Object recognition method, device, equipment and storage medium

Abstract

The present disclosure relates to an object recognition method, apparatus, device and storage medium, the method comprising: obtaining target behavior features of a target object in a target application; obtaining sample behavior features of a sample object of a second business type of the target application; the sample object and the target object are objects in the same field; the sample behavior features and the target behavior features are features of the same type; based on the similarity between the target behavior features and the sample behavior features, determining the association information between the target object and the second business type; based on the association information, obtaining an object recognition result for the target object. The present disclosure determines the association information between the target object and the second business type based on the target behavior features and the sample behavior features; thereby determining the object recognition result of the target object, and improving the accuracy of determining the matching degree between the target object and the second business type.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to an object recognition method, device, equipment and storage medium. Background Art

[0002] With the booming development of short videos and live streaming e-commerce, live streaming shopping is becoming a new channel for consumers to shop online. The platform continues to optimize product functions and services, promote the development and maturity of live streaming e-commerce, and gradually form a content e-commerce ecosystem with influencers bringing goods as the main model. However, as the business deepens, live streaming e-commerce is also recruiting supply chains, product merchants, brand merchants, and service providers. For e-commerce investment promotion business, major e-commerce companies are actively looking for potential merchants to settle in. How to detect the potential willingness of merchants to settle in in real time is of great significance to the growth of e-commerce investment promotion business.

[0003] In the related technologies, most of the current e-commerce companies mainly collect merchants' willingness to settle in manually, mainly through outbound telephone inquiries, and are unable to detect the willingness of all merchants to settle in real time.

[0004] Therefore, it is necessary to provide an object recognition method, apparatus, device and storage medium to improve the accuracy of target object recognition results, thereby improving the accuracy of determining the match between the target object and the second business type. Summary of the invention

[0005] The present disclosure provides an object recognition method, apparatus, device and storage medium to at least solve the problem in the related art that the matching degree between the target object and the second business type cannot be accurately determined. The technical solution of the present disclosure is as follows:

[0006] According to a first aspect of an embodiment of the present disclosure, there is provided an object recognition method, comprising:

[0007] Acquire a target behavior feature of a target object in a target application; the target behavior feature is a feature related to an operation behavior of the target object in at least one first business type of the target application;

[0008] Acquire a sample behavior feature of a sample object of a second business type of the target application; the sample object and the target object are objects of the same field; the sample behavior feature and the target behavior feature are features of the same type;

[0009] Determining, based on the similarity between the target behavior feature and the sample behavior feature, the association information between the target object and the second service type;

[0010] Based on the association degree information, an object recognition result of the target object is obtained.

[0011] In an exemplary embodiment, before determining the association information between the target object and the second service type based on the similarity between the target behavior feature and the sample behavior feature, the method further includes:

[0012] Normalizing the target behavior feature and the sample behavior feature to obtain a normalized target behavior feature and a normalized sample behavior feature;

[0013] Converting the normalized target behavior feature into a first feature vector, and converting the normalized sample behavior feature into a second feature vector;

[0014] Calculating the similarity between the first feature vector and the second feature vector to obtain a similarity calculation result;

[0015] The similarity calculation result is used as the similarity between the target behavior feature and the sample behavior feature.

[0016] In an exemplary embodiment, there are at least two target behavior features, and calculating the similarity between the first feature vector and the second feature vector to obtain a similarity calculation result includes:

[0017] Calculating the similarity between the first feature vector and the second feature vector corresponding to each target behavior feature to obtain a similarity calculation result corresponding to each target behavior feature;

[0018] Determine the target weight corresponding to each target behavior feature;

[0019] Determine a comprehensive similarity calculation result based on the similarity calculation result corresponding to each target behavior feature and the target weight;

[0020] The using the similarity calculation result as the similarity between the target behavior feature and the sample behavior feature includes:

[0021] The comprehensive similarity calculation result is used as the similarity between the target behavior feature and the sample behavior feature.

[0022] In an exemplary embodiment, determining the comprehensive similarity calculation result based on the similarity calculation result corresponding to each target behavior feature and the target weight includes:

[0023] Calculate the product of the similarity calculation result corresponding to each target behavior feature and the target weight to obtain the similarity calculation result of each target behavior feature;

[0024] The sum of the similarity calculation results of each target behavior feature is calculated to obtain the comprehensive similarity calculation result.

[0025] In an exemplary embodiment, determining the target weight corresponding to each target behavior feature includes:

[0026] Obtaining a feature set corresponding to each target behavior feature within a preset time period;

[0027] Based on each feature in each feature set, determine a feature standard deviation and a feature average corresponding to each feature set;

[0028] Determine the coefficient of variation of each target behavior feature based on the feature standard deviation and feature average corresponding to each feature set;

[0029] Based on the coefficient of variation of each target behavior feature, a target weight corresponding to each target behavior feature is determined.

[0030] In an exemplary embodiment, determining the target weight corresponding to each target behavior feature based on the coefficient of variation of each target behavior feature includes:

[0031] Based on the coefficient of variation of each target behavior feature, calculating the sum of the coefficients of variation of each target behavior feature;

[0032] Calculating the ratio of the coefficient of variation of each target behavior feature to the sum of the coefficients of variation to obtain the coefficient of variation ratio of each target behavior feature;

[0033] The coefficient of variation ratio of each target behavior feature is used as the target weight corresponding to each target behavior feature.

[0034] In an exemplary embodiment, there are at least two sample objects, and determining the association information between the target object and the second service type based on the similarity between the target behavior feature and the sample behavior feature includes:

[0035] Calculating the similarity between the target behavior feature and each sample behavior feature to obtain a similarity calculation result corresponding to each sample object;

[0036] Based on the similarity calculation results corresponding to the sample objects, the association information between the target object and the second business type is determined.

[0037] In an exemplary implementation, the determining, based on the similarity calculation results corresponding to the sample objects, the association information between the target object and the second business type includes:

[0038] If the similarity calculation results corresponding to the sample objects are all greater than the preset similarity threshold, it is determined that the target object has a high correlation with the second business type;

[0039] If the similarity calculation result corresponding to any sample object is less than or equal to the preset similarity threshold, it is determined that the correlation between the target object and the second business type is low.

[0040] In an exemplary embodiment, before obtaining the sample behavior characteristics of the sample object of the second business type of the target application, the method further includes:

[0041] Acquire a candidate object set in a second business type of the target application;

[0042] Acquire business data of a second business type of each candidate object in the candidate object set;

[0043] The candidate objects whose business data are within the target interval are determined as the sample objects.

