CN115422463B

CN115422463B - User analysis pushing processing method and system based on big data

Info

Publication number: CN115422463B
Application number: CN202211179150.4A
Authority: CN
Inventors: 高冬; 王莉
Original assignee: Ear Pattern Yuan Intelligent Technology Guangdong Co ltd
Current assignee: Ear Pattern Yuan Intelligent Technology Guangdong Co ltd
Priority date: 2022-09-27
Filing date: 2022-09-27
Publication date: 2024-04-19
Anticipated expiration: 2042-09-27
Also published as: CN115422463A

Abstract

According to the user analysis pushing processing method and system based on big data, through the processing of identifying, matching, combining, screening, mining and the like of multiple types of behavior preference items in the user interaction activity information set, preference analysis and demand mining tasks under a complex user interaction activity scene can be effectively conducted, and the upstream and downstream characteristics of the behavior preference items can be introduced to conduct mining and prediction of pushing decision requirements, so that pushing decision requirements of the user interaction activity information set can be accurately and reasonably obtained through mining and prediction from the complex and multiple behavior preference items, and a reliable analysis basis is provided for follow-up personalized and targeted pushing.

Description

User analysis pushing processing method and system based on big data

Technical Field

The invention relates to the technical field of big data, in particular to a user analysis pushing processing method and system based on big data.

Background

The interest preference of the user is analyzed through the big data technology, and the information matched with the interest preference is pushed, so that the method has the characteristics of high efficiency, humanization, individuation and the like. It is conceivable that the current mass of information would be difficult to accomplish effectively without the intervention of an algorithm. At present, for the push repeat defect of big data, the following method can be generally used: information in a single aspect is prevented from being focused, and popular, single and entertainment content pushed by a platform is actively avoided. Therefore, in order to improve personalized and targeted efficient pushing, it is very important to mine the pushing requirement of the user, however, the related technology is difficult to accurately and reasonably mine the pushing requirement in a complex environment.

Disclosure of Invention

In order to improve the technical problems in the related art, the invention provides a user analysis pushing processing method and system based on big data.

In a first aspect, an embodiment of the present invention provides a method for processing user analysis pushing based on big data, which is applied to a user analysis pushing system, where the method includes: acquiring a user interaction activity information set meeting the requirement of big data mining; performing continuous mining analysis on at least two types of behavior preference items in the user interaction activity information set to obtain directional mining information of each behavior preference item; determining upstream and downstream characteristics between the no less than two types of behavior preference items by combining the directional mining information of the no less than two types of behavior preference items; determining business session interaction information corresponding to the behavior preference items according to set information extraction rules through directional mining information of the behavior preference items; the information extraction rule is used for guiding the number of the business session interaction information of the same behavior preference; and determining push decision requirements of the user interaction activity information set by combining the upstream and downstream characteristics between the at least two types of behavior preference items and business session interaction information of each behavior preference item.

By means of the design, through the processes of identifying, matching, combining, screening, mining and the like of the multiple types of behavior preference items in the user interaction activity information set, preference analysis and demand mining tasks in a complex user interaction activity scene can be effectively conducted, upstream and downstream characteristics of the behavior preference items can be introduced to conduct mining and prediction of push decision requirements, push decision requirements of the user interaction activity information set can be accurately and reasonably obtained through mining and prediction from the complex and various behavior preference items, and reliable analysis basis is provided for follow-up personalized and targeted pushing.

For some independently implementable embodiments, the obtaining a set of user interaction activity information that meets big data mining requirements includes: acquiring not less than two groups of business interaction logs collected by an information collection unit; processing one group of business interaction logs in the two groups of business interaction logs by using a log processing algorithm not lower than two to obtain current user activity data in the corresponding group of business interaction logs; and taking the current user activity data in the service interaction logs not lower than two groups as the user interaction activity information set.

By the design, a plurality of groups of service interaction logs output a group of user activity data in parallel in a plurality of log processing algorithms to form a user interaction activity information set, so that behavior preference items in the user interaction activity information set can be enriched, and the reliability and rationality of pushing decision requirement mining are improved.

For some embodiments that may be implemented independently, the directional mining information includes an identification core, a matching tag, and a matching record of the behavior preferences, and the performing persistent mining analysis on at least two types of behavior preferences in the user interaction information set to obtain directional mining information of each of the behavior preferences includes: performing behavior preference recognition on at least two types of behavior preference items of current user activity data in the user interaction activity information set to obtain recognition cores of the behavior preference items; and matching the corresponding behavior preference items by combining the identification cores of the behavior preference items in the user interaction activity information set to obtain matching labels and matching records corresponding to the behavior preference items.

The method comprises the steps of firstly identifying multiple types of behavior preference items on current user activity data to obtain identification cores of the behavior preference items, and then further matching corresponding behavior preference items based on the identification cores of the behavior preference items so as to ensure the identification accuracy of single behavior preference items.

For some independently implementable embodiments, the at least two types of behavior preferences include a mainstream behavior preference and an edge behavior preference; the mainstream behavior preference reflects a push demand topic; performing behavior preference recognition on at least two types of behavior preference items of current user activity data in the user interaction activity information set to obtain recognition cores of the behavior preference items, wherein the recognition cores comprise: performing behavior preference identification on the main stream behavior preferences of the current user activity data in the user interaction activity information set to obtain an identification core of the main stream behavior preferences; performing behavior preference identification on the edge behavior preference in the current user activity data based on the current user activity data as selected user activity data to obtain a basic identification core of the edge behavior preference; the selected user activity data is extracted according to a set extraction step length; and carrying out preference matching processing on the edge behavior preference in the current user activity data based on the fact that the current user activity data is non-selected user activity data, so as to obtain a candidate identification core of the edge behavior preference.

By the design, the thought groups of the selected user activity data and the unselected user activity data estimation are adopted for the edge behavior preference, so that the abuse of an algorithm can be reduced, the timeliness of behavior preference item identification is improved, and the light weight of the whole scheme is realized.

For some independently implementable embodiments, each current user activity data in the set of user interaction activity information comprises a digital authentication signature; the performing preference matching processing on the edge behavior preference in the current user activity data based on the current user activity data as the non-selected user activity data to obtain a candidate identification core of the edge behavior preference, including: combining the basic identification core of the edge behavior preference, and adjusting a first preference matching processing model; the base identification core is determined from selected user activity data prior to the digital authentication signature preceding the current user activity data; and estimating the distribution variable of the edge behavior preference in the current user activity data through the adjusted first preference matching processing model to obtain a candidate identification core of the edge behavior preference.

According to the design, the first preference matching processing model is adjusted by utilizing the basic identification core of the edge behavior preference identified by the selected user activity data, and the candidate identification core of the edge behavior preference in the unselected user activity data is estimated through the first preference matching processing model, so that the advantages of good timeliness and high accuracy of the preference matching processing are utilized, and the efficiency of pushing demand mining analysis is ensured.

