CN116934395A

CN116934395A - Feature processing method and device, storage medium and electronic equipment

Info

Publication number: CN116934395A
Application number: CN202310725524.6A
Authority: CN
Inventors: 梅杰
Original assignee: Alipay Hangzhou Information Technology Co Ltd
Current assignee: Alipay Hangzhou Information Technology Co Ltd
Priority date: 2023-06-19
Filing date: 2023-06-19
Publication date: 2023-10-24

Abstract

The specification discloses a feature processing method, a device, a storage medium and an electronic apparatus, wherein the method comprises the following steps: collecting object behavior log data of a release recommendation object, carrying out feature data extraction processing on the object behavior log data based on each feature data extraction template to obtain at least one object feature variable, and carrying out information release recommendation processing on the release recommendation object based on the object feature variable.

Description

Feature processing method and device, storage medium and electronic equipment

Technical Field

The present disclosure relates to the field of computer technologies, and in particular, to a feature processing method, a device, a storage medium, and an electronic apparatus.

Background

In the process of using daily application software, a user generally receives push promotion information sent by a service side, for example, the promotion information may be advertisement information for promoting a certain product or thing. These promotional information is significant to advertisers, so how to promote the effect of delivering and recommending the promotional information has been the focus of attention in the industry.

Disclosure of Invention

The specification provides a feature processing method, a device, a storage medium and electronic equipment, wherein the technical scheme is as follows:

In a first aspect, the present specification provides a feature processing method, the method comprising:

acquiring at least one feature data extraction template;

collecting at least one object behavior log data of a recommended object, and carrying out feature data extraction processing on the object behavior log data based on each feature data extraction template to obtain at least one object feature variable;

and carrying out information release recommendation processing on the release recommendation object based on the object characteristic variable.

In a second aspect, the present specification provides a feature processing apparatus, the apparatus comprising:

the template acquisition module is used for acquiring at least one characteristic data extraction template;

the variable generation module is used for collecting at least one object behavior log data of a recommended object, and carrying out feature data extraction processing on the object behavior log data based on each feature data extraction template to obtain at least one object feature variable;

and the release recommendation module is used for carrying out information release recommendation processing on the release recommendation objects based on the object characteristic variables.

In a third aspect, the present description provides a computer storage medium storing at least one instruction adapted to be loaded by a processor and to perform the method steps of one or more embodiments of the present description.

In a fourth aspect, the present description provides a computer program product storing at least one instruction adapted to be loaded by a processor and to perform the method steps of one or more embodiments of the present description.

In a fifth aspect, the present description provides an electronic device, which may include: a processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and to perform the method steps of one or more embodiments of the present description.

The technical scheme provided by some embodiments of the present specification has the following beneficial effects:

in one or more embodiments of the present disclosure, a service platform acquires at least one feature data extraction template, acquires object behavior log data of at least one release recommendation object, performs feature data extraction processing on the object behavior log data based on each feature data extraction template to obtain at least one object feature variable, and then performs information release recommendation processing on the release recommendation object based on the object feature variable.

Drawings

In order to more clearly illustrate the technical solutions of the present specification or the prior art, the following description will briefly introduce the drawings that are required to be used in the embodiments or the prior art descriptions, it is obvious that the drawings in the following description are only some embodiments of the present specification, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic illustration of a scenario of a feature processing system provided herein;

FIG. 2 is a schematic flow chart of a feature processing method provided in the present specification;

FIG. 3 is a schematic view of a scenario of feature data extraction provided in the present specification;

FIG. 4 is a schematic flow chart of a feature data extraction template determination provided in the present specification;

FIG. 5 is a schematic view of a key character decision tree provided herein;

FIG. 6 is a schematic view of a feature processing apparatus provided in the present specification;

FIG. 7 is a schematic diagram of an electronic device provided in the present specification;

FIG. 8 is a schematic diagram of the architecture of the operating system and user space provided herein;

FIG. 9 is an architecture diagram of the android operating system of FIG. 8;

FIG. 10 is an architecture diagram of the IOS operating system of FIG. 8.

Detailed Description

The following description of the embodiments of the present invention will be made apparent from, and elucidated with reference to, the drawings of the present specification, in which embodiments described are only some, but not all, embodiments of the present specification. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are intended to be within the scope of the present disclosure.

In the description of the present specification, it should be understood that the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. In the description of the present specification, it should be noted that, unless expressly specified and limited otherwise, "comprise" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those listed steps or elements but may include other steps or elements not listed or inherent to such process, method, article, or apparatus. The specific meaning of the terms in this specification will be understood by those of ordinary skill in the art in the light of the specific circumstances. In addition, in the description of the present specification, unless otherwise indicated, "a plurality" means two or more. "and/or", describes an association relationship of an association object, and indicates that there may be three relationships, for example, a and/or B, and may indicate: a exists alone, A and B exist together, and B exists alone. The character "/" generally indicates that the context-dependent object is an "or" relationship.

In the related art, in most of the delivering recommendation scenes, when a user object is searched for the recommendation information, the user behavior data of the user object is generally analyzed to extract an object feature variable, the exposure click rate CTR of at least one (advertisement) information creative corresponding to the recommendation information is estimated based on the object feature variable, a proper information creative is individually searched for based on the exposure click rate, the recommendation information corresponding to the information creative is delivered and pushed to the user object, for example, one or a/357640 (advertisement) information creative of some recommendation information is scored based on the exposure click rate, and the recommendation information is delivered to the user in the order of the score from high to low.

In the foregoing process, the expert-side construction feature engineering is often used to extract the object feature variables, and the object feature variables are extracted and then used to perform model entry on the put-in recommendation model. This is inefficient in development and the human experience may introduce errors into the final training effect of the model.

The present specification is described in detail below with reference to specific examples.

Referring to fig. 1, a schematic view of a feature processing system is provided in the present specification. As shown in fig. 1, the feature processing system may include at least a client cluster and a service platform 100.

The client cluster may include at least one client, as shown in fig. 1, specifically including a client 1 corresponding to a user 1, a client 2 corresponding to a user 2, …, and a client n corresponding to a user n, where n is an integer greater than 0.

Each client in the client cluster may be a communication-enabled electronic device including, but not limited to: wearable devices, handheld devices, personal computers, tablet computers, vehicle-mounted devices, smart phones, computing devices, or other processing devices connected to a wireless modem, etc. Electronic devices in different networks may be called different names, for example: a user equipment, an access terminal, a subscriber unit, a subscriber station, a mobile station, a remote terminal, a mobile device, a user terminal, a wireless communication device, a user agent or user equipment, a cellular telephone, a cordless telephone, a personal digital assistant (personaldigital assistant, PDA), an electronic device in a 5G network or future evolution network, and the like.