[0044] In an exemplary implementation, determining the candidate objects of the business data within the target interval as the sample objects includes:

[0045] Determine candidate objects whose business data are within the target interval as objects to be screened, wherein the number of the objects to be screened is at least two;

[0046] Obtain the resources of the target object and the number of concerned objects;

[0047] Determining a target domain of the target object based on the resources of the target object;

[0048] Based on the target field of the target object and the number of the objects of interest, sample objects having the same target field and matching the number of the objects of interest are determined from at least two objects to be screened.

[0049] In an exemplary embodiment, the method further comprises:

[0050] Based on the association information, sharing information is sent to the target object, where the sharing information includes an identifier of the second service type.

[0051] In an exemplary embodiment, the target behavior feature includes at least one of the following:

[0052] The target object's behavior duration, the number of multimedia information released, the exposure duration of multimedia information, the delivery value of virtual resources, the recovery value of virtual resources, the growth number of attention objects and the number of occurrences of target content.

[0053] According to a second aspect of an embodiment of the present disclosure, there is provided an object recognition device, including:

[0054] A target behavior feature acquisition module is configured to execute acquisition of a target behavior feature of a target object in a target application; the target behavior feature is a feature related to an operation behavior of the target object in at least one first business type of the target application;

[0055] A sample behavior feature acquisition module is configured to execute acquisition of sample behavior features of a sample object of a second business type of the target application; the sample object and the target object are objects of the same field; the sample behavior features and the target behavior features are features of the same type;

[0056] A correlation information determination module, configured to determine the correlation information between the target object and the second business type based on the similarity between the target behavior feature and the sample behavior feature;

[0057] The object recognition result determination module is configured to obtain an object recognition result of the target object based on the association degree information.

[0058] In an exemplary embodiment, the device further comprises:

[0059] A normalization processing module is configured to perform normalization processing on the target behavior feature and the sample behavior feature to obtain a normalized target behavior feature and a normalized sample behavior feature;

[0060] A vector conversion module, configured to convert the normalized target behavior feature into a first feature vector and convert the normalized sample behavior feature into a second feature vector;

[0061] a similarity calculation result determination module, configured to calculate the similarity between the first feature vector and the second feature vector to obtain a similarity calculation result;

[0062] The behavior feature similarity determination module is configured to use the similarity calculation result as the similarity between the target behavior feature and the sample behavior feature.

[0063] In an exemplary embodiment, the target behavior features are at least two, and the similarity calculation result determination module includes:

[0064] A similarity calculation result determination unit is configured to calculate the similarity between the first feature vector and the second feature vector corresponding to each target behavior feature to obtain a similarity calculation result corresponding to each target behavior feature;

[0065] A target weight determination unit, configured to determine the target weight corresponding to each target behavior feature;

[0066] A comprehensive similarity calculation result determination unit is configured to determine a comprehensive similarity calculation result based on the similarity calculation result corresponding to each target behavior feature and the target weight;

[0067] The behavior feature similarity determination module includes:

[0068] The behavior feature similarity determination unit is configured to use the comprehensive similarity calculation result as the similarity between the target behavior feature and the sample behavior feature.

[0069] In an exemplary embodiment, the comprehensive similarity calculation result determination unit includes:

[0070] A product calculation subunit is configured to calculate the product of the similarity calculation result corresponding to each target behavior feature and the target weight to obtain the similarity calculation result of each target behavior feature;

[0071] The comprehensive similarity calculation result determination subunit is configured to calculate the sum of the similarity calculation results of each target behavior feature to obtain the comprehensive similarity calculation result.

[0072] In an exemplary embodiment, the target weight determination unit includes:

[0073] A feature set acquisition subunit is configured to acquire a feature set corresponding to each target behavior feature within a preset time period;

[0074] A feature average value determination subunit is configured to determine a feature standard deviation and a feature average value corresponding to each feature set based on each feature in each feature set;

[0075] A coefficient of variation determination subunit, configured to determine the coefficient of variation of each target behavior feature based on a feature standard deviation and a feature average corresponding to each feature set;

[0076] The target weight determination subunit is configured to determine the target weight corresponding to each target behavior feature based on the coefficient of variation of each target behavior feature.

[0077] In an exemplary embodiment, the target weight determination subunit includes:

[0078] A coefficient of variation and calculation subunit, configured to calculate the sum of the coefficients of variation of each target behavior feature based on the coefficient of variation of each target behavior feature;

[0079] A coefficient of variation ratio determination subunit is configured to calculate the ratio of the coefficient of variation of each target behavior feature to the sum of the coefficients of variation to obtain the coefficient of variation ratio of each target behavior feature;

[0080] The feature weight determination subunit is configured to use the coefficient of variation ratio of each target behavior feature as the target weight corresponding to each target behavior feature.

[0081] In an exemplary embodiment, there are at least two sample objects, and the association information determination module includes:

[0082] A similarity calculation result determination unit is configured to calculate the similarity between the target behavior feature and each sample behavior feature to obtain a similarity calculation result corresponding to each sample object;

[0083] The association information determining unit is configured to determine the association information between the target object and the second business type based on the similarity calculation results corresponding to the sample objects.

[0084] In an exemplary embodiment, the association information determination unit includes:

[0085] A first relevance determination subunit is configured to determine that the relevance between the target object and the second business type is high if the similarity calculation results corresponding to the sample objects are all greater than a preset similarity threshold;

[0086] The second association determination subunit is configured to determine that the association between the target object and the second business type is low if a similarity calculation result corresponding to any sample object is less than or equal to the preset similarity threshold.

[0087] In an exemplary embodiment, the device further comprises:

[0088] A candidate object set acquisition module, configured to execute acquisition of a candidate object set in a second business type of the target application;

[0089] A business data acquisition module, configured to execute acquisition of business data of a second business type of each candidate object in the candidate object set;

[0090] The sample object determination module is configured to determine the candidate object of the business data within the target interval as the sample object.

[0091] In an exemplary embodiment, the sample object determination module includes:

[0092] A to-be-screened object determination unit is configured to determine candidate objects of the business data within a target interval as to-be-screened objects, wherein the to-be-screened objects are at least two;

[0093] An object quantity acquisition unit, configured to execute acquisition of resources of the target object and the quantity of the objects of interest;

[0094] a target domain determining unit, configured to determine a target domain of the target object based on resources of the target object;

[0095] The sample object screening unit is configured to determine, based on the target field of the target object and the number of the concerned objects, sample objects having the same target field and matching the number of the concerned objects from at least two objects to be screened.

