commodity and the third commodity.

For the batteries in the first ordering result, the commodity selling points associated with commodity one are: 4800mAh battery +66W super quick charge; the item selling points associated with item two are: 4800mAh battery +66W super quick charge; the item selling points associated with item three are: the 4000mAh battery +66W super quick charge. By combining the commodity selling point performance of the commodities, the commodity sequence corresponding to the battery is as follows: commodity one = commodity two > commodity three.

In combination with the above, there are differences in the commodity orders associated with the cameras, chips and batteries. In this case, the ranking of each second type word in the first ranking result is taken as a main ranking, the commodity ranking associated with each second type word is taken as an auxiliary ranking, the plurality of commodity ranking results are adjusted to be a commodity ranking result, so as to obtain a fifth recommendation list, and the commodity ranking in the fifth recommendation list is as follows: commodity two > Commodity one > Commodity three.

And processing the posts related to each block access record in the first block access records to obtain a plurality of second type words. And then, sequencing according to the occurrence frequency of the second type words to obtain a first sequencing result. And then, carrying out fuzzy matching on each second type word in the first sequencing result and the commodity selling points in the commodity selling point database, associating each second type word with the commodity selling points of the commodities, and determining the commodity sequencing corresponding to each second type word according to the performance of the commodity selling points after associating. And finally, generating a final fifth recommendation list by taking the ordering of the second type words in the first ordering result as a main part and taking the commodity ordering corresponding to each second type word as an auxiliary part. And subsequently recommending commodities for the new user according to the commodity information in the fifth recommendation list, or combining the fifth recommendation list with other recommendation lists to recommend the new user.

903-905 may be executed by a terminal device or a server. If 903-905 is performed by a server, before 903, the method further includes: the server receives a recommendation list acquisition request from the terminal equipment, wherein the recommendation list acquisition request comprises the model of the terminal equipment used by the new user, and the model of the terminal equipment is a first model. After the fifth recommendation list is obtained through 905, the server sends the fifth recommendation list to the terminal device, and the terminal device may recommend a commodity for the new user based on the commodity information in the fifth recommendation list.

Another cold start recommendation method provided in the embodiment of the present application is described below with reference to fig. 10 as an example, where the method further includes:

1001. the model of the terminal device used by the new user to access the first application is determined.

See 601 for details.

1002. And acquiring a commodity access record of the first application.

See 602 for details.

1003. And distinguishing the commodity access records of the first application based on the model of the terminal equipment, and obtaining a user behavior sequence corresponding to each model based on the distinguished commodity access records of the first application.

And distinguishing the commodity access records of the first application according to the equipment models in the user equipment records. After the differentiation, a user behavior sequence P, P = [ P1, P2, P3, … ] is established for each device model. The user behavior sequence P may be determined according to the commodity access record of the first application corresponding to the device model within a preset time period. Wherein, p1, p2 and p3 represent the product granularity. The product granularity may be for all product pools, as well as for each level of product in each product pool. For example, in the cell phone product pool, p1, p2 and p3 can be any one of a folder cell phone, a stand cell phone, a slide cell phone, a rotary cell phone, etc.

1004. A similar device model of the first model is determined.

Determining similar device models for any device model may be determined by the similarity between the user behavior sequence P of the device and the user behavior sequences P of the other devices. Specifically, the determination may be performed by calculating cosine similarity between the user behavior sequence P of the device and the user behavior sequence P of another device.

For example, if the product of the similar device model to be determined is a mobile phone, similarity calculation may be performed between mobile phones of other models in the mall and the mobile phone of the model, so as to obtain a similarity calculation result. After the calculation, a plurality of similarity results can be obtained, and the device model with the similarity result ranking 3 can be generally selected as the similar device model.

By combining 601, the model of the device used by the new user who logs in the first application is the first model, and through similarity calculation, the similar device model of the first model can be obtained, and the similar device models of the first model can include multiple similar device models, which can be specifically selected according to actual situations, and the present application does not limit this.

1005. And determining a second commodity access record corresponding to the similar equipment model, and obtaining a third recommendation list according to the second commodity access record.

And the third recommendation list comprises commodity information and a scoring result corresponding to the commodity.

And distinguishing the commodity access records of the first application according to the equipment models to obtain second commodity access records corresponding to similar equipment models. And scoring the commodities related to the second commodity access record to obtain a scoring result of the commodities corresponding to the similar equipment models. Specific scoring processes may be referenced 604. And then sorting the commodities according to the grading height of the commodities. In combination with 1004, if 3 similar device models are selected, 3 sorting results can be obtained.