The service platform 100 may be a separate server device, such as: rack-mounted, blade, tower-type or cabinet-type server equipment or hardware equipment with stronger computing capacity such as workstations, mainframe computers and the like is adopted; the server cluster may also be a server cluster formed by a plurality of servers, and each server in the server cluster may be formed in a symmetrical manner, wherein each server is functionally equivalent and functionally equivalent in a transaction link, and each server may independently provide services to the outside, and the independent provision of services may be understood as no assistance of another server is needed.

In one or more embodiments of the present disclosure, the service platform 100 may establish a communication connection with at least one client in the client cluster, and complete data interaction during feature processing based on the communication connection, such as online transaction data interaction, for example, the service platform 100 may obtain object behavior log data of a recommended object placed on the client based on a feature processing method of the present disclosure; as another example, the service platform 100 may make information delivery recommendations to a delivery recommendation object on a client.

It should be noted that, the service platform 100 establishes a communication connection with at least one client in the client cluster through a network for interactive communication, where the network may be a wireless network, or may be a wired network, where the wireless network includes, but is not limited to, a cellular network, a wireless local area network, an infrared network, or a bluetooth network, and the wired network includes, but is not limited to, an ethernet network, a universal serial bus (universal serial bus, USB), or a controller area network. In one or more embodiments of the specification, techniques and/or formats including HyperTextMark-up Language (HTML), extensible markup Language (Extensible Markup Language, XML), and the like are used to represent data exchanged over a network (e.g., target compression packets). All or some of the links may also be encrypted using conventional encryption techniques such as secure socket layer (Secure Socket Layer, SSL), transport layer security (Transport LayerSecurity, TLS), virtual private network (Virtual Private Network, VPN), internet protocol security (Internet Protocol Security, IPsec), and the like. In other embodiments, custom and/or dedicated data communication techniques may also be used in place of or in addition to the data communication techniques described above.

The embodiments of the feature processing system provided in the present specification and the feature processing methods in one or more embodiments belong to the same concept, and an execution subject corresponding to the feature processing methods related to one or more embodiments in the present specification may be the service platform 100 described above; the execution subject corresponding to the feature processing method in one or more embodiments of the present disclosure may also be an electronic device corresponding to a client, and specifically determined based on an actual application environment. The implementation process of the embodiment of the feature processing system may be described in detail in the following method embodiments, which are not described herein.

Based on the schematic view of the scenario shown in fig. 1, a feature processing method provided in one or more embodiments of the present disclosure is described in detail below.

Referring to fig. 2, a flow diagram of a feature processing method, which may be implemented in dependence on a computer program and may be run on a feature processing device based on von neumann system, is provided for one or more embodiments of the present description. The computer program may be integrated in the application or may run as a stand-alone tool class application. The feature processing device may be a service platform.

Specifically, the feature processing method comprises the following steps:

S102: acquiring at least one feature data extraction template;

in the information delivery recommendation scene, a plurality of information recommendation creatives are configured for one information recommendation content by the service platform, creative recall is performed by the service platform according to the configuration, and exposure of one information recommendation creative to a platform object is selected in a personalized manner through a delivery recommendation algorithm model according to behavior data such as historical behavior preference of the platform object (such as a platform user), so that the overall exposure click rate is improved.

Furthermore, in order to make personalized recommendation as accurate as possible during information delivery recommendation, at present, a machine learning model (for example, a machine learning model is used as an information delivery model) is mostly utilized to estimate CTR (click rate), and one of the keys for improving the effect of the information delivery model is reasonable selection and use of features.

It should be noted that the machine learning Model according to one or more embodiments of the present disclosure includes, but is not limited to, fitting of one or more of a convolutional neural network (Convolutional Neural Network, CNN) Model, a deep neural network (Deep Neural Network, DNN) Model, a recurrent neural network (Recurrent Neural Networks, RNN), a Model, an embedding (embedding) Model, a gradient lifting decision tree (Gradient Boosting DecisionTree, GBDT) Model, a logistic regression (Logistic Regression, LR) Model, a Diffusion Model (DM) Diffusion Model, and the like.

The information delivery model is characterized in that a multi-layer neural network is utilized to learn the parameter entering characteristics of the model, and the relation among the parameter entering characteristics is actively excavated.

Optionally, in the automation field of the information delivery feature engineering, the feature data extraction template may automatically extract key characters to generate a decision tree model, a father node in the decision tree model is a log attribute, a child node under the father node is a key character token, a leaf node under the child node is a data extraction template of a behavior log, the decision tree model is trained based on the object behavior log of the platform, and after the decision tree model converges, at least one feature data extraction template is obtained based on at least one leaf node corresponding to the decision tree model.

Optionally, in the automation field of the information delivery feature engineering, the clustering model may be relied on to perform cluster analysis processing on the object behavior logs of the platform, and after classifying the object behavior logs belonging to the same class, similar feature conditions in the data are extracted from the object behavior logs of the same class as a template, that is, a data extraction template.

S104: collecting at least one object behavior log data of a recommended object, and carrying out feature data extraction processing on the object behavior log data based on each feature data extraction template to obtain at least one object feature variable;

The release recommendation object can be a user object determined by the service platform based on a release recommendation rule, and the release recommendation rule can be custom setting based on an actual application scene.

Specifically, the service platform may provide functional transactions (such as consuming finance, online shopping, payment, express delivery, etc.) to the outside, the users of one or more clients perform related operations based on the functional transactions provided by the service platform, such as selecting a certain functional transaction, etc., the service platform may record user behavior data of feature dimensions related to the functional transactions under the condition of obtaining user authorization or full authorization of each party, where the user behavior may be classified into functional transaction behavior, user traffic behavior, and user object information, the functional transaction behavior may be, for example, an object query account feature, an object repayment behavior feature, an object data staging behavior feature, etc., the user traffic behavior may be a click behavior, an exposure behavior, an access behavior, etc., and the user object behavior data may be often characterized in the form of object behavior log data.

The feature data extraction template comprises one or more template parameter items, wherein the template parameter items are used for indicating that dynamic parameter information corresponding to the template parameter items is obtained from object behavior log data, and the parameter information corresponding to the same template parameter item is different in the object behavior log data of different object users, namely the dynamic parameter information.

Specifically, the service platform may determine dynamic parameter information corresponding to the template parameter item of the feature data extraction template from the object behavior log data, and generate the object feature variable based on the dynamic parameter information corresponding to the template parameter item of the feature character extraction template.