[0096] In an exemplary embodiment, the device further comprises:

[0097] The sharing information sending module is configured to send sharing information to the target object based on the association information, where the sharing information includes an identifier of the second business type.

[0098] In an exemplary embodiment, the target behavior feature includes at least one of the following:

[0099] The target object's behavior duration, the number of multimedia information released, the exposure duration of multimedia information, the delivery value of virtual resources, the recovery value of virtual resources, the growth number of attention objects and the number of occurrences of target content.

[0100] According to a third aspect of an embodiment of the present disclosure, there is provided an electronic device, including:

[0101] processor;

[0102] a memory for storing instructions executable by the processor;

[0103] The processor is configured to execute the instructions to implement the object recognition method as described above.

[0104] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the object recognition method as described above.

[0105] According to a fifth aspect of an embodiment of the present disclosure, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the object recognition method as described above is implemented.

[0106] The technical solution provided by the embodiments of the present disclosure brings at least the following beneficial effects:

[0107] The present disclosure obtains the target behavior feature of the target object in the target application; the target behavior feature is a feature related to the operation behavior of the target object in at least one first business type of the target application; obtains the sample behavior feature of the sample object of the second business type of the target application; the sample object and the target object are objects in the same field; the sample behavior feature and the target behavior feature are features of the same type; based on the similarity between the target behavior feature and the sample behavior feature, the association information between the target object and the second business type is determined; thereby, the accuracy of determining the association information between the target object and the second business type can be improved; and then based on the association information, an object recognition result for the target object is obtained. The present disclosure determines the association information between the target object and the second business type based on the target behavior feature and the sample behavior feature; thereby, the object recognition result of the target object is determined, the accuracy of the target object recognition result is improved, and the accuracy of the matching degree between the target object and the second business type is improved.

[0108] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0109] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.

[0110] Figure 1 The figure is a diagram showing an application environment of an object recognition method according to an exemplary embodiment.

[0111] Figure 2 The figure is a flow chart of an object recognition method according to an exemplary embodiment.

[0112] Figure 3 The figure is a flowchart of a method for determining a sample object according to an exemplary embodiment.

[0113] Figure 4 The present invention is a flowchart showing a method for determining the similarity between a target behavior feature and a sample behavior feature according to an exemplary embodiment.

[0114] Figure 5 The present invention is a flowchart of a method for determining a comprehensive similarity calculation result according to an exemplary embodiment.

[0115] Figure 6The present invention is a flowchart showing a method for determining a target weight corresponding to each target behavior feature according to an exemplary embodiment.

[0116] Figure 7 is a schematic structural diagram of an object recognition system according to an exemplary embodiment;

[0117] Figure 8 The figure is a block diagram of an object recognition device according to an exemplary embodiment.

[0118] Fig. 9 It is a block diagram of an electronic device for object recognition according to an exemplary embodiment. DETAILED DESCRIPTION

[0119] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.

[0120] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0121] The data involved in the present disclosure may be data authorized by the user or fully authorized by all parties.

[0122] See also Figure 1 , which is a diagram of an application environment of an object recognition method according to an exemplary embodiment. The application environment may include a server 01 and a client 02.

[0123] Specifically, in the embodiments of this specification, the server 01 may include an independently operated server, or a distributed server, or a server cluster composed of multiple servers, and may also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. Server 01 may include a network communication unit, a processor, and a memory, etc. Specifically, the server 01 can be used to obtain the target behavior characteristics of the target object in the target application and the sample behavior characteristics of the sample object of the second business type of the target application; and based on the similarity between the target behavior characteristics and the sample behavior characteristics, determine the association information between the target object and the second business type; and based on the association information, obtain the object recognition result of the target object.

[0124] Specifically, in the embodiments of this specification, the client 02 may include physical devices such as smart phones, desktop computers, tablet computers, laptop computers, digital assistants, smart wearable devices, and vehicle-mounted terminals, and may also include software running in physical devices, such as web pages provided by some service providers to users, or applications provided by these service providers to users. Specifically, the client 02 may be a terminal corresponding to the target object, used to display the sharing information including the second business type identifier.

[0125] Figure 2 is a flow chart of an object recognition method according to an exemplary embodiment. Figure 2 As shown, this method can be applied to Figure 1 The server 01 shown includes the following steps.

[0126] In step S21, a target behavior feature of a target object in a target application is obtained; the target behavior feature is a feature related to an operation behavior of the target object in at least one first business type of the target application;

[0127] In some embodiments, the target application may include but is not limited to live broadcast, short video and other applications; the first business type may include but is not limited to non-e-commerce businesses such as sports videos and food making videos in the target application; the target object may be an object that publishes multimedia information but does not sell goods, such as an anchor who publishes food making videos, sports videos, music videos, etc. on a live broadcast platform but does not sell goods online.

[0128] In some embodiments, the target behavior feature may be one or more, and the target behavior feature may include at least one of the following:

[0129] The target object's behavior duration, the number of multimedia information released, the exposure duration of multimedia information, the delivery value of virtual resources, the recovery value of virtual resources, the growth number of attention objects and the number of occurrences of target content.

[0130] In some embodiments, the target object's behavior duration can be the user's daily average live broadcast duration. Since the user's live broadcast behavior may fluctuate over time, the data from the past 30 days can be used as a reference standard when calculating the characteristics of the day. In the past 30 days, the user's live broadcast duration > 1 hour is a valid day, the number of valid days is d, the total live broadcast duration of valid days is m, and the average live broadcast duration h = m / d is taken.

[0131] In some embodiments, the number of multimedia information released can be the number of short videos released by the user per day. Since the number of short videos released by the user may fluctuate over time, the data in the past 30 days can be used as a reference standard when calculating the characteristics of the day. In the past 30 days, the total number of short videos released by the user in 30 days is s1, the number of valid days is d1, and the average number of short videos released is v1=s1 / d1.

[0132] In some embodiments, the exposure duration of multimedia information can be the viewing duration of the user's daily short video. Since the viewing duration of the short video posted by the user may fluctuate over time, the data in the past 30 days can be used as a reference standard when calculating the characteristics of the day. In the past 30 days, the viewing duration of the short video posted by the user in 30 days is s2, the number of days is d2, and the average viewing duration is v2=s2 / d2.