And based on the similarity calculation result of each similar equipment model and the first model, integrally sorting the obtained 3 sorting results from high to low according to the size of the similarity calculation result to obtain a third recommendation list. And subsequently recommending commodities for the new user according to the commodity information in the third recommendation list and the scoring result corresponding to the commodities, or combining the third recommendation list and other recommendation lists to recommend the new user.

1003, 1005 may be executed by the terminal device, or may be executed by the server. If 1003-: the server receives a recommendation list acquisition request from the terminal equipment, wherein the recommendation list acquisition request comprises the model of the terminal equipment used by the new user, and the model of the terminal equipment is a first model. After the third recommendation list is obtained through 1005, the server sends the third recommendation list to the terminal device, and the terminal device recommends commodities for the new user based on the commodity information in the third recommendation list.

By mining the commodities with similar equipment models and user preferences, the dilemma without effective data can be broken through from a brand new aspect, and interested commodities are recommended for the user, so that the cold start problem is solved.

Another cold start recommendation method provided in the embodiment of the present application is described below with reference to fig. 11 as an example, where the method includes 4 parts, which are respectively the first part: a first recommendation list generation method; a second part: generating methods of a second recommendation list, a fourth recommendation list and a fifth recommendation list; and a third part: a third recommendation list generation method and a fourth part: and (3) fusion process.

Wherein the first part: the first recommendation list generation method specifically includes: when a new user logs in the first application, the commodity access record of the first application can be obtained through the cloud server of the first application. Then, the commodity access records are distinguished according to the model of the terminal device used by the new user, a first commodity access record corresponding to the model (first model) of the terminal device used by the new user is obtained, and finally, a first recommendation list is obtained according to the first commodity access record. The detailed generation process of the first recommendation list is shown in 601-604, and will not be described in detail here.

A second part: the generation methods of the second recommendation list, the fourth recommendation list and the fifth recommendation list specifically include a generation method of the second recommendation list, a generation method of the fourth recommendation list and a generation method of the fifth recommendation list.

The generation method of the second recommendation list comprises the following steps: the method comprises the steps that a cloud server of a second application is used for obtaining a version access record of the second application, then the version access record is distinguished according to the model of a terminal device used by a new user, a first version access record corresponding to the model (a first model) of the terminal device used by the new user is obtained, and then a user preference version and a user group corresponding to the user preference version are obtained according to the first version access record. For example, the user preference section is a section a, a section B, and a section C. The A edition block corresponds to a user group a, the B edition block corresponds to a user group B, and the C edition block corresponds to a user group C. And determining the commodities preferred by the users in the user group according to the names of the users in the user group and the commodity access records, and finally generating a second recommendation list according to the commodities preferred by the users. The detailed generation process of the second recommendation list is shown in 701-705, and detailed description is omitted here.

The generation method of the fourth recommendation list comprises the following steps: the method comprises the steps of firstly obtaining a version access record of a second application through a cloud server of the second application, secondly distinguishing the version access record according to the model of a terminal device used by a new user to obtain a first version access record corresponding to the model (a first model) of the terminal device used by the new user, then respectively determining a first type word corresponding to a post related to each version access record according to the post related to each version access record in the first version access record, and establishing a mapping relation between the post and the first type word. The user preferred posts are then determined based on the first version of the access record. And determining a first type word corresponding to the post preferred by the user according to the mapping relation, finally determining commodity information corresponding to the first type word by using a fuzzy matching algorithm, and generating a fourth recommendation list according to the commodity information corresponding to the first type word. The detailed generation process of the fourth recommendation list is shown in 801-.

The generation method of the fifth recommendation list comprises the following steps: the method comprises the steps of obtaining a version access record of a second application through a cloud server of the second application, distinguishing the version access record according to the model of a terminal device used by a new user to obtain a first version access record corresponding to the model (a first model) of the terminal device used by the new user, determining a second type word according to the first version access record, associating a selling point according to the second type word, and generating a fifth recommendation list according to the selling point. The detailed generation process of the fifth recommendation list is shown in 901-905, and will not be described in detail here.

And a third part: the generation method of the third recommendation list specifically comprises the following steps: and acquiring a commodity access record of the first application through a cloud server of the first application. Then, based on the model of the terminal device, the commodity access records of the first application are distinguished, a user behavior sequence corresponding to each model is obtained based on the distinguished commodity access records of the first application, then the model of the similar device of the first model is determined, then the second commodity access record corresponding to the model of the similar device is determined, and a third recommendation list is obtained according to the second commodity access record. The detailed generation process of the third recommendation list is shown in 1001-1005 and will not be described in detail here.