In a possible implementation manner, as shown in fig. 3, fig. 3 is a schematic view of a scenario of feature data extraction, and for feature variables of log data of any object behavior in an actual application stage, the feature variables go through a log collection stage, a template mining stage, and a variable extraction stage. And obtaining a plurality of characteristic data extraction templates through a log acquisition stage and a template mining stage, then arranging the characteristic data extraction templates online into specific release recommended transactions, such as consumption financial transactions, then extracting dynamic parameter information corresponding to template parameter items from object behavior log data by using one or more characteristic data extraction templates, and generating object characteristic variables based on the dynamic parameter information corresponding to the template parameter items of the characteristic character extraction templates.

Optionally, the template parameter item of the feature data extraction template may be composed of one or more template characters, and parameter information between the template characters in the object behavior log data and information before and after the template characters may be understood as dynamic parameter information;

Optionally, the template parameter item of the feature data extraction template may be composed of one or more template data type items, object features with different dimensions correspond to different template data types, and the object behavior log data obtain the data segment corresponding to the corresponding template parameter item, namely the dynamic parameter information.

S106: and carrying out information release recommendation processing on the release recommendation object based on the object characteristic variable.

In the information delivery recommendation scene, a service platform can perform unified configuration of information recommendation creatives in an information content operation service, a plurality of information recommendation creatives are configured for one information recommendation content to be delivered, the service platform can perform creative recall according to the configuration, after an object feature variable is generated according to a behavior data set feature data extraction template such as historical behavior preference of a platform object (such as a platform user), the object feature variable is used as a model input of a delivery recommendation algorithm model, the object feature variable is input into the delivery recommendation algorithm model, one or more target information recommendation creatives can be selected individually through the delivery recommendation algorithm model, and the service platform performs delivery recommendation of information content corresponding to the specified information recommendation creative for the delivery recommendation object based on the target information recommendation.

The information recommendation creative is an expression form of recommendation information and can be a picture, a text, a video or a combination of pictures and texts.

Currently, in most scenarios, manually constructed feature engineering is often used, with a portion of expert experience added at the feature level. This is inefficient in development and the human experience may introduce errors into the final training effect of the model. Based on the above, one or more embodiments of the present disclosure automatically extract feature variables by pre-configuring a feature data extraction template, automatically determine the feature data extraction template by accumulating data based on the behavior of a released recommended object for a period of time, automatically perform feature analysis and variable extraction by using the feature data extraction template, improve the development efficiency of feature engineering in a released recommended scene, and reduce the expert experience cost and possible errors depending on the feature engineering in the released recommended scene by means of the characteristics of the released recommended model, by using the feature processing method of one or more embodiments of the present disclosure, without the intervention of an expert side, automatically produce one or more feature data extraction templates, and deploy the one or more feature data extraction templates to the on-line released recommended scene, automatically extract object feature variables by means of the feature data extraction template, thereby ensuring the data quality and feature extraction efficiency of feature parameters;

Optionally, in the field of automation of information delivery feature engineering, the feature data extraction template may automatically extract key characters to generate a decision tree model, and obtain at least one feature data extraction template based on at least one leaf node corresponding to the decision tree model, please refer to fig. 4, fig. 4 is a schematic flow chart of determining a feature data extraction template according to one or more embodiments of the present disclosure. Specific:

s202: collecting at least one first object behavior log data;

the first object behavioral log data is template mining sample data used by a feature data extraction template mining stage.

Illustratively, in the template mining stage, the service platform may record user behavior data of feature dimensions related to the functional transaction under the condition that one or more sample user object authorizations or each party is fully authorized, where the user object behavior data is often characterized in the form of first object behavior log data.

S204: constructing an initial key character decision tree aiming at the first object behavior log data;

specifically, each system log of a platform is used as a data source, first object behavior log data are collected in real time, the similarity degree of different first object behavior log data is calculated under the same data attribute, an initial key character decision tree is constructed, decision tree node routing is carried out through a key word token of abstract first object behavior log data, a data extraction template aiming at the first object behavior log data is extracted and stored in a decision tree leaf node, and after the template is stabilized or reaches a set threshold value, the content of the corresponding change of the data extraction template is collected as a user characteristic variable;

Illustratively, the constructing an initial key character decision tree for the first object behavior log data may be:

a2: determining at least one reference data attribute, and respectively acquiring second object behavior log data with the same reference data attribute from the first object behavior log data;

the reference data attribute may be a fit of one or more of log data length, log data generation time, data generation scenario, etc.

Optionally, the reference data attribute may be one or more different preset reference data attributes;

alternatively, the reference data attribute may be determined based on a data attribute (which is a fit of one or more of a log data length, a log data generation time, a data generation scenario, etc.) corresponding to the first set of first object behavior log data used in the initial tree building stage.

In one possible embodiment, the reference data attribute of the log data may be a log data length.

It should be noted that, training of each key character decision tree only uses the object behavior log data of the same data attribute to perform iterative training, and the object behavior log data of different data attributes correspond to different key character decision trees.

Based on this, after determining the reference data attribute of each first object behavior log data, a class of second object behavior log data of the corresponding reference data attribute needs to be determined. FIG. 5 is a schematic view of a key character decision tree, as shown in FIG. 5;

for example, taking a reference data attribute as a data length as an example, after starting, taking each system log of a service platform as a data source, collecting user behavior data in real time to record first object behavior log data, then analyzing and processing the first object behavior log data, firstly determining the reference data attribute of each first object behavior log data as the data length, and dividing second object behavior log data according to the difference of the data lengths.

A4: determining at least one key characteristic character of the second object behavior log data, and determining a first log characteristic character template corresponding to the second object behavior log data;

optionally, key segmentation characters may be set, where the key segmentation characters are used as a key feature character, and when considering that the object behavior log data is recorded, feature type data of different user objects are often distinguished by one or more special characters, where the feature type data refer to data in different feature data structure forms, the key feature characters may be one or more of bracket characters, comma characters, and the like, and parameter information between the key feature characters and information before and after the key feature characters in the object behavior log data may be understood as dynamic parameter information characterizing behavior features of the user of different objects, where the establishment of a template using subsequent feature characters is available.

Optionally, a data parameter item may be set, where the data parameter item is used as a key feature character, and object features in different dimensions in the object behavior log data correspond to different data parameter items, and the object behavior log data is data composed of data information corresponding to the different data parameter items.

It should be noted that the key feature characters may include data parameter items and key segmentation characters.

In a possible implementation, the second object behavior log data may be identified, and at least one key feature character is extracted.