[0133] In some embodiments, the virtual resource delivery value may be the user's commercial delivery amount. Since the user's commercial delivery amount may fluctuate over time, the data in the past 30 days may be used as a reference standard when calculating the characteristics of the day. In the past 30 days, the user's commercial delivery amount is s3, the number of effective delivery days is d3, and the average commercial delivery amount is v3 = s3 / d3.

[0134] In some embodiments, the recovery value of virtual resources can be the user's return on investment (ROI). Since the user's commercial investment ROI may fluctuate over time, the data in the past 30 days can be used as a reference standard when calculating the characteristics of the day. In the past 30 days, the user's total investment ROI is s, the number of effective investment days is d, and the average investment ROI is v4=s4 / d4.

[0135] In some embodiments, the follow object may be an object that follows the target object, for example, the number of fans of the target object, and the growth number of the follow object may be the daily increase in the number of fans of the user. Since the increase in the number of fans of the user may fluctuate over time, the data in the past 30 days may be used as a reference standard when calculating the characteristics of the day. In the past 30 days, the increase in the number of fans of the user is s5, the number of days is d5, and the average increase in the number of fans is v5=s5 / d5.

[0136] In some embodiments, the target content may be commercial content; for example, in a sports video, the commercial content may be the target object's reply to a fan's question about a sports equipment purchase link. The number of times the target content appears may be the number of times the user's live broadcast content mentions the commercial content. Since the user's live broadcast content may contain commercial intent, this indicator may fluctuate over time. When calculating the characteristics of the day, the data within the past 30 days can be used as a reference standard. In the past 30 days, the number of times users mentioned commercial content was s6, and the number of valid days was d6. The average number of times the user's live broadcast content mentioned commercial content was v6 = s6 / d6.

[0137] In some embodiments, when obtaining the target behavior characteristics of the target object, multiple types of behavior characteristics of the target object can be obtained in advance, and the target behavior characteristics that meet the business indicators can be screened out; for example, the target behavior characteristics can be screened out based on the business indicator of the number of fans; outliers can also be screened out, missing values ​​can be filled in, and other operations can be performed from multiple behavior characteristics of the same type of the target object.

[0138] In the disclosed embodiment, any one or more characteristics of the target object's behavior duration, the number of multimedia information releases, the exposure duration of multimedia information, the virtual resource delivery value, the virtual resource recovery value, the growth number of objects of attention, and the number of occurrences of the target content can be used as target behavior characteristics according to the needs of business indicators, thereby improving the flexibility in selecting target behavior characteristics.

[0139] In step S23, sample behavior features of sample objects of the second business type of the target application are obtained; the sample objects and the target objects are objects of the same field; the sample behavior features and the target behavior features are features of the same type.

[0140] In the disclosed embodiment, the second business type may be an e-commerce business in the target application, such as live streaming with goods; the sample object is usually a live streamer in the target application, who can sell goods online. The sample object and the target object are objects in the same field, which means that the sample object and the target object may be objects in the same industry; for example, the sample object and the target object may both be live streamers in the food industry. The sample behavior feature and the target behavior feature are features of the same type, and if there is one target behavior feature, then there is also one sample behavior feature; the sample behavior feature and the target behavior feature may also be multiple features of the same type.

[0141] In an embodiment of the present disclosure, sample behavior characteristics of sample objects of the second business type of the target application are obtained; multiple types of behavior characteristics of sample objects can be obtained in advance, and sample behavior characteristics that meet business indicators can be screened out; for example, sample behavior characteristics can be screened out based on the business indicator of the number of fans; and operations such as outlier screening and missing value filling can also be performed on multiple behavior characteristics of the same type of sample objects.

[0142] In the present disclosure, Figure 3 As shown, before obtaining the sample behavior characteristics of the sample objects of the second business type of the target application, the sample objects of the second business type of the target application may be determined first, and the method further includes:

[0143] In step S221, a candidate object set in the second business type of the target application is obtained;

[0144] In the embodiment of the present disclosure, the candidate object set in the second business type of the target application may be a set of all hosts of the second business type in the target application, and sample objects may be screened out according to the set.

[0145] In step S223, business data of the second business type of each candidate object in the candidate object set is obtained;

[0146] In the disclosed embodiment, the business data of the candidate object may include but is not limited to the sales volume, the number of followers, and the website transaction amount (Gross Merchandise Volume, GMV) of the candidate object. GMV is a transaction indicator of e-commerce platform enterprises; it mainly refers to the total amount of the order, including the paid and unpaid parts. Sample objects can be screened out from the candidate object set based on the business data.

[0147] In step S225, candidate objects whose business data are within the target interval are determined as the sample objects.

[0148] In the embodiment of the present disclosure, the target interval can be determined based on the business data in the second business type of each candidate object in the candidate object set, and the median corresponding interval of the business data can be used as the target interval; for example, the median of a merchant group with a GMV between 500,000 and 2,000,000 can be determined as a sample object.

[0149] In the embodiment of the present disclosure, sample objects can be screened out according to the business data of candidate objects, and objects centered with the business data of the second business type in the target application can be used as sample objects, thereby improving the accuracy of subsequent determination of the association information between the target object and the second business type.

[0150] In some embodiments, determining the candidate object of the business data within the target interval as the sample object includes:

[0151] Determine candidate objects whose business data are within the target interval as objects to be screened, wherein the number of the objects to be screened is at least two;

[0152] In the embodiment of the present disclosure, candidate objects whose business data are within the target interval may be determined as preliminarily screened objects.

[0153] Obtain the resources of the target object and the number of concerned objects;

[0154] In the embodiment of the present disclosure, the target object's resources may be the object of the target object's live broadcast, such as food, etc. The target object's resources and the number of fans may be obtained. The target object's resources may be used to determine the field described by the target object.

[0155] Determining a target domain of the target object based on the resources of the target object;

[0156] In the disclosed embodiment, the target domain of the target object may be determined according to the resources of the target object, thereby facilitating the screening of sample objects in the same domain as the target object.

[0157] Based on the target field of the target object and the number of the objects of interest, sample objects having the same target field and matching the number of the objects of interest are determined from at least two objects to be screened.

[0158] In the disclosed embodiment, objects to be screened that are in the same field as the target object and have the same number of fans as the target object may be determined as sample objects.

[0159] In the disclosed embodiment, taking into account the differences in industries and potential traffic acquisition, in actual comparison, merchants with the same main industry and the same fan range can be selected for result comparison, thereby further improving the matching degree between the sample object and the second business type, thereby improving the accuracy of subsequent determination of the correlation information between the target object and the second business type.

[0160] In step S25, based on the similarity between the target behavior feature and the sample behavior feature, the association information between the target object and the second business type is determined.