For some embodiments that may be implemented independently, the matching the corresponding behavior preference item with the identification core of each behavior preference item in the user interaction information set to obtain a matching tag and a matching record corresponding to the behavior preference item includes: loading the identification cores of all the behavior preference items in the user interaction activity information set to a second preference matching processing model to obtain matching labels of the behavior preference items; and determining a matching record corresponding to the behavior preference item based on the identification cores corresponding to the same behavior preference item and the matching labels corresponding to the behavior preference item in a group of business interaction logs.

By means of the design, the second preference matching processing model is used for outputting the matching labels of the behavior preference items and determining the matching records, single identification windows mined by each group of activity information can be associated, and the single identification windows can be treated as the behavior preference items binding the same matching labels for processing, so that resource expenditure can be saved in the subsequent behavior preference item processing process.

For some independently implementable embodiments, the at least two types of behavioral preferences include the following: virtual mall browsing items, cross-border e-commerce attention items, VR service preference items, MR service preference items, and hot topic preference items; the determining the upstream and downstream characteristics between the no less than two types of behavior preference items by combining the directional mining information of the no less than two types of behavior preference items comprises the following steps: combining the distributed variable joint analysis results between the recognition cores of the virtual mall browsing items and the recognition cores of the cross-border e-commerce attention items to determine cross-border mall associated information corresponding to the same e-commerce preference items; combining a distribution variable joint analysis result between the recognition core of the cross-border e-commerce attention item and the recognition core of the VR service preference item to determine somatosensory demand description characteristics between the cross-border e-commerce attention item and the VR service preference item; and determining a first project participation characteristic or a second project participation characteristic between the hot topic preference item and the MR service preference item according to the distribution variable of the identification core of each hot topic preference item and the MR service preference item and the matching record of each hot topic preference item and the MR service preference item.

The method comprises the steps that through the adoption of the identification cores of all behavior preference items, virtual mall browsing items and cross-border e-commerce attention items corresponding to the same business objects are subjected to linkage analysis, and cross-border e-commerce attention items with somatosensory demand description characteristics and VR service preference items are subjected to linkage analysis; and meanwhile, combining the identification cores and the matching records of the behavior preference items, carrying out linkage analysis on the cross-border e-commerce attention item and the MR service preference item, and facilitating analysis and determination of the business interaction log as a push decision requirement.

For some embodiments that may be implemented independently, the directional mining information further includes a merit factor of the behavior preference, and determining, by the directional mining information of each behavior preference, the business session interaction information corresponding to the behavior preference according to a set information extraction rule includes: determining initial session interaction information by combining the identification core of the first behavior preference item based on the fact that the merit factor of the first behavior preference item in the current user activity data reaches a first merit score; the first behavior preference item is any one of the at least two types of behavior preference items; inputting the initial session interaction information into a temporary storage space of the first behavior preference item; and determining the session interaction information recorded in the temporary storage space as the business session interaction information corresponding to the first behavior preference item based on the temporary storage space of the first behavior preference item reaching a preset screening starting requirement.

In such a design, for each behavior preference item, whether the recognized goodness index in the current user activity data reaches a first goodness score or not is firstly evaluated, namely a screening condition is obtained, then initial session interaction information of the behavior preference item is determined and is input into a temporary storage space, and finally service session interaction information with the best evaluation value recorded in the temporary storage space is screened out according to a preset screening starting requirement, so that the accuracy of the subsequent knowledge vector mining is improved while the space utilization rate is improved.

For some independently implementable embodiments, said entering said initial session interaction information into a temporary space of said first behavioral preference comprises: and recording the initial session interaction information into the temporary storage space of the first behavior preference item unconditionally on the basis that the number of the session interaction information recorded in the temporary storage space of the first behavior preference item does not reach a storage limit value.

By means of the design, the number of the session interaction information recorded in the temporary storage space is compared with the storage limit value, initial session interaction information is recorded in the temporary storage space on the basis that the temporary storage space of the behavior preference item is not saturated, and the omission of the initial session interaction information of the behavior preference item is avoided.

For some independently implementable embodiments, said entering said initial session interaction information into a temporary space of said first behavioral preference comprises: removing the first session interaction information in the temporary storage space on the basis that the number of the session interaction information recorded in the temporary storage space of the first behavior preference reaches the storage limit value; the quality coefficient of the first session interaction information is lower than that of the initial session interaction information; and inputting the initial session interaction information into the temporary storage space of the first behavior preference.

By comparing the number of the session interaction information recorded in the temporary storage space with the storage limit value, the session interaction information with poor recognition evaluation value in the temporary storage space is removed firstly on the basis of judging the saturation of the temporary storage space of the behavior preference, then the initial session interaction information is recorded in the temporary storage space, overload of the temporary storage space of the behavior preference is reduced, and loss of the session interaction information with high evaluation value is caused.

For some independently implementable embodiments, the method further comprises: and initializing the temporary storage space of the first behavior preference item based on the temporary storage space of the first behavior preference item reaching the screening starting requirement.

By the design, when the temporary storage space of the first behavior preference item reaches the screening starting requirement, the screening task is completed, and the initialization of the temporary storage space is convenient for the normal operation of the screening task of the next user interaction information set.

For some independently implementable embodiments, the preset screening initiation requirements include at least one of: screening timeliness requirements, screening intermittent requirements, matching time-consuming screening conditions, good and bad scoring screening requirements and matching termination screening requirements; the screening timeliness requirement reflects that the time length for continuously positioning the first behavior preference reaches a first set time length; the screening intermittent requirement reflects that the gap period matched with the first behavior preference reaches a set gap value; the time-consuming screening condition reflects that the total time length of the first behavior preference item is matched to reach a second set time length; the second set duration is not less than the first set duration; the priority score screening requirement reflects session interaction information with priority coefficients reaching second priority scores in the temporary storage space of the first behavior preference; the second merit score is greater than the first merit score; the match termination screening requirement reflects that the overall hierarchy matches the first behavioral preference until the collected business interaction log is complete.

By the design, screening starting requirements of each behavior preference item can be flexibly set, so that the number of business session interaction information used for analysis of the same behavior preference item is limited based on multiple aspects, the evaluation value of behavior preference item screening is improved, and the processing efficiency is improved.

For some independently implementable embodiments, said determining a push decision requirement of said set of user interaction information in combination with said business session interaction information for each of said behavioral preferences and upstream features between said at least two types of behavioral preferences comprises: carrying out knowledge vector mining and/or element vector mining on the business session interaction information of each behavior preference to obtain the demand prediction information corresponding to the behavior preference; combining the demand prediction information of the at least two types of behavior preference items by combining upstream and downstream characteristics between the at least two types of behavior preference items; and determining push decision requirements of the user interaction activity information set based on the requirement prediction information of the at least two types of behavior preference items which are combined.

According to the design, big data mining is conducted on the screened business session interaction information, and the requirement prediction information is combined by combining the upstream and downstream characteristics among different behavior preference items so as to determine the push decision requirement of the user interaction information set, so that the precision and rationality of push decision requirement mining and prediction under the complex session environment are improved.