Illustratively, after determining at least one key feature character of the second object behavior log data, determining a first log feature character template corresponding to the second object behavior log data; for example, in an initial tree building stage of an initial key character decision tree, template data abstraction can be performed on the second object behavior log data based on key feature characters, so as to obtain a first log feature character template. The first log feature character template is composed of one or more template parameter items, and the sequence of the parameter items can be reserved among the template parameter items.

A6: and respectively constructing an initial key character decision tree aiming at the second object behavior log data based on the key feature character, the first log feature character template and the reference data attribute.

Specifically, the service platform may respectively construct initial parent nodes for the second object behavior log data based on the reference data attribute;

for example, taking a parameter data attribute as a data length as an example, the data lengths of different object behavior log data are different, based on the fact that as shown in fig. 5, a plurality of initial key character decision trees can be divided based on the different data lengths, and the parameter data attribute such as the data length is taken as an initial father node corresponding to the initial key character decision tree, wherein the value of the initial father node is the corresponding data length, and the data length is 1 and 2.

Specifically, the service platform can respectively construct initial child nodes under the initial parent nodes based on the key feature characters, and respectively construct initial leaf nodes under the initial child nodes based on the first log feature character templates;

for example, as shown in fig. 5, for a key character decision tree of an initial parent node of data length 1, there may be key feature characters-Token 1, key feature characters-Token 2. Dividing the first log feature character template of the corresponding second object behavior log data into initial leaf nodes according to key feature characters-Token, wherein the value of the initial leaf nodes is the first log feature character template.

Specifically, the service platform may generate an initial key character decision tree for the second object behavior log data based on the initial parent node, the initial child node, and the initial leaf node.

For example, as shown in fig. 5, the decision tree of the initial key character is shown by the dashed boxes in fig. 5, and different dashed boxes correspond to different decision trees of the initial key character.

It should be noted that, in the initial tree building stage, the number of initial key character decision trees, the number of initial parent nodes of the decision trees, the number of initial child nodes, and the number of initial leaf nodes are determined by the number of one-round object behavior log data of the service platform, for example, the number of one-round object behavior log data of the service platform is 1, then typically, the number of initial key character decision trees is 1, the number of initial parent nodes of the decision trees is 1, the number of initial child nodes may be one or more, and the number of initial leaf nodes is 1.

S206: determining similarity information corresponding to each piece of first object behavior log data, and performing tree node updating processing on the initial key character decision tree based on the first object behavior log data and the similarity parameters;

It can be understood that after the initial key character decision tree with the number of at least 1 is created, the initial tree creation stage is completed, and then the next batch of first object behavior log data is adopted to perform tree node update processing on the initial key character decision tree, wherein the tree node update processing comprises tree node update processing of the existing decision tree, decision tree new addition processing and other processing links.

Illustratively, in the process of updating the data nodes, the processes of creating and updating the templates are determined by calculating the similarity of the templates. And the data extraction strictness is controlled by controlling the similarity and the depth of the decision tree, and a plurality of data extraction templates can be obtained for deployment after the decision tree is stable, so that the data characteristic extraction is performed after the deployment.

B2: determining a first data attribute of the first object behavior log data, acquiring a reference data attribute on an initial parent node of each initial key character decision tree, determining data attribute similar information based on the first data attribute and the reference data attribute, and updating node data attribute of the initial parent node based on the data attribute similar information;

illustratively, the next batch of first object behavior log data is adopted to update the tree node of the initial key character decision tree, first, the first data attribute of the first object behavior log data, such as one or more of the log data length, the log data generation time, the data generation scene and the like, can be determined, and the first data attribute corresponds to the reference data attribute on the initial father node;

Illustratively, the node data attribute update is to update the node data attribute of the corresponding initial key character decision tree by adopting the first object behavior log data with the same data attribute;

optionally, whether the reference data attribute is matched with the first data attribute or not can be detected, an attribute matching result is obtained, and the attribute matching result is used as data attribute similar information;

for example, for a first object behavior log data i:

the reference data attribute is assumed to be the data length 1 corresponding to the initial key character decision tree a, the data length 2 corresponding to the initial key character decision tree b and the data length 3 corresponding to the initial key character decision tree c;

the method includes the steps that 1, a first data attribute of first object behavior log data i is data length 4, an attribute matching result is that the data attribute of the first object behavior log data i is similar to that of an initial key character decision tree b, and at the moment, data attribute similar information is of an attribute similar type;

the schematic 2, the first data attribute of the first object behavior log data i is data length 4, the attribute matching result is that the data attribute of the first object behavior log data i is dissimilar to the data attribute of any initial key character decision tree, and the data attribute similar information is of an attribute dissimilar type;

Further, if the data attribute similarity information is of an attribute dissimilarity type, performing father node addition processing on the initial father node to obtain a first initial father node, performing data attribute configuration on the first initial father node based on the reference data attribute, and taking the first initial father node as the initial father node to execute at least one reference key feature character for determining the first object behavior log data, thereby obtaining key feature characters of target initial child nodes under the initial father nodes of each initial key character decision tree;

for example, taking fig. 5 as an example, assume that there is no initial key character decision tree with a data length of 4 in fig. 5, and at this time, a new initial key character decision tree is created based on the first object behavior log data i, that is: performing new addition processing on the initial parent node of the new initial key character decision tree based on the first data attribute to obtain a first initial parent node, performing data attribute configuration on the corresponding first initial parent node based on the reference data attribute, and subsequently performing processing by referring to the mode of newly creating the initial key character decision tree, namely: taking the first initial parent node as the initial parent node to execute the step of 'determining at least one reference key feature character of the first object behavior log data, acquiring key feature characters of a target initial child node under the initial parent node of each initial key character decision tree', of one or more embodiments of the present specification;

Further, if the data attribute similarity information is of a similar attribute type, performing data attribute maintenance on the initial parent node to determine a second initial parent node, and using the second initial parent node as the initial parent node to execute the determination of at least one reference key feature character of the first object behavior log data, thereby obtaining key feature characters of a target initial child node under the initial parent node of each initial key character decision tree.

For example, where data attributes are generally highly similar, the initial parent node may be directly taken as the second initial parent node;

for another example, where the data attributes are generally highly similar, a data fit may be performed based on the first data attribute and the reference data attribute, generating a new reference data attribute as the value of the second initial parent node.