[0161] In the disclosed embodiment, the target behavior feature and the sample behavior feature may be converted into feature vectors, and then the similarity between the two feature vectors may be calculated. For example, the cosine similarity between the two features may be calculated, and the similarity between the two feature vectors may be used as the similarity between the target behavior feature and the sample behavior feature.

[0162] In the disclosed embodiment, the similarity between the target object and the sample object can be determined by the similarity between the target behavior feature of the target object and the sample behavior feature of the sample object, thereby accurately determining the association information between the target object and the second business type.

[0163] In the embodiment of the present disclosure, there are at least two sample objects, and determining the association information between the target object and the second service type based on the similarity between the target behavior feature and the sample behavior feature includes:

[0164] Calculating the similarity between the target behavior feature and each sample behavior feature to obtain a similarity calculation result corresponding to each sample object;

[0165] In some embodiments, if there are multiple sample objects, the similarity between the target behavior feature and the sample behavior feature of each sample object may be calculated, thereby obtaining similarity calculation results corresponding to each sample object.

[0166] Based on the similarity calculation results corresponding to the sample objects, the association information between the target object and the second business type is determined.

[0167] In some embodiments, the similarity calculation results corresponding to each sample object can be determined based on the similarity between the target behavior characteristics of the target object and the sample behavior characteristics of each sample object, thereby determining the association information between the target object and the second business type. The association information represents the matching degree between the target object and the second business type, thereby determining the intended result of the target object being converted from the first business type to the second business type.

[0168] In the disclosed embodiment, the intended result of the target object being converted from a first business type to a second business type can be determined by calculating the similarity between the target object and multiple sample objects; an increase in the number of sample objects is conducive to improving the accuracy of determining the intended result of the target object being converted from a first business type to a second business type.

[0169] In some embodiments, the determining, based on the similarity calculation results corresponding to the sample objects, the association information between the target object and the second business type includes:

[0170] If the similarity calculation results corresponding to the sample objects are all greater than the preset similarity threshold, it is determined that the target object has a high correlation with the second business type;

[0171] In the embodiment of the present disclosure, the target object has a high correlation with the second service type, which indicates that the probability of the target object being converted from the first service type to the second service type is high.

[0172] In some embodiments, the determining, based on the similarity calculation results corresponding to the sample objects, the association information between the target object and the second business type includes:

[0173] If the similarity calculation result corresponding to any sample object is less than or equal to the preset similarity threshold, it is determined that the correlation between the target object and the second business type is low.

[0174] In the embodiment of the present disclosure, the correlation between the target object and the second service type is low, which means that the probability of the target object being converted from the first service type to the second service type is low.

[0175] In some embodiments, the determining, based on the similarity calculation results corresponding to the sample objects, the association information between the target object and the second business type includes:

[0176] If there are a preset number of sample objects whose corresponding similarity calculation results are less than or equal to the preset similarity threshold, it is determined that the association between the target object and the second business type is low.

[0177] In the embodiment of the present disclosure, the preset number can be set according to actual conditions. Multiple sample objects can constitute a sample object set, which can be set according to the total number of sample objects in the sample set. For example, the preset number can be set to 10% of the total number of sample objects. If there are a preset number of objects in the sample object set whose similarity with the target object is less than or equal to the preset similarity threshold, it can be determined that the target object has a low correlation with the second business type.

[0178] In some embodiments, a preset similarity threshold may be set according to actual conditions, so as to determine the association information between the target object and the second business type according to the preset similarity threshold. When there is one sample object, the association information between the target object and the second business type may be determined according to the similarity between the one sample object and the target object. When there are multiple sample objects, the association information between the target object and the second business type may be determined according to the similarity between the target object and multiple sample objects.

[0179] In the disclosed embodiment, a preset similarity threshold can be set to determine the association information between the target object and the second business type based on the similarity calculation result between the target object and the sample object; for multiple sample objects, the association information between the target object and the second business type can be determined based on the feature similarity between each sample object and the target object, thereby improving the accuracy of the object recognition result for determining the target object.

[0180] In the present disclosure, Figure 4 As shown, before determining the association information between the target object and the second business type based on the similarity between the target behavior feature and the sample behavior feature, the method further includes:

[0181] In step S241, the target behavior feature and the sample behavior feature are normalized to obtain a normalized target behavior feature and a normalized sample behavior feature;

[0182] In some embodiments, in order to eliminate the influence of extreme values, the target behavior features and the sample behavior features may be normalized; specifically, the normalization may be performed according to the values ​​corresponding to the target behavior features and the sample behavior features; for example, the normalization logic is as follows:

[0183] (1) For values ​​between the 10th and 90th percentiles, linear processing logic is used, and the score is set to (x-10th percentile) / (90th percentile-10th percentile);

[0184] (2) For values ​​above the 90th percentile, an inverse proportional function is used, and the score is set to 0.9 + 0.1 * (1-90th percentile / x);

[0185] (3) For values ​​below the 10th percentile, linear processing logic is adopted and the score is set to 0.1*(x / 10th percentile).

[0186] In step S243, the normalized target behavior feature is converted into a first feature vector, and the normalized sample behavior feature is converted into a second feature vector;

[0187] In some embodiments, the normalized features may be converted into feature vectors, and then the feature similarity may be calculated.

[0188] In step S245, the similarity between the first feature vector and the second feature vector is calculated to obtain a similarity calculation result;

[0189] In some embodiments, the similarity between the target object and the sample object may be obtained by calculating the similarity between two feature vectors.

[0190] In some embodiments, Figure 5 As shown, the target behavior features are at least two, and the similarity between the first feature vector and the second feature vector is calculated to obtain a similarity calculation result, including:

[0191] In step S2451, the similarity between the first feature vector and the second feature vector corresponding to each target behavior feature is calculated to obtain a similarity calculation result corresponding to each target behavior feature;

[0192] In some embodiments, the target behavior features of the target object may include multiple types of features, and at least two target behavior features mean that there are at least two types of target behavior features; in this case, there are also multiple sample behavior features of the sample object; the similarity calculation result corresponding to each target behavior feature can be obtained by calculating the similarity between the feature vector corresponding to each target behavior feature and the corresponding sample behavior feature.

[0193] In step S2453, the target weight corresponding to each target behavior feature is determined;

[0194] In some embodiments, a target weight corresponding to each target behavior feature may be set according to the importance of each target behavior feature relative to the target object, and the target weight represents the importance of each target behavior feature relative to the target object.