For some independently implementable embodiments, the behavioral preference is a virtual mall browsing item having a number of the business session interaction information of not less than 2; the knowledge vector mining and/or element vector mining are performed on the business session interaction information of each behavior preference item to obtain the demand prediction information corresponding to the behavior preference item, which comprises the following steps: respectively carrying out knowledge vector mining on the service session interaction information which is not lower than two of the virtual mall browsing items to obtain knowledge description blocks which are not lower than two of the virtual mall browsing items; splicing the knowledge description blocks which are not lower than two to obtain a linkage descriptor of the virtual mall browsing item; element vector mining is carried out on the business session interaction information with the highest quality index in the business session interaction information to obtain element description blocks of the virtual mall browsing items; and determining the demand prediction information of the virtual mall browsing item based on the linkage descriptor of the virtual mall browsing item and the element description block of the virtual mall browsing item.

By means of the design, the feature output quality of the descriptors can be improved through vector mining and splicing of the service session interaction information of the plurality of virtual mall browsing items, and meanwhile, the credibility of demand mining prediction can be improved through element vector mining of the service session interaction information with the best evaluation value.

For some independently implementable embodiments, the method operates with set rules by not less than the executing nodes in both log processing algorithms; the setting rule reflects the generation result of the last execution node between every two adjacent execution nodes and is used as the input information of the next execution node through an information transition strategy.

The method is designed by adopting a multi-log processing algorithm to design a plurality of execution nodes to execute the user analysis pushing processing method based on big data, and the information transition strategy is matched between every two adjacent execution nodes, so that the processing efficiency of the whole scheme can be improved.

In a second aspect, the invention also provides a user analysis pushing system, which comprises a processor and a memory; the processor is in communication with the memory, and the processor is configured to read and execute a computer program from the memory to implement the method described above.

In a third aspect, the present invention also provides a server, including a processor and a memory; the processor is in communication with the memory, and the processor is configured to read and execute a computer program from the memory to implement the method described above.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.

Fig. 1 is a flow chart of a user analysis push processing method based on big data provided by an embodiment of the invention.

Fig. 2 is a schematic diagram of a communication architecture of an application environment of a user analysis push processing method based on big data according to an embodiment of the present invention.

Detailed Description

Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the invention. Rather, they are merely examples of apparatus and methods consistent with aspects of the invention as detailed in the accompanying claims.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order.

The method embodiment provided by the embodiment of the invention can be executed in a user analysis pushing system, computer equipment or similar computing devices. Taking the example of running on a user analysis push system, the user analysis push system 10 may include one or more processors 102 (the processor 102 may include, but is not limited to, a microprocessor MCU or a processing device such as a programmable logic device FPGA) and a memory 104 for storing data, and optionally the user analysis push system may also include a transmission device 106 for communication functions. It will be appreciated by those of ordinary skill in the art that the above-described structure is merely illustrative and is not intended to limit the structure of the user analysis push system. For example, the user analysis push system 10 may also include more or fewer components than shown above, or have a different configuration than shown above.

The memory 104 may be used to store a computer program, for example, a software program of application software and a module, such as a computer program corresponding to a method for processing user analysis and push based on big data in the embodiment of the present invention, and the processor 102 executes the computer program stored in the memory 104 to perform various functional applications and data processing, that is, implement the method described above. Memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include memory remotely located with respect to the processor 102, which may be connected to the user analysis push system 10 via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The transmission means 106 is arranged to receive or transmit data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider of the user analysis push system 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, simply referred to as a NIC) that can connect to other network devices through a base station to communicate with the internet. In one example, the transmission device 106 may be a Radio Frequency (RF) module, which is used to communicate with the internet wirelessly.

Based on this, referring to fig. 1, fig. 1 is a flow chart of a user analysis push processing method based on big data according to an embodiment of the present invention, where the method is applied to a user analysis push system, and further may include the following technical solutions described in the following.

The PROCESS110 obtains a set of user interaction activity information that meets the requirements of big data mining.

In the embodiment of the invention, the user interaction activity information set meeting the big data mining requirement can be a user interaction activity information set within a set period or a user interaction activity information set correspondingly determined according to a demand analysis instruction, wherein the user interaction activity information set comprises two or more user interaction activity information. Further, the user interaction information acquisition mode is not limited, and the user interaction information can be acquired directly through the user analysis pushing system, and also can be received through other systems. Further, the user interaction information may be activity records corresponding to different digital services, including but not limited to e-commerce, digital office, teleconferencing, smart city services, and the like.

And performing continuous mining analysis on at least two types of behavior preference items in the user interaction activity information set by using the PROCESS120 to obtain directional mining information of each behavior preference item.

In the embodiment of the invention, the at least two types of behavior preference items are important behavior preference items or behavior preference items with higher attention in the pushing demand mining process, such as virtual mall browsing items, cross-border e-commerce attention items, VR service preference items, MR service preference items, hot topic preference items and the like based on user pushing demand mining. In the actual implementation process, various behavior preference items in each user interaction information in the user interaction information set are mined and positioned uninterruptedly, in other words, various behavior preference items can be continuously mined and tracked.

Further, the process of behavior preference identification may be understood as performing distributed variable identification and grouping operations of multiple types of behavior preferences in user interaction information using a DNN network. The targeted mining information for the behavior preferences may include identification kernels (i.e., identification windows) for the behavior preferences, categories of the behavior preferences, matching tags (such as continuously analyzed location tags), matching records (continuously analyzed location tracks), and the like. In addition, the behavior preference item can be further understood as a user behavior preference event, and the user behavior preference event can reflect the preference or tendency of the user, so that the mining analysis of the related push requirements can be performed based on the user behavior preference event.

For some embodiments that may be implemented independently, the obtained set of user interaction information may also be normalized prior to identifying the behavior preferences of the user interaction information, so as to weaken the group characteristics of the user interaction information and highlight the personalized characteristics of the user interaction information.

PROCESS130, in conjunction with the directional mining information of the at least two types of behavior preferences, determines upstream and downstream features between the at least two types of behavior preferences.

In the embodiment of the invention, after the directional mining information of each behavior preference in the same user interaction information is obtained, the upstream and downstream characteristics between the virtual mall browsing items and the cross-border electronic commerce attention items corresponding to the same electronic commerce preference can be mined according to the identification cores and the matching labels of the behavior preference; the upstream and downstream features corresponding to the somatosensory demand description features can be mined according to the relationship between the recognition kernels of the two behavior preference items; and the upstream and downstream characteristics of the project participation characteristics can be mined according to the distribution variables of the identification window and the matching records.

And the PROCESS140 determines the business session interaction information corresponding to the behavior preference according to the set information extraction rule through the directional mining information of each behavior preference.

In the embodiment of the invention, after all the identified behavior preference items in the user interaction information set are matched and the upstream and downstream characteristics are determined, the identification evaluation values of the behavior preference items can be calculated, so that the business session interaction information with the quality coefficient reaching the requirement is determined to be subjected to subsequent processing analysis.