B4: determining at least one reference key feature character of the first object behavior log data, acquiring key feature characters of a target initial child node under an initial father node of each initial key character decision tree, determining key character similar information based on the reference key feature characters and the key feature characters, and updating node key characters of the initial child nodes based on the key character similar information;

Calculating the key character similarity of the reference key feature character and the key feature character;

for example, the reference key feature character and the key feature character are input into a first calculation formula, so as to obtain the similarity of the key character, wherein the first calculation formula satisfies the following formula:

wherein, equal () represents the character similarity calculation process, token ₁ Is key character ₂ The first target value is 0, and the second target value is 1 when the first target value is the reference key feature character;

if the similarity of the key characters is a first target value, performing new sub-node adding processing on the initial sub-node to obtain a first initial sub-node, performing key character configuration on the first initial sub-node based on the reference key feature characters, and taking the first initial sub-node as the initial sub-node to execute a second log feature character template for determining the first object behavior log data, thereby obtaining a first log feature character template of the initial sub-node under each initial key character decision tree initial sub-node;

if the similarity of the key characters is a second target value, carrying out key character maintenance on the initial sub-node to determine a second initial sub-node, and taking the second initial sub-node as the initial sub-node to execute a second log feature character template for determining the first object behavior log data, thereby obtaining a first log feature character template of the initial sub-node under each initial key character decision tree initial sub-node;

For example, where the key characters are generally highly similar, the initial child node may be directly used as the second initial child node;

for another example, where the key characters are generally highly similar, a data fit may be performed based on the reference key feature character and the key feature character, generating a new reference key feature character as the value of the second initial child node.

B6: determining a second log feature character template of the first object behavior log data, acquiring a first log feature character template of an initial leaf node under an initial sub-node of each initial key character decision tree, determining feature character template similarity information based on the first log feature character template and the second log feature character template, and updating the node character template of the initial leaf node based on the feature character template similarity information.

Optionally, a feature character template similarity of the first log feature character template and the second log feature character template may be calculated;

for example, the first log feature character template and the second log feature character template may be input into a second calculation formula to obtain feature character template similarity;

updating the node character template of the initial leaf node through the second log feature character template based on the feature character template similarity and the similarity threshold;

The second calculation formula satisfies the following formula:

wherein simSequence represents template similarity calculation processing, sequence ₁ (i) Character model for representing first log featureIth character of character, sequence in board ₂ (i) Representing the ith feature character, equivalent (sequence) ₁ (i)，sequence ₂ (i) I-th feature character in the first log feature character template and the i-th feature character in the second log feature character template. And n is the number of characters corresponding to the first log feature character template and the second log feature character template.

Optionally, performing node character template updating on the initial leaf node based on the feature character template similarity information may be:

if the similarity of the characteristic character templates is smaller than the similarity threshold, performing leaf node newly-added processing on the initial leaf nodes to obtain first initial leaf nodes, and performing character template configuration on the first initial leaf nodes based on the second log characteristic character templates;

and if the similarity of the feature character templates is greater than or equal to the similarity threshold, performing leaf node maintenance processing on the initial leaf nodes to obtain second initial leaf nodes, and performing character template updating on the second initial leaf nodes based on the second log feature character templates.

Illustratively, the threshold value set for the similarity of the feature character templates at the similarity threshold value,

illustratively, in each round of node character template updating process, the similarity of the feature character templates is smaller than a similarity threshold, and at the moment, a character template needs to be newly added when the similarity of the feature character templates is lower, namely, an initial leaf node is firstly subjected to leaf node new addition to obtain a first initial leaf node, and then character template configuration is performed on the first initial leaf node based on a second log feature character template.

Illustratively, in each round of node character template updating process, the similarity of the feature character templates is greater than or equal to a similarity threshold, and at this time, the similarity of the feature character templates is higher, and the character template updating can be performed on the second initial leaf node based on the second log feature character template.

Illustratively, the new addition or the new modification of the leaf node is decided through the similarity calculation of the character templates.

S208: and acquiring the updated key character decision tree of the initial key character decision tree, and determining at least one characteristic data extraction template from the key character decision tree.

In a possible implementation manner, if a first log feature character template of an initial leaf node in the initial key character decision tree is determined to meet a template data stabilization condition, determining that updating processing is completed for the initial key character decision tree, and taking the initial key character decision tree as the updated key character decision tree;

Optionally, the template data stabilizing condition may be that, for any initial leaf node, if no new first log feature character template is added to or an existing first log feature character template is updated in a reference period (for example, 30 minutes), it is determined that an initial key character decision tree i completes an updating process of the initial key character decision tree, at this time, a certain initial key character decision tree i is used as an updated key character decision tree, then values of all leaf nodes in the key character decision tree, that is, log feature character templates are updated as feature data extraction templates, and then the feature data extraction process can be deployed on a line;

optionally, the template data stabilizing condition may be that no new first log feature character template is added to an initial leaf node in a certain data volume content (for example, a batch of first object behavior log data with reference data volume (for example, 1000) or an update of an existing first log feature character template is performed, it is determined that an initial key character decision tree i completes an update process of the initial key character decision tree, at this time, a certain initial key character decision tree i is used as an updated key character decision tree, then values of all leaf nodes in the key character decision tree, that is, log feature character templates, are updated as feature data extraction templates, and then the feature data extraction process may be performed by deploying the feature data extraction template on a line.

In a possible implementation manner, determining a template attribute parameter of a first log feature character template of an initial leaf node in the initial key character decision tree, if the template attribute parameter meets a template parameter threshold, determining that updating processing is completed for the initial key character decision tree, and taking the initial key character decision tree as the updated key character decision tree.

The template attribute parameters can be the number of first log feature character templates under the same initial leaf node, the number of feature characters contained in the templates, the template data length and the like, the number of the first log feature character templates characterizes the depth of a decision tree to a certain extent, and the severity of data extraction is controlled based on the depth of the decision tree;

the template parameter threshold is a threshold value set for a template attribute parameter, when the template attribute parameter of a certain initial key character decision tree i meets the template parameter threshold value (such as being greater than or equal to the template parameter threshold value), the certain initial key character decision tree i is determined to complete the updating process of the initial key character decision tree, at the moment, the certain initial key character decision tree i is used as an updated key character decision tree, then the values of all leaf nodes in the key character decision tree, namely, the log feature character template, are updated as feature data extraction templates, and then the feature data extraction processing can be carried out by deploying the feature data extraction templates on a line.

It should be noted that, the updating process of the initial key character decision tree is completed in a certain initial key character decision tree i, and the rest of non-initial key character decision trees i need to be continuously updated until each room key character decision tree completes the updating process.

In one or more embodiments of the present disclosure, feature analysis is automatically performed by pre-configuring a feature data extraction template, so as to extract object feature variables, thereby improving the development efficiency of feature engineering in a release recommendation scene, and subsequently performing information release recommendation by means of the object feature variables, reducing the expert experience cost dependent on the feature engineering in the release recommendation scene, and avoiding errors caused by the dependence on the expert experience. .