[0195] In some embodiments, Figure 6 As shown, the determining of the target weight corresponding to each target behavior feature includes:

[0196] In step S24531, a feature set corresponding to each target behavior feature within a preset time period is obtained;

[0197] In some embodiments, the preset period may be one month or one year, and the specific time may be set according to actual conditions. A feature set of target behavior features within the preset period may be obtained, and the feature set may include target behavior features at different time points in the preset period.

[0198] In step S24533, based on each feature in each feature set, the feature standard deviation and feature average corresponding to each feature set are determined;

[0199] In some embodiments, for features x1, ..., xn in the feature set, the feature standard deviation may be calculated by the following formula:

[0200]

[0201] Among them, S is the feature standard deviation, xi is the i-th feature in the feature set, and n is the number of features in the feature set. It refers to the characteristic average value of x1-xn.

[0202] In step S24535, based on the feature standard deviation and feature average corresponding to each feature set, the coefficient of variation of each target behavior feature is determined;

[0203] In some embodiments, the ratio of the feature standard deviation corresponding to each feature set to the corresponding feature average value may be determined as the coefficient of variation V of each target behavior feature; the calculation formula is as follows:

[0204]

[0205] In step S24537, based on the coefficient of variation of each target behavior feature, the target weight corresponding to each target behavior feature is determined.

[0206] In some embodiments, the coefficient of variation method can be understood as the coefficient of variation of each feature, and then the mean is introduced to eliminate the influence of different means on the comparison of the degree of variation. Finally, it is multiplied by the standardization coefficient to eliminate the data imbalance caused by the dimension difference. For example, for the viewing time of a feature short video, it is easy to be tens of thousands, while whether it is a high-quality video is only 1 or 0. Therefore, they need to be placed in the range of 0-1 before comparison.

[0207] In the disclosed embodiment, the target weight of each target behavior feature can be determined based on the feature set corresponding to each target behavior feature within a preset time period, thereby determining the importance of each target behavior feature, thereby improving the accuracy of the feature similarity calculation results between the target object and the sample object, thereby improving the accuracy of the target object's intention to enter the platform.

[0208] In some embodiments, determining the target weight corresponding to each target behavior feature based on the coefficient of variation of each target behavior feature includes:

[0209] Based on the coefficient of variation of each target behavior feature, calculating the sum of the coefficients of variation of each target behavior feature;

[0210] Calculating the ratio of the coefficient of variation of each target behavior feature to the sum of the coefficients of variation to obtain the coefficient of variation ratio of each target behavior feature;

[0211] The coefficient of variation ratio of each target behavior feature is used as the target weight corresponding to each target behavior feature.

[0212] In some embodiments, the target weight Wj corresponding to each target behavior feature is calculated as follows:

[0213]

[0214] Among them, Wj is the target weight corresponding to the j-th target behavior feature, and Vj is the coefficient of variation corresponding to the ith target behavior feature.

[0215] In the disclosed embodiment, the target weight of each target behavior feature can be determined based on the coefficient of variation corresponding to each target behavior feature, and the proportion of the coefficient of variation corresponding to each target behavior feature after the coefficient of variation corresponding to all target behavior features can be used as the target weight of each target behavior feature, thereby calculating the feature similarity calculation result between the target object and the sample object based on the weight, thereby improving the accuracy of the feature similarity calculation result.

[0216] In step S2455, based on the similarity calculation results corresponding to each target behavior feature and the target weight, a comprehensive similarity calculation result is determined.

[0217] In some embodiments, determining the comprehensive similarity calculation result based on the similarity calculation result corresponding to each target behavior feature and the target weight may include:

[0218] Determining an initial similarity calculation result for each target behavior feature based on the similarity calculation result corresponding to each target behavior feature and the target weight;

[0219] In the disclosed embodiment, the product of the similarity calculation result corresponding to each target behavior feature and the corresponding target weight may be used as the initial similarity calculation result of each target behavior feature.

[0220] In the disclosed embodiment, the features of the sample object can be processed accordingly according to the same processing method as the features of the target object; when calculating the similarity calculation result corresponding to the target behavior feature, the sample behavior feature corresponding to the target behavior feature needs to be subjected to the same normalization processing before calculating the similarity between the two, thereby improving the accuracy of the calculation result.

[0221] Based on the initial similarity calculation result of each target behavior feature, a comprehensive similarity calculation result is determined.

[0222] In the disclosed embodiment, the sum of the initial similarity calculation results of each target behavior feature may be determined as the comprehensive similarity calculation result.

[0223] In the disclosed embodiment, for multiple target behavior features, after determining the weight corresponding to each target behavior feature, that is, determining the importance of each feature, the comprehensive similarity calculation result can be calculated, thereby improving the accuracy of the similarity calculation between the target object and the sample object.

[0224] In some embodiments, the determining of the comprehensive similarity calculation result based on the similarity calculation result corresponding to each target behavior feature and the target weight includes:

[0225] Calculate the product of the similarity calculation result corresponding to each target behavior feature and the target weight to obtain an updated similarity calculation result of each target behavior feature;

[0226] The sum of the updated similarity calculation results of each target behavior feature is calculated to obtain the comprehensive similarity calculation result.

[0227] In the disclosed embodiment, after obtaining the similarity calculation result corresponding to each target behavior feature, the similarity calculation result corresponding to each target behavior feature can be updated according to the target weight of each target behavior feature, thereby obtaining an updated similarity calculation result corresponding to each target behavior feature, thereby improving the accuracy of the comprehensive similarity calculation result corresponding to the target object.

[0228] In step S247, the similarity calculation result is used as the similarity between the target behavior feature and the sample behavior feature.

[0229] In the disclosed embodiment, when determining the similarity between the target behavior feature and the sample behavior feature, the target behavior feature and the sample behavior feature can be first normalized and then converted into feature vectors to calculate the similarity, thereby eliminating the influence of extreme value features in the features and improving the accuracy of the similarity between the target behavior feature and the sample behavior feature.

[0230] In some embodiments, the using the similarity calculation result as the similarity between the target behavior feature and the sample behavior feature includes:

[0231] The comprehensive similarity calculation result is used as the similarity between the target behavior feature and the sample behavior feature.

[0232] In step S27, an object recognition result of the target object is obtained based on the association information.

[0233] In the embodiment of the present disclosure, the object recognition result may include: a first recognition result in which the target object has a high degree of match with the second business type, and a second recognition result in which the target object has a low degree of match with the second business type; if the target object has a high degree of association with the second business type, it means that the two have a high degree of match; if the target object has a low degree of association with the second business type, it means that the two have a low degree of match.