Wherein the information extraction rule is used for guiding the number of the business session interaction information of the same behavior preference; the screening can be understood as that after the priority coefficients of the business session interaction information in each group of activity information in the whole matching recording process are sequentially adjusted, the behavior preference item on the business session interaction information with the highest evaluation value is selected for processing.

Further, the matching behavior preference item refers to a behavior preference item with a matching tag, which is continuously present in a series of user interaction information, and the business session interaction information is a part of information set divided from the user interaction information based on the identified behavior preference item identification core, and the behavior preference item information (including the identification core of the behavior preference item, the behavior preference item information set and other identification information) of the matching behavior preference item in a certain user interaction information is included for later knowledge vector mining, element description block analysis and the like.

It can be understood that the preference coefficients are determined for each behavior preference item respectively, and the thought groups determined for different behavior preference items can be flexibly selected, and only the preference coefficient interval of the behavior preference items is ensured to be [0,1]. Therefore, the service session interaction information with the optimal evaluation value can be conveniently selected for feature mining analysis after being adjusted according to the order of the quality coefficients.

And the PROCESS150 determines the push decision requirement of the user interaction information set by combining the upstream and downstream characteristics between the at least two types of behavior preference items and the business session interaction information of each behavior preference item.

In the embodiment of the invention, knowledge vector mining and element description block analysis are carried out on the service session interaction information of each behavior preference item which is screened and output, and the processing results are combined to be used as push decision requirements of a user interaction activity information set by combining with upstream and downstream characteristics among different behavior preference items, and prediction processing or regression analysis processing can be carried out according to the upstream and downstream characteristics among at least two types of behavior preference items and the service session interaction information of each behavior preference item, so that push decision requirements are obtained, and the push decision requirements can be used for guiding big data push aiming at users, thereby improving push precision and pertinence. .

Further, after the knowledge description block and the element description block of the service session interaction information of all the behavior preference items which are screened and output are extracted, all the behavior preference items are combined, and then each service session interaction information of each group of service interaction logs is processed in an iterative manner until the processing is completed.

It will be appreciated that the PROCESS110-PROCESS150 of the present embodiment may run in a loop, i.e., continue to obtain the latest set of user interaction information for processing until all service session interaction information processing is completed or a stop indication is received.

In the embodiment of the invention, through carrying out the processes of identifying, matching, combining, screening, mining and the like on the multiple types of behavior preference items in the user interaction information set, the preference analysis and the requirement mining task under the complex user interaction scene can be effectively processed, and the upstream and downstream characteristics of the behavior preference items can be introduced to carry out the mining and the prediction of the push decision requirement, so that the push decision requirement of the user interaction information set can be accurately and reasonably mined and predicted from the complex and various behavior preference items, and a reliable analysis basis is provided for the follow-up personalized and targeted push.

For some independently implementable embodiments, the PROCESS110 through PROCESS150 run with no less than two execution nodes (e.g., as related functional modules) in a log processing algorithm (e.g., as a computer thread) to set rules (e.g., synchronous parallel processing); the setting rule reflects the generation result of the last executing node as the input information of the next executing node through an information transition strategy (for example, a buffer space can be set to transmit data information in the node processing process).

By the design, different execution nodes operate according to different log processing algorithms, and the processing efficiency of the whole scheme can be improved. For example, a plurality of execution nodes are configured to run the user analysis pushing processing method based on big data based on a multi-log processing algorithm, and an information transition strategy (such as setting a cache space) is matched between every two adjacent execution nodes, so that the processing efficiency of the overall scheme can be improved.

In some examples, the directed mining information includes an identification core, a matching tag, and a matching record of the behavior preference. The following is a related exemplary description of a user analysis push processing method based on big data.

The PROCESS210 obtains not less than two sets of service interaction logs collected by the information collection unit.

And processing one group of service interaction logs in the two groups of service interaction logs by the processing algorithm not lower than two logs to obtain current user activity data in the corresponding group of service interaction logs.

In the embodiment of the invention, a data extractor is used in each log processing algorithm to process a group of business interaction logs or interaction log texts (such as activity data extraction operation), for example, a log processing algorithm algorithm_A is used to process a1 st group of business interaction logs, and a log processing algorithm algorithm_B is used to process a2 nd group of business interaction logs, so that a plurality of log processing algorithms can be used to process a plurality of groups of business interaction logs simultaneously, and current user activity data in each group of business interaction logs can be obtained simultaneously.

And using the current user activity data in the service interaction logs of not lower than two groups as the user interaction activity information set by the PROCESS 230.

In the embodiment of the invention, a plurality of groups of business interaction logs extract current user activity data in a plurality of log processing algorithms at the same time to form a data set, namely a user interaction activity information set.

And performing behavior preference identification on at least two types of behavior preference items of the current user activity data in the user interaction activity information set by using the PROCESS240 to obtain identification cores of the behavior preference items.

In the embodiment of the invention, the identification core of the behavior preference is the identification window. The behavior preference identification can be performed on a group of user interaction information through the unary regression sub-network, and the behavior preference identification can be performed on a group of user interaction information through the multiple regression sub-network.

For some embodiments that can be implemented independently, for each behavior preference, behavior preference identification is performed by using a unary regression sub-network in turn, so as to obtain an identification core of the behavior preference. For example, for the virtual mall browsing item with the unique identifiable user tag, the unary regression sub-network is used for identifying the virtual mall browsing item in each piece of user interaction information, so that the identification accuracy of the virtual mall browsing item is ensured.

In other examples, behavior preference identification is performed on multiple types of behavior preferences simultaneously using multiple regression sub-networks, resulting in identification kernels for each behavior preference in a set of user interaction information. For example, for edge behavior preferences such as VR service preference items, MR service preference items, hot topic preference items and the like, the multiple regression sub-network is utilized to conduct behavior preference identification in user interaction information selected according to a certain period, so that behavior preference identification efficiency is improved.

And (2) matching the corresponding behavior preference items by combining the identification cores of the behavior preference items in the user interaction information set through the PROCESS250 to obtain matching labels and matching records corresponding to the behavior preference items.

In the embodiment of the invention, the second preference matching processing model outputs the matching label of the behavior preference item, and the same behavior preference item in the front and back information generally has the same unique matching label and can be used for carrying out subsequent correlation or screening analysis of the behavior preference item.

In the actual implementation process, firstly, multiple types of behavior preference items on the current user activity data are identified to obtain identification cores of the behavior preference items, then, corresponding behavior preference items are further matched based on the identification cores of the behavior preference items, and single identification windows mined out by each group of activity information are associated through a matching tag to form a matching record so as to ensure the identification accuracy of the single behavior preference items.

PROCESS260, in conjunction with the directional mining information of the at least two types of behavior preferences, determines upstream and downstream features between the at least two types of behavior preferences.

In the embodiment of the invention, the directional mining information comprises an identification core, a matching tag and a matching record of the behavior preference. The upstream and downstream features between the at least two types of behavior preference items can comprise cross-border mall association information, somatosensory requirement description features, first item participation features or second item participation features and the like, so that the upstream and downstream features can be further understood as association relations.