The feature processing apparatus provided in this specification will be described in detail with reference to fig. 6. The feature processing apparatus shown in fig. 6 is used to execute the method of the embodiment shown in fig. 1 to 5 of the present specification, and for convenience of explanation, only the portion relevant to the present specification is shown, and specific technical details are not disclosed, please refer to the embodiment shown in fig. 1 to 5 of the present specification.

Referring to fig. 6, a schematic structural diagram of the feature processing apparatus of the present specification is shown. The feature processing apparatus 1 may be implemented as all or part of an electronic device by software, hardware or a combination of both. According to some embodiments, the feature processing apparatus 1 includes a template acquisition module 11, a variable generation module 12, and a delivery recommendation module 13, specifically configured to:

a template acquisition module 11 for acquiring at least one feature data extraction template;

the variable generating module 12 is configured to collect object behavior log data of at least one recommended object, and perform feature data extraction processing on the object behavior log data based on each feature data extraction template to obtain at least one object feature variable;

and the release recommendation module 13 is used for carrying out information release recommendation processing on the release recommendation objects based on the object characteristic variables.

Optionally, the template obtaining module 11 is specifically configured to:

collecting at least one first object behavior log data;

constructing an initial key character decision tree aiming at the first object behavior log data, and determining similarity information corresponding to each first object behavior log data;

Performing tree node updating processing on the initial key character decision tree based on the first object behavior log data and the similarity parameter;

and acquiring the updated key character decision tree of the initial key character decision tree, and determining at least one characteristic data extraction template from the key character decision tree.

Optionally, the template obtaining module 11 is specifically configured to:

determining at least one reference data attribute, and respectively acquiring second object behavior log data with the same reference data attribute from the first object behavior log data;

determining at least one key characteristic character of the second object behavior log data, and determining a first log characteristic character template corresponding to the second object behavior log data;

and respectively constructing an initial key character decision tree aiming at the second object behavior log data based on the key feature character, the first log feature character template and the reference data attribute.

Optionally, the template obtaining module 11 is specifically configured to:

respectively constructing initial father nodes aiming at the second object behavior log data based on the reference data attributes;

Respectively constructing initial child nodes under the initial parent nodes based on the key feature characters, and respectively constructing initial leaf nodes under the initial child nodes based on the first log feature character templates;

an initial key character decision tree for the second object behavior log data is generated based on the initial parent node, the initial child node, and the initial leaf node.

Optionally, the template obtaining module 11 is specifically configured to:

determining a first data attribute of the first object behavior log data, acquiring a reference data attribute on an initial parent node of each initial key character decision tree, determining data attribute similar information based on the first data attribute and the reference data attribute, and updating node data attribute of the initial parent node based on the data attribute similar information;

determining at least one reference key feature character of the first object behavior log data, acquiring key feature characters of a target initial child node under an initial father node of each initial key character decision tree, determining key character similar information based on the reference key feature characters and the key feature characters, and updating node key characters of the initial child nodes based on the key character similar information;

Determining a second log feature character template of the first object behavior log data, acquiring a first log feature character template of an initial leaf node under an initial sub-node of each initial key character decision tree, determining feature character template similarity information based on the first log feature character template and the second log feature character template, and updating the node character template of the initial leaf node based on the feature character template similarity information.

Optionally, the template obtaining module 11 is specifically configured to:

detecting whether the reference data attribute is matched with the first data attribute or not to obtain an attribute matching result, and taking the attribute matching result as data attribute similar information;

if the data attribute similarity information is of an attribute dissimilarity type, performing father node addition processing on the initial father node to obtain a first initial father node, performing data attribute configuration on the first initial father node based on the reference data attribute, using the first initial father node as the initial father node to execute at least one reference key feature character for determining the first object behavior log data, and obtaining key feature characters of target initial child nodes under the initial father nodes of each initial key character decision tree;

And if the data attribute similarity information is of an attribute similarity type, carrying out data attribute maintenance on the initial parent node to determine a second initial parent node, and taking the second initial parent node as the initial parent node to execute the determination of at least one reference key characteristic character of the first object behavior log data to obtain key characteristic characters of a target initial child node under the initial parent node of each initial key character decision tree.

Optionally, the template obtaining module 11 is specifically configured to: calculating the similarity of the reference key feature character and the key character of the key feature character;

And if the similarity of the key characters is a second target value, carrying out key character maintenance on the initial sub-node to determine a second initial sub-node, taking the second initial sub-node as the initial sub-node to execute a second log feature character template for determining the first object behavior log data, and obtaining a first log feature character template of the initial sub-node under each initial key character decision tree initial sub-node.

Optionally, the template obtaining module 11 is specifically configured to:

calculating the similarity of the feature character templates of the first log feature character template and the second log feature character template;

and updating the node character template of the initial leaf node through the second log characteristic character template based on the characteristic character template similarity and the similarity threshold.

Optionally, the template obtaining module 11 is specifically configured to:

determining that a first log feature character template of an initial leaf node in the initial key character decision tree meets a template data stabilization condition, and determining that updating processing is completed for the initial key character decision tree, wherein the initial key character decision tree is used as the updated key character decision tree; or alternatively, the first and second heat exchangers may be,

determining a template attribute parameter of a first log feature character template of an initial leaf node in the initial key character decision tree, if the template attribute parameter meets a template parameter threshold, determining that updating processing is completed for the initial key character decision tree, and taking the initial key character decision tree as the updated key character decision tree.

Optionally, the variable generating module 12 is specifically configured to:

and determining dynamic parameter information corresponding to the template parameter items of the feature data extraction template from the object behavior log data, and generating object feature variables based on the dynamic parameter information corresponding to the template parameter items of the key feature character extraction template.

It should be noted that, in the feature processing apparatus provided in the foregoing embodiment, when executing the feature processing method, only the division of the foregoing functional modules is used as an example, in practical application, the foregoing functional allocation may be performed by different functional modules, that is, the internal structure of the device is divided into different functional modules, so as to complete all or part of the functions described above. In addition, the feature processing device and the feature processing method provided in the foregoing embodiments belong to the same concept, which embody detailed implementation procedures in the method embodiments, and are not described herein again.

The foregoing description is provided for the purpose of illustration only and does not represent the advantages or disadvantages of the embodiments.

The present disclosure further provides a computer storage medium, where the computer storage medium may store a plurality of instructions, where the instructions are adapted to be loaded by a processor and execute the feature processing method according to the embodiment shown in fig. 1 to 5, and the specific execution process may refer to the specific description of the embodiment shown in fig. 1 to 5, which is not repeated herein.