The fourth part: and (3) fusion process. The method comprises the following steps: after the first recommendation list, the second recommendation list, the third recommendation list, the fourth recommendation list and the fifth recommendation list are obtained, the commodity information in the first recommendation list, the second recommendation list, the third recommendation list, the fourth recommendation list and the fifth recommendation list is subjected to fusion processing, and a final recommendation list is obtained.

For example, the fusion processing may be to select the commodity information corresponding to part of the commodities from the first recommendation list, the second recommendation list, the third recommendation list, the fourth recommendation list and the fifth recommendation list according to a set ratio. The selected items are then sorted in the order that the first recommendation list is first, followed by the second, third, fourth, and fifth recommendation lists. The set ratio may be 3:2:1:1: 1.

And deleting the repeated commodities after sorting, and supplementing the commodities from the corresponding recommendation list in order after deleting. Illustratively, the commodity corresponding to the commodity information in the third recommendation list is found to be consistent with the commodity corresponding to the commodity information in the first recommendation list, the commodity information in the third recommendation list is deleted, and then the next commodity information of the deleted commodity information in the third recommendation list is filled in the vacant position to generate the final recommendation list. The quantity of the commodities in the final recommendation list is related to the transmission parameter value of the front end. The front end transmits the number of the commodities needing to be recommended and displayed to the rear end in a recNum form, and after a final recommendation list is obtained, the rear end sequentially intercepts the recNum commodities and returns the commodities to be displayed to a user.

Or after the first recommendation list, the second recommendation list, the third recommendation list, the fourth recommendation list and the fifth recommendation list are obtained, the fusion processing may be to arbitrarily select two or more lists from the first recommendation list, the second recommendation list, the third recommendation list, the fourth recommendation list and the fifth recommendation list to perform sorting and deduplication of the commodity information, so as to generate a final recommendation list. For example, the product information in the first recommendation list, the second recommendation list, the third recommendation list, and the fourth recommendation list may be deduplicated and sorted to generate a final recommendation list, the product information in the first recommendation list, the second recommendation list, the third recommendation list, and the fifth recommendation list may be deduplicated and sorted to generate a final recommendation list, or the product information in the first recommendation list, the second recommendation list, and the third recommendation list may be deduplicated and sorted to generate a final recommendation list as shown in fig. 12.

In order to implement the functions, the electronic device includes a hardware structure and/or a software module for performing each function. Those of skill in the art will readily appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as hardware or combinations of hardware and computer software. Whether a function is performed as hardware or computer software drives hardware depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

In the embodiment of the present application, the electronic device may be divided into the functional modules according to the method example, for example, each functional module may be divided corresponding to each function, or two or more functions may be integrated into one processing module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. It should be noted that, in the embodiment of the present application, the division of the module is schematic, and is only one logic function division, and there may be another division manner in actual implementation.

Other embodiments of the present application provide an electronic device, which may include: a communication module, a memory, and one or more processors. The communication module, the memory and the processor are coupled. The memory is for storing computer program code comprising computer instructions.

Another embodiment of the present application provides a chip system, as shown in fig. 13, which includes at least one processor 1301 and at least one interface circuit 1302. The processor 1301 and the interface circuit 1302 may be interconnected by wires. For example, the interface circuit 1302 may be used to receive signals from other devices. Also for example, the interface circuit 1302 may be used to transmit signals to other devices, such as the processor 1301.

For example, the interface circuitry 1302 may read instructions stored in a memory in the device and send the instructions to the processor 1301. The instructions, when executed by the processor 1301, may cause the electronic device to perform various ones of the embodiments described above. Of course, the chip system may further include other discrete devices, which is not specifically limited in this embodiment of the present application.

Embodiments of the present application also provide a computer-readable storage medium, which includes computer instructions, when the computer instructions are executed on an electronic device, causing the electronic device to perform various functions or operations performed by the electronic device (e.g., a mobile phone) in the above-described method embodiments.

Embodiments of the present application further provide a computer program product, which, when running on a computer, causes the computer to perform the functions or operations performed by the electronic device (e.g., a mobile phone) in the above method embodiments.

Through the description of the above embodiments, it is clear to those skilled in the art that, for convenience and simplicity of description, the foregoing division of the functional modules is merely used as an example, and in practical applications, the above function distribution may be completed by different functional modules according to needs, that is, the internal structure of the device may be divided into different functional modules to complete all or part of the above described functions.