Further, the cross-border mall associated information mainly refers to linkage analysis of virtual mall browsing items and cross-border electronic commerce attention items corresponding to the same business object; the somatosensory demand description features refer to upstream and downstream features of the cross-border e-commerce attention item and the VR service preference item, which are respectively corresponding to the somatosensory demand description features; the first item engagement feature and the second item engagement feature each refer to upstream and downstream features between the cross-border e-commerce attention item and the MR service preference item.

Further, the upstream and downstream features may be determined by the following: combining the distributed variable joint analysis results between the recognition cores of the virtual mall browsing items and the recognition cores of the cross-border e-commerce attention items to determine cross-border mall associated information corresponding to the same e-commerce preference items; combining a distribution variable joint analysis result between the recognition core of the cross-border e-commerce attention item and the recognition core of the VR service preference item to determine somatosensory demand description characteristics between the cross-border e-commerce attention item and the VR service preference item; and determining a first project participation characteristic or a second project participation characteristic between the hot topic preference item and the MR service preference item by combining distribution variables (such as position information) of the identification cores of the hot topic preference item and the MR service preference item respectively and matching records of the hot topic preference item and the MR service preference item respectively.

By means of the design, the recognition cores of the behavior preference items are utilized to carry out linkage analysis on the virtual mall browsing items and the cross-border e-commerce attention items corresponding to the same business objects, and the cross-border e-commerce attention items with the somatosensory demand description characteristics and the VR service preference items are subjected to linkage analysis. And further combining the identification cores and the matching records of the behavior preference items, carrying out linkage analysis on the cross-border e-commerce attention item and the MR service preference item, and facilitating the prediction mining processing of the push decision requirement.

The PROCESS270 determines the business session interaction information corresponding to the behavior preferences according to the set information extraction rules through the directional mining information of each behavior preference.

In the embodiment of the invention, the information extraction rule is used for guiding the number of the business session interaction information of the same behavior preference.

And the PROCESS280 combines the upstream and downstream characteristics between the at least two types of behavior preference items and the business session interaction information of each behavior preference item to determine the push decision requirement of the user interaction activity information set.

In the embodiment of the invention, a group of user activity data is output in parallel in a plurality of log processing algorithms for a plurality of groups of business interaction logs to form a user interaction activity information set, then behavior preference identification is carried out on the user interaction activity information set simultaneously to obtain the identification core of each behavior preference, and the corresponding behavior preference is further matched based on the identification core of the behavior preference so as to ensure the identification accuracy of a single behavior preference, thereby realizing the accurate identification of multiple types of behavior preference and improving the accuracy and reliability of requirement mining.

For some examples, the at least two types of behavior preferences include a mainstream behavior preference and an edge behavior preference; the mainstream behavior preferences reflect push demand topics. Based on this, the PROCESS120 "performs continuous mining analysis on at least two types of behavior preference items in the user interaction activity information set, and obtains directional mining information of each behavior preference item" may be implemented by the following technical scheme.

And obtaining behavior preference identification for the mainstream behavior preference of the current user activity data in the user interaction activity information set by the PROCESS310, and obtaining an identification core of the mainstream behavior preference.

The main stream behavior preference refers to a behavior preference item which can uniquely reflect a push demand theme in the current user activity data. For some e-commerce services, the virtual mall browsing item can be used as a mainstream behavior preference, and the user tag can be uniquely confirmed.

In the actual implementation process, the unary regression sub-network is used for identifying the virtual mall browsing item for each group of activity information in the service interaction log, so that the identification core of the virtual mall browsing item in each group of activity information, namely the item capturing window, is obtained, and the accuracy of the virtual mall browsing item is ensured.

And performing behavior preference identification on the edge behavior preference in the current user activity data based on the current user activity data as the selected user activity data by using the PROCESS320 to obtain a basic identification core of the edge behavior preference.

Further, the selected user activity data is extracted according to a set extraction step. The edge behavior preference is other behavior preference items of the business session interaction information besides the main stream behavior preference, such as a cross-border e-commerce attention item, a VR service preference item, an MR service preference item, a hot topic preference item and the like which are related to the behavior attention event.

And in the selected user activity data, identifying various edge behavior preferences simultaneously by using a multiple regression sub-network to obtain basic identification cores of the edge behavior preferences. The simultaneous identification of multiple behavioral preferences is typically performed only for cases where the current user activity data is selected user activity data, thereby avoiding excessive use of resources.

And processing 330, based on the current user activity data as non-selected user activity data, performing preference matching processing on the edge behavior preference in the current user activity data to obtain a candidate identification core of the edge behavior preference.

Further, each piece of current user activity data in the user interaction activity information set contains a digital authentication signature, and after the basic identification core of the edge behavior preference in the selected user activity data is identified in the last step, the candidate identification core (updated identification core) of each edge behavior preference in the non-selected user activity data after the estimation of the basic identification core of each edge behavior preference is utilized. Therefore, the abuse of the algorithm can be reduced for the edge behavior preference, the timeliness of behavior preference item identification is improved, and the light weight of the whole scheme is realized.

For some examples, the first preference matching process model is adjusted in conjunction with the underlying recognition kernel of the edge behavior preference; the base identification core is determined from selected user activity data prior to the digital authentication signature preceding the current user activity data; and estimating the distribution variable of the edge behavior preference in the current user activity data through the adjusted first preference matching processing model to obtain a candidate identification core of the edge behavior preference.

And the PROCESS340 loads the identification cores of all the behavior preference items in the user interaction information set to a second preference matching processing model to obtain matching labels of the behavior preference items.

In the embodiment of the present invention, for all current user activity data in the user interaction information set, after the main stream behavior preference identification of the PROCESS310 and the combination and estimation of the identification of the edge behavior preference in the PROCESS320 to the PROCESS330, the distribution variables and the categories of all the behavior preference items on the user interaction information are determined, and the second preference matching processing model (multi-classification model) can be utilized to perform multi-behavior preference item matching, so as to obtain the matching label of each identification window.

PROCESS350 determines a matching record corresponding to the behavioral preference based on the identification core corresponding to the same behavioral preference and a matching tag corresponding to the behavioral preference in a set of business interaction logs.

In the embodiment of the invention, the identification cores which are mined by the current user activity data and carry the same matching labels are associated and processed as the matching records of the behavior preference items, so that the subsequent screening analysis is facilitated.

For example, the PROCESS320-PROCESS330 employs a set of selected user activity data identification and non-selected user activity data estimation for edge behavior preferences because multiple regression sub-network scaling used across e-commerce concerns, VR service preferences, MR service preferences, and hot topic preferences is relatively large and time consuming. And by adopting the thought groups of the selected user activity data identification and the unselected user activity data estimation, the efficient advantage of the preference matching processing is utilized, and the processing efficiency can be improved. The interval of the data groups of the selected user activity data in the preference matching process is generally 4-8 groups, the speed of the data groups is difficult to be increased, if the interval of the data groups is too small, the accuracy of single behavior preference estimation is reduced, and the error of the identification core of the estimated behavior preference is larger.