The present disclosure further provides a computer program product, where at least one instruction is stored, where the at least one instruction is loaded by the processor and executed by the processor, where the specific execution process may refer to the specific description of the embodiment shown in fig. 1 to 5, and details are not repeated herein.

Referring to fig. 7, a block diagram of an electronic device according to an exemplary embodiment of the present disclosure is shown. The electronic device in this specification may include one or more of the following: processor 110, memory 120, input device 130, output device 140, and bus 150. The processor 110, the memory 120, the input device 130, and the output device 140 may be connected by a bus 150.

Processor 110 may include one or more processing cores. The processor 110 utilizes various interfaces and lines to connect various portions of the overall electronic device, perform various functions of the electronic device 100, and process data by executing or executing instructions, programs, code sets, or instruction sets stored in the memory 120, and invoking data stored in the memory 120. Alternatively, the processor 110 may be implemented in at least one hardware form of digital signal processing (digital signal processing, DSP), field-programmable gate array (field-programmable gatearray, FPGA), programmable logic array (programmable logic Array, PLA). The processor 110 may integrate one or a combination of several of a central processor (central processing unit, CPU), an image processor (graphics processing unit, GPU), and a modem, etc. The CPU mainly processes an operating system, a user interface, an application program and the like; the GPU is used for being responsible for rendering and drawing of display content; the modem is used to handle wireless communications. It will be appreciated that the modem may not be integrated into the processor 110 and may be implemented solely by a single communication chip.

The memory 120 may include a random access memory (random Access Memory, RAM) or a read-only memory (ROM). Optionally, the memory 120 includes a non-transitory computer readable medium (non-transitory computer-readable storage medium). Memory 120 may be used to store instructions, programs, code, sets of codes, or sets of instructions. The memory 120 may include a stored program area and a stored data area, wherein the stored program area may store instructions for implementing an operating system, which may be an Android (Android) system, including an Android system-based deep development system, an IOS system developed by apple corporation, including an IOS system-based deep development system, or other systems, instructions for implementing at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing various method embodiments described below, and the like. The storage data area may also store data created by the electronic device in use, such as phonebooks, audiovisual data, chat log data, and the like.

Referring to FIG. 8, the memory 120 may be divided into an operating system space in which the operating system is running and a user space in which native and third party applications are running. In order to ensure that different third party application programs can achieve better operation effects, the operating system allocates corresponding system resources for the different third party application programs. However, the requirements of different application scenarios in the same third party application program on system resources are different, for example, under the local resource loading scenario, the third party application program has higher requirement on the disk reading speed; in the animation rendering scene, the third party application program has higher requirements on the GPU performance. The operating system and the third party application program are mutually independent, and the operating system often cannot timely sense the current application scene of the third party application program, so that the operating system cannot perform targeted system resource adaptation according to the specific application scene of the third party application program.

In order to enable the operating system to distinguish specific application scenes of the third-party application program, data communication between the third-party application program and the operating system needs to be communicated, so that the operating system can acquire current scene information of the third-party application program at any time, and targeted system resource adaptation is performed based on the current scene.

Taking an operating system as an Android system as an example, as shown in fig. 9, a program and data stored in the memory 120 may be stored in the memory 120 with a Linux kernel layer 320, a system runtime library layer 340, an application framework layer 360 and an application layer 380, where the Linux kernel layer 320, the system runtime library layer 340 and the application framework layer 360 belong to an operating system space, and the application layer 380 belongs to a user space. The Linux kernel layer 320 provides the underlying drivers for various hardware of the electronic device, such as display drivers, audio drivers, camera drivers, bluetooth drivers, wi-Fi drivers, power management, and the like. The system runtime layer 340 provides the main feature support for the Android system through some C/c++ libraries. For example, the SQLite library provides support for databases, the OpenGL/ES library provides support for 3D graphics, the Webkit library provides support for browser kernels, and the like. Also provided in the system runtime library layer 340 is a An Zhuoyun runtime library (Android run) which provides mainly some core libraries that can allow developers to write Android applications using the Java language. The application framework layer 360 provides various APIs that may be used in building applications, which developers can also build their own applications by using, for example, campaign management, window management, view management, notification management, content provider, package management, call management, resource management, location management. At least one application program is running in the application layer 380, and these application programs may be native application programs of the operating system, such as a contact program, a short message program, a clock program, a camera application, etc.; and may also be a third party application developed by a third party developer, such as a game-like application, instant messaging program, photo beautification program, etc.

Taking an operating system as an IOS system as an example, the programs and data stored in the memory 120 are shown in fig. 10, the IOS system includes: core operating system layer 420 (Core OS layer), core service layer 440 (CoreServices layer), media layer 460 (Media layer), and touchable layer 480 (Cocoa Touch Layer). The core operating system layer 420 includes an operating system kernel, drivers, and underlying program frameworks that provide more hardware-like functionality for use by the program frameworks at the core services layer 440. The core services layer 440 provides system services and/or program frameworks required by the application, such as a Foundation (Foundation) framework, an account framework, an advertisement framework, a data storage framework, a network connection framework, a geographic location framework, a sports framework, and the like. The media layer 460 provides an interface for applications related to audiovisual aspects, such as a graphics-image related interface, an audio technology related interface, a video technology related interface, an audio video transmission technology wireless play (AirPlay) interface, and so forth. The touchable layer 480 provides various commonly used interface-related frameworks for application development, with the touchable layer 480 being responsible for user touch interactions on the electronic device. Such as a local notification service, a remote push service, an advertisement framework, a game tool framework, a message User Interface (UI) framework, a User Interface UIKit framework, a map framework, and so forth.

Among the frameworks illustrated in fig. 10, frameworks related to most applications include, but are not limited to: the infrastructure in core services layer 440 and the UIKit framework in touchable layer 480. The infrastructure provides many basic object classes and data types, providing the most basic system services for all applications, independent of the UI. While the class provided by the UIKit framework is a basic UI class library for creating touch-based user interfaces, iOS applications can provide UIs based on the UIKit framework, so it provides the infrastructure for applications to build user interfaces, draw, process and user interaction events, respond to gestures, and so on.

The manner and principle of implementing data communication between the third party application program and the operating system in the IOS system may refer to the Android system, and this description is not repeated here.