[0234] In some embodiments, the method further comprises:

[0235] Based on the association information, sharing information is sent to the target object, where the sharing information includes an identifier of the second service type.

[0236] In some embodiments, based on the association information, sending the sharing information to the target object includes:

[0237] If the association information indicates that the target object has a high association with the second business type, sharing information is sent to the target object.

[0238] In some embodiments, if the association information indicates that the association between the target object and the second business type is low, the sharing information is not sent to the target object.

[0239] In some embodiments, the sharing information may be link information of the second business type, and may include identification information of the second business type; the target object may quickly convert from the first business type to the second business type based on the sharing information. If it is determined that the target object has a high degree of association with the second business type, the sharing information may be sent to the target object, so that the target object may quickly convert to the second business type based on the identification of the second business type in the sharing information.

[0240] In the disclosed embodiment, when it is determined that the target object has a high correlation with the second business type, sharing information is sent to the target object, so that the target object can be quickly converted from the first business type to the second business type according to the sharing information, thereby improving the conversion rate of the sharing information.

[0241] In a specific embodiment, Figure 7 As shown, Figure 7 The structure diagram of an object recognition system includes: a feature service module, a feature production module, a feature analysis model (Data Analysis, DA), a data storage module, a sample data acquisition module and a data analysis module;

[0242] Among them, the feature service module includes a feature query module and a feature management module. The feature query module is used to query merchant features, and the feature management module is used to edit merchant features; the feature production module includes a feature configuration submodule, a feature production submodule, a scheduling submodule and a monitoring submodule; the feature configuration submodule is used to configure features, the feature production submodule is used to combine existing features to produce new combined features, the scheduling submodule is used to schedule features, and the monitoring submodule is used to monitor features and detect abnormal features in a timely manner; the feature analysis module includes an indicator extraction submodule, a missing value filling submodule, and an abnormal value (skewed value) processing submodule. Block, normalization and weight calculation submodules; the indicator extraction submodule is used to select specific features according to actual conditions, the missing value filling submodule is used to fill the missing values ​​of features, and the outlier (skewed value) processing submodule is used to process outliers and skewed values, and can screen out values; the normalization and weight calculation submodule is used to normalize the features and calculate the weight of each feature; the data storage module includes a data warehouse tool (HIVE), an open source stream processing platform (Kafka), a key-value database (key-value, KV) and a remote procedure call module (Remote Procedure Call, RPC). Kafka is a high-throughput distributed publish-subscribe messaging system that can process all action stream data of consumers on the website. HIVE is a data warehouse tool based on Hadoop, which is used for data extraction, transformation, and loading. This is a mechanism that can store, query and analyze large-scale data stored in Hadoop. The HIVE data warehouse tool can map structured data files to a database table; Hadoop is a distributed system infrastructure. A key-value database is a database that stores data in key-value pairs; RPC is to call a function or method (collectively referred to as a service) on another machine (server) from one machine (client) by passing parameters and get the returned result.

[0243] The sample data collection module is used to collect data from sample users throughout their life cycle, including videos posted by users, live broadcast data, commercial launch data, and fan interaction data. After sample data collection, it can be used as leads for merchants.

[0244] Among them, the data analysis module is used to combine merchant clues and the characteristics of each merchant to determine merchants with a higher willingness to join the platform.

[0245] The present disclosure obtains the target behavior feature of the target object in the target application; the target behavior feature is a feature related to the operation behavior of the target object in at least one first business type of the target application; obtains the sample behavior feature of the sample object of the second business type of the target application; the sample object and the target object are objects in the same field; the sample behavior feature and the target behavior feature are features of the same type; based on the similarity between the target behavior feature and the sample behavior feature, the association information between the target object and the second business type is determined; thereby, the accuracy of determining the association information between the target object and the second business type can be improved; and then based on the association information, an object recognition result for the target object is obtained. The present disclosure determines the association information between the target object and the second business type based on the target behavior feature and the sample behavior feature; thereby, the object recognition result of the target object is determined, the accuracy of the target object recognition result is improved, and the accuracy of the matching degree between the target object and the second business type is improved.

[0246] Figure 8 is a block diagram of an object recognition device according to an exemplary embodiment. Figure 8 , the device comprises:

[0247] The target behavior feature acquisition module 810 is configured to execute acquisition of the target behavior feature of the target object in the target application; the target behavior feature is a feature related to the operation behavior of the target object in at least one first business type of the target application;

[0248] The sample behavior feature acquisition module 820 is configured to execute acquisition of sample behavior features of a sample object of a second business type of the target application; the sample object and the target object are objects of the same field; the sample behavior features and the target behavior features are features of the same type;

[0249] The association information determination module 830 is configured to determine the association information between the target object and the second business type based on the similarity between the target behavior feature and the sample behavior feature;

[0250] The object recognition result determination module 840 is configured to obtain an object recognition result for the target object based on the association degree information.

[0251] In some embodiments, the apparatus further comprises:

[0252] A normalization processing module is configured to perform normalization processing on the target behavior feature and the sample behavior feature to obtain a normalized target behavior feature and a normalized sample behavior feature;

[0253] A vector conversion module, configured to convert the normalized target behavior feature into a first feature vector and convert the normalized sample behavior feature into a second feature vector;

[0254] a similarity calculation result determination module, configured to calculate the similarity between the first feature vector and the second feature vector to obtain a similarity calculation result;

[0255] The behavior feature similarity determination module is configured to use the similarity calculation result as the similarity between the target behavior feature and the sample behavior feature.

[0256] In some embodiments, the target behavior features are at least two, and the similarity calculation result determination module includes:

[0257] A similarity calculation result determination unit is configured to calculate the similarity between the first feature vector and the second feature vector corresponding to each target behavior feature to obtain a similarity calculation result corresponding to each target behavior feature;

[0258] A target weight determination unit, configured to determine the target weight corresponding to each target behavior feature;

[0259] A comprehensive similarity calculation result determination unit is configured to determine a comprehensive similarity calculation result based on the similarity calculation result corresponding to each target behavior feature and the target weight;

[0260] The behavior feature similarity determination module includes:

[0261] The behavior feature similarity determination unit is configured to use the comprehensive similarity calculation result as the similarity between the target behavior feature and the sample behavior feature.