In the actual implementation process, after the main stream behavior preference identification is carried out on each piece of user interaction activity information to obtain the identification core of the main stream behavior preference, the current business session interaction information is judged to be two types of selected user activity data or non-selected user activity data, and the identification core of the edge behavior preference is obtained by using the thought group estimated by the selected user activity data and the non-selected user activity data. In some embodiments, implementation of the behavior preference identification matching process provided for embodiments of the present invention may include the following.

PROCESS311 identifies the identifying core of the mainstream behavior preferences.

PROCESS312 determines whether the current user activity data is selected user activity data. If the determination is yes, then executing PROCESS313; if the determination is negative, then PROCESS315 is performed.

PROCESS313, a basic identification core that identifies edge behavior preferences;

PROCESS314, tuning the first preference matching PROCESS model;

here, the first preference matching process model (single classification model) refers to a process of estimating candidate recognition kernels of each behavior preference item in the next several groups of user activity data according to the basic recognition kernels of each behavior preference item in the selected user activity data, wherein the first preference matching process model is a process of adjusting the first preference matching process model, namely, taking the distribution variable of each recognition kernel in the selected user activity data as a basic distribution variable required by the first preference matching process model to match the new distribution variable of the estimated behavior preference item.

PROCESS315 estimates candidate recognition kernels for edge behavior preferences through a first preference matching PROCESS model.

Further, the same first preference matching process model is used for matching different edge behavior preferences in the current user activity data, and the first preference matching process model can be used for matching and estimating a plurality of different identification window distribution variables at a time. The first preference matching process model has a targeted matching process for the several classes of edge behavior preferences.

PROCESS316 loads the identification kernel of all the identified behavior preferences into the second preference matching PROCESS model, and matches each behavior preference in the current user activity data separately.

In the embodiment of the invention, the second preference matching processing model is used for carrying out linkage analysis on the identification cores corresponding to the same behavior preference items in the adjacent group of user activity data, and a unique matching label is configured, and the main process is to input the identification cores of all the behavior preference items in the current user activity data, generate the matching label of the identification core corresponding to each behavior preference item, and the same behavior preference items in the front and back information have the same unique matching label.

In the embodiment of the invention, the first preference matching processing model is adjusted by utilizing the basic identification core of the edge behavior preference identified by the selected user activity data, and the candidate identification core of the edge behavior preference in the unselected user activity data is estimated by the first preference matching processing model, so that the advantages of good timeliness and high accuracy of the preference matching processing are utilized, and the efficiency of pushing demand mining analysis is ensured. By means of the design, the second preference matching processing model is used for outputting the matching labels of the behavior preference items and determining the matching records, single identification windows mined by each group of activity information can be associated, and the single identification windows can be treated as the behavior preference items binding the same matching labels for processing, so that resource expenditure can be saved in the subsequent behavior preference item processing process.

For some examples, the directional mining information also includes a merit factor (quality score or quality rating value of the behavior preference) for the behavior preference. Based on this, the PROCESS140 "determines the business session interaction information corresponding to the behavior preferences according to the set information extraction rule through the directional mining information of each behavior preference" may include the following.

PROCESS410 determines initial session interaction information based on the first performance preference's merit factor reaching a first merit score in the current user activity data in conjunction with the first performance preference's identification core.

In the embodiment of the present invention, the first behavior preference is any one of the at least two types of behavior preference, and the first merit score reflects a lowest merit coefficient as a screening condition, for example, may be 0.3.

For some examples, if the goodness coefficient of the identified first behavior preference item is not lower than the first goodness score, determining a local information set corresponding to the identification core of the first behavior preference item from the current user activity data as initial session interaction information.

For other examples, if the goodness coefficient of the identified first behavioral preference is below the first goodness score, the identification of the first behavioral preference is deleted.

And the PROCESS420 records the initial session interaction information into the temporary storage space of the first behavior preference item.

Whether the temporary storage space of the first behavior preference item is saturated or not can be judged first, and whether the temporary storage space is directly recorded with the initial session interaction information or the session interaction information with the recorded lowest evaluation value is covered with the initial session interaction information is determined based on the judging result.

For some examples, the initial session interaction information is unconditionally entered into the temporary space of the first behavioral preference based on the number of session interaction information recorded in the temporary space of the first behavioral preference not reaching a storage limit. Therefore, the number of the session interaction information recorded in the temporary storage space is compared with the storage limit value, and the initial session interaction information is recorded in the temporary storage space on the basis that the temporary storage space of the behavior preference item is not saturated, so that the initial session interaction information of the behavior preference item is effectively reduced and deleted.

For other examples, the first session interaction information in the temporary storage space of the first behavioral preference is removed based on the number of session interaction information recorded in the temporary storage space reaching the storage limit; the quality coefficient of the first session interaction information is lower than that of the initial session interaction information; and inputting the initial session interaction information into the temporary storage space of the first behavior preference. Therefore, the conversation interaction information with poor recognition evaluation value in the temporary storage space is removed firstly on the basis of judging the saturation of the temporary storage space of the behavior preference items by comparing the number of the conversation interaction information recorded in the temporary storage space with the storage limit value, and then the initial conversation interaction information is recorded in the temporary storage space, so that the overload of the temporary storage space of the behavior preference items is effectively reduced, and the loss of the conversation interaction information with good evaluation value is caused.

And the PROCESS430 determines the session interaction information recorded in the temporary storage space as the service session interaction information corresponding to the first behavior preference item based on the temporary storage space of the first behavior preference item reaching a preset screening start requirement.

In the embodiment of the present invention, the preset screening start requirement includes at least one of the following: screening timeliness requirements, screening intermittence requirements, matching time-consuming screening conditions, good and bad scoring screening requirements, and matching termination screening requirements. Through a plurality of flexible screening starting requirements, noise interference is reduced, and screening timeliness and screening evaluation values of behavior preference items are improved.

In the embodiment of the invention, the screening timeliness requirement reflects that the duration of uninterruptedly positioning the first behavior preference item reaches the first set duration. For example, given a duration, i.e., a first set duration, a screening output is initiated when a behavior preference is once it is positioned uninterrupted beyond this duration. The screening intermittent requirement reflects that a gap period matching the first behavioral preference reaches a set gap value. The time-consuming screening condition reflects that the total duration of the matching of the first behavior preference reaches a second set duration. The priority score screening requirement reflects session interaction information with priority coefficients reaching second priority scores in the temporary storage space of the first behavior preference; the second merit score is greater than the first merit score. For example, when the merit factor of the evaluation value session interaction information recorded by the behavior preference item reaches the second merit score, the output is filtered, and then the filtering can be considered to be omitted. Thus, behavior preference items with certain high evaluation values can be ensured to be extracted. The match termination screening requirement reflects that the overall hierarchy matches the first behavioral preference until the collected business interaction log is complete. In other words, after the matching of the behavior preference items is completed, the best evaluation value session interaction information in the whole course of the matching record of the behavior preference items is filtered and output, and the best evaluation value session interaction information is usually used as an initial scheme. For general behavior preference items, the best business session interaction information output is selected from the matching whole course.