The input device 130 is configured to receive input instructions or data, and the input device 130 includes, but is not limited to, a keyboard, a mouse, a camera, a microphone, or a touch device. The output device 140 is used to output instructions or data, and the output device 140 includes, but is not limited to, a display device, a speaker, and the like. In one example, the input device 130 and the output device 140 may be combined, and the input device 130 and the output device 140 are a touch display screen for receiving a touch operation thereon or thereabout by a user using a finger, a touch pen, or any other suitable object, and displaying a user interface of each application program. Touch display screens are typically provided on the front panel of an electronic device. The touch display screen may be designed as a full screen, a curved screen, or a contoured screen. The touch display screen can also be designed to be a combination of a full screen and a curved screen, and a combination of a special-shaped screen and a curved screen is not limited in this specification.

In addition, those skilled in the art will appreciate that the configuration of the electronic device shown in the above-described figures does not constitute a limitation of the electronic device, and the electronic device may include more or less components than illustrated, or may combine certain components, or may have a different arrangement of components. For example, the electronic device further includes components such as a radio frequency circuit, an input unit, a sensor, an audio circuit, a wireless fidelity (wireless fidelity, wiFi) module, a power supply, and a bluetooth module, which are not described herein.

In this specification, the execution subject of each step may be the electronic device described above. Optionally, the execution subject of each step is an operating system of the electronic device. The operating system may be an android system, an IOS system, or other operating systems, which is not limited in this specification.

The electronic device of the present specification may further have a display device mounted thereon, and the display device may be various devices capable of realizing a display function, for example: cathode ray tube displays (cathode ray tubedisplay, CR), light-emitting diode displays (light-emitting diode display, LED), electronic ink screens, liquid crystal displays (liquid crystal display, LCD), plasma display panels (plasmadisplay panel, PDP), and the like. A user may utilize a display device on electronic device 101 to view displayed text, images, video, etc. The electronic device may be a smart phone, a tablet computer, a gaming device, an AR (Augmented Reality ) device, an automobile, a data storage device, an audio playing device, a video playing device, a notebook, a desktop computing device, a wearable device such as an electronic watch, electronic glasses, an electronic helmet, an electronic bracelet, an electronic necklace, an electronic article of clothing, etc.

In the electronic device shown in fig. 7, the processor 110 may be configured to call an application program stored in the memory 120, and specifically perform the following operations:

acquiring at least one feature data extraction template;

In one embodiment, the processor 110, in executing the acquiring at least one feature data extraction template, performs the steps of:

collecting at least one first object behavior log data;

In one embodiment, the processor 110, in executing the initial key character decision tree that builds the log data for the first object behavior, performs the steps of:

In one embodiment, the processor 110, in executing the initial key character decision tree for the second object behavior log data based on the key feature character, the first log feature character template, and the reference data attribute, respectively, performs the steps of:

In one embodiment, the processor 110 performs the tree node update process on the initial key character decision tree based on the first object behavior log data and the similarity parameter after executing the determining the similarity information corresponding to each of the first object behavior log data, and performs the following steps:

In one embodiment, the processor 110, when executing the determining data attribute similarity information based on the first data attribute and the reference data attribute, performs node data attribute update on the initial parent node based on the data attribute similarity information, and performs the following steps:

In one embodiment, the processor 110, when executing the determining key character similarity information based on the reference key feature character and the key feature character, performs node key character update on the initial sub-node based on the key character similarity information, and performs the following steps:

calculating the similarity of the reference key feature character and the key character of the key feature character;

In one embodiment, the processor 110, when executing the determining feature character template similarity information based on the first log feature character template and the second log feature character template, performs node character template updating on the initial leaf node based on the feature character template similarity information, and executes the following steps:

In one embodiment, the processor 110 performs the following steps in performing the node character template update on the initial leaf node through the second log feature character template based on the feature character template similarity and a similarity threshold:

In one embodiment, the processor 110, upon executing the key character decision tree updated from the initial key character decision tree, determines at least one feature data extraction template from the key character decision tree, performs the steps of:

In one embodiment, the processor 110 performs the feature data extraction process on the object behavior log data based on each of the feature data extraction templates to obtain at least one object feature variable, and performs the following steps:

Those skilled in the art will appreciate that implementing all or part of the above-described methods in accordance with the embodiments may be accomplished by way of a computer program stored on a computer readable storage medium, which when executed may comprise the steps of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a read-only memory, a random access memory, or the like.

It should be noted that, information (including but not limited to user equipment information, user personal information, etc.), data (including but not limited to data for analysis, stored data, presented data, etc.), and signals according to the embodiments of the present disclosure are all authorized by the user or are fully authorized by the parties, and the collection, use, and processing of relevant data is required to comply with relevant laws and regulations and standards of relevant countries and regions. For example, object behavior log data and the like referred to in the present specification are all acquired with sufficient authorization.

The foregoing disclosure is only illustrative of the preferred embodiments of the present invention and is not to be construed as limiting the scope of the claims, which follow the meaning of the claims of the present invention.

Claims

1. A method of feature processing, the method comprising:

acquiring at least one feature data extraction template;

2. The method of claim 1, the obtaining at least one feature data extraction template comprising:

collecting at least one first object behavior log data;

3. The method of claim 2, the constructing an initial key character decision tree for the first object behavioral log data, comprising:

4. The method of claim 3, the constructing an initial key character decision tree for the second object behavioral log data based on the key feature character, the first log feature character template, and the reference data attribute, respectively, comprising:

5. The method of claim 2, wherein determining similarity information corresponding to each of the first object behavior log data, and performing tree node update processing on the initial key character decision tree based on the first object behavior log data and the similarity parameter, comprises:

6. The method of claim 5, the determining data attribute similarity information based on the first data attribute and the reference data attribute, the node data attribute updating the initial parent node based on the data attribute similarity information, comprising:

7. The method of claim 5, the determining key character similarity information based on the reference key feature character and the key feature character, the updating node key characters for the initial child node based on the key character similarity information, comprising:

8. The method of claim 5, the determining feature character template similarity information based on the first and second log feature character templates, the node character template updating the initial leaf node based on the feature character template similarity information, comprising:

9. The method of claim 8, the node character template updating of the initial leaf node by the second log feature character template based on the feature character template similarity and a similarity threshold, comprising:

10. The method of claim 2, the obtaining the updated key character decision tree of the initial key character decision tree, determining at least one feature data extraction template from the key character decision tree, comprising:

11. The method according to claim 1, wherein the performing feature data extraction processing on the object behavior log data based on each feature data extraction template to obtain at least one object feature variable includes:

12. A feature processing apparatus for use with a service platform, the apparatus comprising:

13. A computer storage medium storing a plurality of instructions adapted to be loaded by a processor and to perform the method steps of any one of claims 1 to 11.

14. A computer program product storing at least one instruction for loading by a processor and performing the method steps of any one of claims 1 to 11.

15. An electronic device, comprising: a processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and to perform the method steps of any of claims 1-11.