In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described embodiments of the apparatus are merely illustrative, and for example, a module or a unit may be divided into only one logic function, and may be implemented in other ways, for example, a plurality of units or components may be combined or integrated into another apparatus, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

Units described as separate parts may or may not be physically separate, and parts displayed as units may be one physical unit or a plurality of physical units, may be located in one place, or may be distributed to a plurality of different places. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application may be essentially or partially contributed to by the prior art, or all or part of the technical solutions may be embodied in the form of a software product, where the software product is stored in a storage medium and includes several instructions to enable a device (which may be a single chip, a chip, or the like) or a processor (processor) to execute all or part of the methods of the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

The above description is only an embodiment of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions within the technical scope of the present disclosure should be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A cold start recommendation method is applied to a server and is characterized by comprising the following steps:

receiving a recommendation list acquisition request from a terminal device, wherein the recommendation list acquisition request comprises the model of the terminal device used by a new user, and the model of the terminal device is a first model;

determining a first commodity access record and a first version access record corresponding to a first model, wherein the first commodity access record is obtained by screening from a commodity access record of a first application, the first version access record is obtained by screening from a version access record of a second application, the first application is a shopping application, and the second application is a community application;

obtaining a first recommendation list according to the first commodity access record;

obtaining a second recommendation list according to the first version access record;

determining a similar model of the first model and a second commodity access record corresponding to the similar model, and obtaining a third recommendation list according to the second commodity access record, wherein the first recommendation list, the second recommendation list and the third recommendation list all comprise commodity information and a scoring result corresponding to commodities;

removing duplication and sorting commodity information in the first recommendation list, the second recommendation list and the third recommendation list to obtain a final recommendation list, wherein the final recommendation list is used for recommending commodities for a new user;

and sending the final recommendation list to the terminal equipment.

2. The method of claim 1, wherein after the deriving the third recommendation list, the method further comprises:

determining a first type word corresponding to the post related to each block access record in the first block access record based on the post related to each block access record in the first block access record, and creating a mapping relation between the post and the first type word, wherein the first type word is a word related to a product name in the post;

determining a post preferred by a user based on the first version access record;

determining a first type word corresponding to the post preferred by the user according to the mapping relation;

determining commodity information corresponding to the first type of words by using a fuzzy matching algorithm, and generating a fourth recommendation list according to the commodity information corresponding to the first type of words;

the removing duplication and sorting of the commodity information in the first recommendation list, the second recommendation list and the third recommendation list to obtain a final recommendation list includes:

and performing duplicate removal and sorting on the commodity information in the first recommendation list, the second recommendation list, the third recommendation list and the fourth recommendation list to obtain the final recommendation list.

3. The method of claim 1, wherein after the deriving the third recommendation list, the method further comprises:

determining a second type word corresponding to the post related to each block access record in the first block access record based on the post related to each block access record in the first block access record, wherein the second type word is a word related to product characteristics in the post;

sequencing according to the occurrence frequency of all the second type words to generate a first sequencing result;

generating a fifth recommendation list based on the commodity selling points associated with the second type words in the first sequencing result, wherein the fifth recommendation list comprises commodity information corresponding to the commodity selling points;

removing duplication and sorting commodity information in the first recommendation list, the second recommendation list, the third recommendation list and the fifth recommendation list to obtain a final recommendation list;

or,

and performing duplicate removal and sorting on the commodity information in the first recommendation list, the second recommendation list, the third recommendation list, the fourth recommendation list and the fifth recommendation list to obtain the final recommendation list.

4. The method of claim 1, wherein deriving a first recommendation list based on the first item access record comprises:

scoring the commodities related to each commodity access record in the first commodity access record to obtain a scoring result of the commodities related to each commodity access record;

accumulating the scoring results of the same commodity to obtain the scoring results of all commodities;

and sorting the commodities from high to low according to the scoring results of all commodities to obtain a first recommendation list.

5. The method of claim 4, wherein the scoring the items involved in each item access record in the first item access record to obtain a scoring result of the items involved in each item access record comprises:

based on the first interaction behavior of the user and the commodities in each commodity access record, scoring the commodities related to each commodity access record to obtain a first scoring result;

based on the occurrence time of the first interaction behavior in each commodity access record, scoring the commodities related to each commodity access record to obtain a second scoring result;

and multiplying the first scoring result and the second scoring result to obtain the scoring result of the commodity related to each commodity access record.

6. The method according to claim 5, wherein the scoring the commodities related to each commodity access record based on the first interaction behavior of the user with the commodities in each commodity access record to obtain a first scoring result comprises:

and comparing the first interaction behaviors of the user and the commodities in each commodity access record with behaviors in a behavior scoring rule to obtain a first scoring result, wherein the behavior scoring rule comprises the behaviors and scoring results corresponding to the behaviors.