The PROCESS440 initializes the scratch space of the first behavior preference based on the scratch space of the first behavior preference reaching the filter initiation requirement.

In the embodiment of the invention, when the temporary storage space of the first behavior preference item reaches the screening starting requirement, the screening task is completed, and the initialization of the temporary storage space is convenient for the normal implementation of the screening task of the next user interaction activity information set.

In the embodiment of the invention, for each behavior preference item, whether the recognized quality index in the current user activity data reaches a first quality score or not is firstly estimated, namely a screening condition, then initial session interaction information of the behavior preference item is determined and is input into a temporary storage space, and finally service session interaction information with the best evaluation value recorded in the temporary storage space is determined according to a preset screening starting requirement, so that the accuracy of the subsequent knowledge vector mining is improved while the space utilization rate is improved. Meanwhile, a plurality of information extraction rules are used, the screening requirement of each application environment of each behavior preference item is flexibly set, and the resource utilization rate is improved.

In some independent embodiments, the PROCESS150 "in combination with the upstream and downstream characteristics between the at least two types of behavior preferences and the business session interaction information for each of the behavior preferences, determining the push decision requirement of the set of user interaction information may include the following.

And performing knowledge vector mining and/or element vector mining on the business session interaction information of each behavior preference to obtain the demand prediction information corresponding to the behavior preference by the PROCESS 510.

In the embodiment of the invention, knowledge vector mining (feature extraction) is carried out on behavior preference items related to users through a knowledge vector mining model to obtain knowledge description blocks (feature vectors) with certain dimensionality, and element vector mining is carried out through an element vector mining model to obtain element description blocks such as e-commerce behavior preference element description blocks, VR behavior preference element description blocks and the like; and for other unimportant behavior preference items such as VR service preference items, element vector mining is carried out only through an element vector mining model, so that corresponding element description blocks are obtained.

In the actual implementation process, the behavior preference items of different categories are respectively analyzed. For a virtual mall browsing item, respectively carrying out knowledge vector mining on the service session interaction information which is not lower than two service session interaction information of the virtual mall browsing item to obtain the knowledge description blocks which are not lower than two service session interaction information; splicing the knowledge description blocks which are not lower than two to obtain a linkage descriptor of the virtual mall browsing item; element vector mining is carried out on the business session interaction information with the highest quality index in the business session interaction information to obtain element description blocks of the virtual mall browsing items; and determining the demand prediction information of the virtual mall browsing item based on the linkage descriptor of the virtual mall browsing item and the element description block of the virtual mall browsing item. Therefore, the feature output quality of the descriptors can be improved by vector mining and splicing of the service session interaction information of the plurality of virtual mall browsing items; meanwhile, element vector mining is carried out on the business session interaction information with the best evaluation value, so that the credibility of demand mining prediction can be improved.

For cross-border e-commerce attention items, generally, screening and outputting business session interaction information, directly using a DNN (digital network) to carry out knowledge vector mining of VR (virtual reality) behavior preferences, then using a VR behavior preference element description block model to carry out element vector mining of VR behavior preferences, and then using characteristics of the VR behavior preferences and an element description block as demand prediction information of the cross-border e-commerce attention items. For the VR service preference, feature pyramid networks are used for element vector mining, and then interactive topic element description blocks such as topic viewpoint distribution variables and topic viewpoint contents are used as demand prediction information of the VR service preference. And for MR service preference items, performing mixed reality element vector mining by using an element description block model, and taking the mixed reality element vector mining as demand prediction information of the MR service preference items.

PROCESS520 combines the demand forecast information for the at least two types of behavior preferences in combination with the upstream and downstream features between the at least two types of behavior preferences;

In the embodiment of the invention, the requirement prediction information of each behavior preference item is combined through the upstream and downstream characteristics among different behavior preference items, so that excessive data processing pressure can be avoided.

PROCESS530 determines a push decision requirement for the set of user interaction activity information based on the requirement prediction information for the at least two types of behavioral preferences to complete the combination.

In the embodiment of the invention, after the knowledge descriptors and element description blocks of all kinds of service session interaction information screened and output are mined, the push decision requirement of the user interaction information set is jointly determined based on the requirement prediction information of all behavior preference items.

In the embodiment of the invention, the screened business session interaction information is subjected to big data mining, and the requirement prediction information is integrated by combining the upstream and downstream characteristics among different behavior preference items, so that the requirement is used as the push decision requirement of the user interaction activity information set, and the precision and rationality of the push decision requirement mining and prediction under the complex session environment are improved.

Based on the same or similar inventive concept, please refer to fig. 2, a schematic architecture diagram of an application environment 30 of a user analysis push processing method based on big data is also provided, which includes a user analysis push system 10 and a service user terminal 20 that communicate with each other, where the user analysis push system 10 and the service user terminal 20 implement or partially implement the technical solutions described in the above method embodiments during operation.

Further, there is also provided a readable storage medium having stored thereon a program which when executed by a processor implements the above-described method.

Further, the invention also provides a server, which comprises a processor and a memory; the processor is in communication with the memory, and the processor is configured to read and execute a computer program from the memory to implement the method described above.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other manners. The apparatus and method embodiments described above are merely illustrative, for example, flow diagrams and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

In addition, functional modules in the embodiments of the present invention may be integrated together to form a single part, or each module may exist alone, or two or more modules may be integrated to form a single part.

The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a media service server 10, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a read-only memory (ROM), a random access memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. The user analysis pushing processing method based on big data is characterized by being applied to a user analysis pushing system, and comprises the following steps:

Obtaining a user interaction activity information set meeting the requirement of big data mining, and performing continuous mining analysis on at least two types of behavior preference items in the user interaction activity information set to obtain directional mining information of each behavior preference item;

Determining upstream and downstream characteristics between the no less than two types of behavior preference items by combining the directional mining information of the no less than two types of behavior preference items; determining business session interaction information corresponding to the behavior preference items according to set information extraction rules through directional mining information of the behavior preference items; wherein the information extraction rule is used for guiding the number of the business session interaction information of the same behavior preference;

determining push decision requirements of the user interaction activity information set by combining the upstream and downstream characteristics between the at least two types of behavior preference items and business session interaction information of each behavior preference item;

Wherein the at least two types of behavior preferences include a mainstream behavior preference and an edge behavior preference; the mainstream behavior preference reflects a push demand topic; performing behavior preference recognition on at least two types of behavior preference items of current user activity data in the user interaction activity information set to obtain recognition cores of the behavior preference items, wherein the recognition cores comprise: performing behavior preference identification on the main stream behavior preferences of the current user activity data in the user interaction activity information set to obtain an identification core of the main stream behavior preferences; performing behavior preference identification on the edge behavior preference in the current user activity data based on the current user activity data as selected user activity data to obtain a basic identification core of the edge behavior preference; the selected user activity data is extracted according to a set extraction step length; performing preference matching processing on the edge behavior preference in the current user activity data based on the current user activity data as non-selected user activity data to obtain a candidate identification core of the edge behavior preference; wherein, each current user activity data in the user interaction activity information set contains a digital authentication signature; the performing preference matching processing on the edge behavior preference in the current user activity data based on the current user activity data as the non-selected user activity data to obtain a candidate identification core of the edge behavior preference, including: combining the basic identification core of the edge behavior preference, and adjusting a first preference matching processing model; the base identification core is determined from selected user activity data prior to the digital authentication signature preceding the current user activity data; estimating a distribution variable of the edge behavior preference in the current user activity data through the adjusted first preference matching processing model to obtain a candidate identification core of the edge behavior preference; the matching of the corresponding behavior preference items by combining the identification cores of the behavior preference items in the user interaction activity information set to obtain matching labels and matching records corresponding to the behavior preference items comprises the following steps: loading the identification cores of all the behavior preference items in the user interaction activity information set to a second preference matching processing model to obtain matching labels of the behavior preference items; determining a matching record corresponding to the behavior preference based on the identification cores corresponding to the same behavior preference and the matching labels corresponding to the behavior preference in a group of business interaction logs;

Wherein the at least two types of behavior preferences include: virtual mall browsing items, cross-border e-commerce attention items, VR service preference items, MR service preference items, and hot topic preference items; the determining the upstream and downstream characteristics between the no less than two types of behavior preference items by combining the directional mining information of the no less than two types of behavior preference items comprises the following steps: combining the distributed variable joint analysis results between the recognition cores of the virtual mall browsing items and the recognition cores of the cross-border e-commerce attention items to determine cross-border mall associated information corresponding to the same e-commerce preference items; combining a distribution variable joint analysis result between the recognition core of the cross-border e-commerce attention item and the recognition core of the VR service preference item to determine somatosensory demand description characteristics between the cross-border e-commerce attention item and the VR service preference item; determining a first project participation characteristic or a second project participation characteristic between the hot topic preference item and the MR service preference item by combining the distribution variable of the identification core of each hot topic preference item and the MR service preference item and the matching record of each hot topic preference item and the MR service preference item;

The determining, by the directional mining information of each behavior preference, the service session interaction information corresponding to the behavior preference according to a set information extraction rule, includes: determining initial session interaction information by combining the identification core of the first behavior preference item based on the fact that the merit factor of the first behavior preference item in the current user activity data reaches a first merit score; wherein the first behavior preference is any one of the at least two types of behavior preference; inputting the initial session interaction information into a temporary storage space of the first behavior preference item; based on the temporary storage space of the first behavior preference item reaching a preset screening starting requirement, determining the session interaction information recorded in the temporary storage space as business session interaction information corresponding to the first behavior preference item; the entering the initial session interaction information into the temporary storage space of the first behavior preference item includes: recording the initial session interaction information into the temporary storage space of the first behavior preference item unconditionally on the basis that the number of the session interaction information recorded in the temporary storage space of the first behavior preference item does not reach a storage limit value; the entering the initial session interaction information into the temporary storage space of the first behavior preference item includes: removing the first session interaction information in the temporary storage space on the basis that the number of the session interaction information recorded in the temporary storage space of the first behavior preference reaches the storage limit value; wherein, the goodness coefficient of the first session interaction information is lower than that of the initial session interaction information; inputting the initial session interaction information into a temporary storage space of the first behavior preference item;

Wherein the method further comprises: initializing the temporary storage space of the first behavior preference item based on the temporary storage space of the first behavior preference item reaching the screening starting requirement;

wherein, the preset screening start requirement comprises at least one of the following: screening timeliness requirements, screening intermittent requirements, matching time-consuming screening conditions, good and bad scoring screening requirements and matching termination screening requirements; wherein the screening timeliness requirement reflects that the duration of continuously positioning the first behavior preference reaches a first set duration; the screening intermittent requirement reflects that the gap period matched with the first behavior preference reaches a set gap value; the time-consuming screening condition reflects that the total time length of the first behavior preference item is matched to reach a second set time length; the second set duration is not less than the first set duration; the priority score screening requirement reflects session interaction information that the priority coefficient reaches a second priority score exists in the temporary storage space of the first behavior preference; the second merit score is greater than the first merit score; the matching termination screening requirement reflects that the whole layer matches the first behavior preference until the collected business interaction log is completed;

Wherein the determining the push decision requirement of the user interaction information set by combining the upstream and downstream features between the at least two types of behavior preference items and the business session interaction information of each behavior preference item includes: carrying out knowledge vector mining and/or element vector mining on the business session interaction information of each behavior preference to obtain the demand prediction information corresponding to the behavior preference; combining the demand prediction information of the at least two types of behavior preference items by combining upstream and downstream characteristics between the at least two types of behavior preference items; determining push decision requirements of the user interaction activity information set based on the requirement prediction information of the at least two types of behavior preference items which are combined; the behavior preference items are virtual mall browsing items, and the number of the business session interaction information of the virtual mall browsing items is not less than 2; the knowledge vector mining and/or element vector mining are performed on the business session interaction information of each behavior preference item to obtain the demand prediction information corresponding to the behavior preference item, which comprises the following steps: respectively carrying out knowledge vector mining on the service session interaction information which is not lower than two of the virtual mall browsing items to obtain knowledge description blocks which are not lower than two of the virtual mall browsing items; splicing the knowledge description blocks which are not lower than two to obtain a linkage descriptor of the virtual mall browsing item; element vector mining is carried out on the business session interaction information with the highest quality index in the business session interaction information to obtain element description blocks of the virtual mall browsing items; determining demand prediction information of the virtual mall browsing item based on the linkage descriptors of the virtual mall browsing item and the element description blocks of the virtual mall browsing item;

The method for obtaining the user interaction activity information set meeting the big data mining requirement comprises the following steps:

Acquiring not less than two groups of business interaction logs collected by an information collection unit;

Processing one group of business interaction logs in the two groups of business interaction logs by using a log processing algorithm not lower than two to obtain current user activity data in the corresponding group of business interaction logs;

taking the current user activity data in the service interaction logs which are not lower than two groups as the user interaction activity information set;

The directional mining information comprises an identification core, a matching tag and a matching record of the behavior preference items, and the continuous mining analysis is carried out on at least two types of behavior preference items in the user interaction activity information set to obtain the directional mining information of each behavior preference item, and the method comprises the following steps:

performing behavior preference recognition on at least two types of behavior preference items of current user activity data in the user interaction activity information set to obtain recognition cores of the behavior preference items;

and matching the corresponding behavior preference items by combining the identification cores of the behavior preference items in the user interaction activity information set to obtain matching labels and matching records corresponding to the behavior preference items.

2. A user analysis pushing system, comprising a processor and a memory; the processor is communicatively connected to the memory, and the processor is configured to read the computer program from the memory and execute the computer program to implement the method of claim 1.

3. A server comprising a processor and a memory; the processor is communicatively connected to the memory, and the processor is configured to read the computer program from the memory and execute the computer program to implement the method of claim 1.