7. The method of claim 5, wherein the scoring the item involved in each item access record based on the occurrence time of the first interaction behavior in each item access record to obtain a second scoring result comprises:

and inputting the occurrence time of the first interactive behavior into a time attenuation function based on the occurrence time of the first interactive behavior in each commodity access record to obtain a second scoring result.

8. The method of claim 1, wherein deriving a second recommendation list based on the first version of the access record comprises:

determining a user preference block and a user group corresponding to the user preference block based on the first block access record;

and determining the commodities preferred by the users in the user group from the commodity access records based on the names of the users in the user group, and generating a second recommendation list based on the commodities preferred by the users.

9. The method of claim 8, wherein determining the user preference block and the user group corresponding to the user preference block based on the first version access record comprises:

and based on the second interaction behavior between the user and the layout in each layout access record in the first layout access record and the occurrence time of the second interaction behavior, scoring the layout related to each layout access record to obtain the user group corresponding to the user preference layout and the user preference layout.

10. The method of claim 1, wherein prior to said determining a similar model number of said first model, said method further comprises:

distinguishing the commodity access records according to equipment models, and obtaining a user behavior sequence corresponding to each equipment model based on the distinguished commodity access records, wherein the user behavior sequence comprises products accessed by users;

the determining a similar model of the first model includes:

determining the similarity between the user behavior sequence of the first model and the user behavior sequence of the second model, wherein when the similarity between the user behavior sequence of the first model and the user behavior sequence of the second model meets a threshold value, the second model is the similar model of the first model, and the second models are all models except the first model;

determining a second commodity access record corresponding to the similar model, and obtaining a third recommendation list according to the second commodity access record, wherein the third recommendation list comprises:

distinguishing the commodity access records according to equipment models to obtain second commodity access records corresponding to the similar models;

based on the second commodity access record, scoring the commodities related to the second commodity access record to obtain a commodity scoring result corresponding to the similar model;

and generating a third recommendation list based on the commodity scoring result.

11. The method according to claim 1 or 2, characterized in that the model of the terminal device used by the new user is determined to be the first model through an application program interface, API, exposed by the operating system of the terminal device.

12. The method according to claim 1 or 2, wherein the commodity access record of the first application and the block access record of the second application are obtained through page burial points, the page burial points comprise handwriting burial points and non-burial points, the handwriting burial points are located at preset detection positions, if data are detected, the data are reported, and the non-burial points are all data reported.

13. The method of claim 5 or 6, wherein the first interaction behavior of the user with the commodity comprises clicking into a commodity detail page, collecting the commodity, joining a shopping cart, clicking to purchase the commodity immediately, submitting a commodity purchase order, paying for the commodity order successfully, canceling the commodity collection, sharing the commodity, searching for the commodity, confirming that the commodity arrival notification is successful, consulting the commodity, commenting on the commodity, and clicking on the commodity in the browsing footprint.

14. The method of claim 9, wherein the second interaction behavior of the user with the layout comprises browsing of posts by the user under the layout, comments by the user under the posts, replies to comments by the user under the posts, sharing of posts by the user, favorite posts by the user, endorsement of posts by the user, detailed information of the user that clicks on the comment posts by the user, posts posted by the user that clicks on the comment posts, and appreciation of posts by the user.

15. A cold start recommendation method is applied to terminal equipment and is characterized by comprising the following steps:

responding to an operation of a new user for logging in a first application, and sending a request for obtaining a recommendation list to a server, wherein the recommendation list request comprises the model of terminal equipment used by the new user, and the model of the terminal equipment is a first model;

receiving a final recommendation list sent by the server;

and recommending commodities for the new user according to the final recommendation list.

16. A server, characterized in that the server comprises: a wireless communication module, memory, and one or more processors; the wireless communication module, the memory and the processor are coupled;

wherein the memory is to store computer program code comprising computer instructions; the computer instructions, when executed by the processor, cause the server to perform the method of any of claims 1-14.

17. A terminal device, characterized in that the terminal device comprises: a wireless communication module, memory, and one or more processors; the wireless communication module, the memory and the processor are coupled;

wherein the memory is to store computer program code comprising computer instructions; the computer instructions, when executed by the processor, cause the terminal device to perform the method of claim 15.

18. A computer-readable storage medium comprising computer instructions;

the computer instructions, when executed on a terminal device, cause the server to perform the method of any of claims 1-14, cause the terminal device to perform the method of claim 15.