[0262] In some embodiments, the comprehensive similarity calculation result determination unit includes:

[0263] A product calculation subunit is configured to calculate the product of the similarity calculation result corresponding to each target behavior feature and the target weight to obtain the similarity calculation result of each target behavior feature;

[0264] The comprehensive similarity calculation result determination subunit is configured to calculate the sum of the similarity calculation results of each target behavior feature to obtain the comprehensive similarity calculation result.

[0265] In some embodiments, the target weight determination unit includes:

[0266] A feature set acquisition subunit is configured to acquire a feature set corresponding to each target behavior feature within a preset time period;

[0267] A feature average value determination subunit is configured to determine a feature standard deviation and a feature average value corresponding to each feature set based on each feature in each feature set;

[0268] A coefficient of variation determination subunit, configured to determine the coefficient of variation of each target behavior feature based on a feature standard deviation and a feature average corresponding to each feature set;

[0269] The target weight determination subunit is configured to determine the target weight corresponding to each target behavior feature based on the coefficient of variation of each target behavior feature.

[0270] In some embodiments, the target weight determination subunit includes:

[0271] A coefficient of variation and calculation subunit, configured to calculate the sum of the coefficients of variation of each target behavior feature based on the coefficient of variation of each target behavior feature;

[0272] A coefficient of variation ratio determination subunit is configured to calculate the ratio of the coefficient of variation of each target behavior feature to the sum of the coefficients of variation to obtain the coefficient of variation ratio of each target behavior feature;

[0273] The feature weight determination subunit is configured to use the coefficient of variation ratio of each target behavior feature as the target weight corresponding to each target behavior feature.

[0274] In some embodiments, the number of the sample objects is at least two, and the association information determination module includes:

[0275] A similarity calculation result determination unit is configured to calculate the similarity between the target behavior feature and each sample behavior feature to obtain a similarity calculation result corresponding to each sample object;

[0276] The association information determining unit is configured to determine the association information between the target object and the second business type based on the similarity calculation results corresponding to the sample objects.

[0277] In some embodiments, the association information determination unit includes:

[0278] A first relevance determination subunit is configured to determine that the relevance between the target object and the second business type is high if the similarity calculation results corresponding to the sample objects are all greater than a preset similarity threshold;

[0279] The second association determination subunit is configured to determine that the association between the target object and the second business type is low if a similarity calculation result corresponding to any sample object is less than or equal to the preset similarity threshold.

[0280] In some embodiments, the apparatus further comprises:

[0281] A candidate object set acquisition module, configured to execute acquisition of a candidate object set in a second business type of the target application;

[0282] A business data acquisition module, configured to execute acquisition of business data of a second business type of each candidate object in the candidate object set;

[0283] The sample object determination module is configured to determine the candidate object of the business data within the target interval as the sample object.

[0284] In some embodiments, the sample object determination module includes:

[0285] A to-be-screened object determination unit is configured to determine candidate objects of the business data within a target interval as to-be-screened objects, wherein the to-be-screened objects are at least two;

[0286] An object quantity acquisition unit, configured to execute acquisition of resources of the target object and the quantity of the objects of interest;

[0287] a target domain determining unit, configured to determine a target domain of the target object based on resources of the target object;

[0288] The sample object screening unit is configured to determine, based on the target field of the target object and the number of the concerned objects, sample objects having the same target field and matching the number of the concerned objects from at least two objects to be screened.

[0289] In some embodiments, the apparatus further comprises:

[0290] The sharing information sending module is configured to send sharing information to the target object based on the association information, where the sharing information includes an identifier of the second business type.

[0291] In some embodiments, the target behavior feature includes at least one of the following:

[0292] The target object's behavior duration, the number of multimedia information released, the exposure duration of multimedia information, the delivery value of virtual resources, the recovery value of virtual resources, the growth number of attention objects and the number of occurrences of target content.

[0293] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0294] Fig. 9 is a block diagram of an electronic device for object recognition according to an exemplary embodiment. The electronic device may be a terminal, and its internal structure diagram may be as shown in FIG. Fig. 9 As shown. The electronic device includes a processor, a memory, a network interface, a display screen and an input device connected via a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, an object recognition method is implemented. The display screen of the electronic device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the electronic device can be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the electronic device, or an external keyboard, touchpad or mouse, etc.

[0295] Those skilled in the art will understand that Fig. 9 The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present disclosure, and does not constitute a limitation on the electronic device to which the scheme of the present disclosure is applied. The specific electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0296] In an exemplary embodiment, there is also provided an electronic device, comprising:

[0297] A processor; a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement the above-mentioned object recognition method.

[0298] In an exemplary embodiment, a computer-readable storage medium including instructions is also provided, such as a memory including instructions, and the above instructions can be executed by a processor of an electronic device to complete the above object recognition method. Optionally, the computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0299] In an exemplary embodiment, a computer program product is also provided, including a computer program, and when the computer program is executed by a processor, the object recognition method described above is implemented.

[0300] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0301] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0302] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. An object recognition method, characterized in that: include: Obtain target behavior characteristics of the target object in the target application; The target behavior feature is a feature related to the operation behavior of the target object in at least one first business type of the target application; the target behavior feature includes at least one of the following: the behavior duration of the target object, the number of multimedia information released, the exposure duration of multimedia information, the delivery value of virtual resources, the recovery value of virtual resources, the growth number of attention objects, and the number of occurrences of target content; Determine candidate objects of the business data within the target interval as objects to be screened, wherein the number of the objects to be screened is at least two; the candidate objects are objects in the second business type of the target application; Based on the target field of the target object and the number of objects of interest, determining sample objects having the same target field and matching the number of objects of interest from at least two objects to be screened; Acquire sample behavior characteristics of sample objects of a second business type of the target application; The sample behavior features and the target behavior features are features of the same type; the sample behavior features and the target behavior features are of the same number; Normalizing the corresponding values ​​of the target behavior feature and the sample behavior feature according to the tenth and ninetieth digits to obtain normalized target behavior features and normalized sample behavior features; Determining the similarity between the normalized target behavior feature and the normalized sample behavior feature as the similarity between the target behavior feature and the sample behavior feature; Determining, based on the similarity between the target behavior feature and the sample behavior feature, the association information between the target object and the second service type; Based on the association degree information, an object recognition result of the target object is obtained.

2. The method according to claim 1, characterized in that Before determining the association information between the target object and the second service type based on the similarity between the target behavior feature and the sample behavior feature, the method further includes: Converting the normalized target behavior feature into a first feature vector, and converting the normalized sample behavior feature into a second feature vector; Calculating the similarity between the first feature vector and the second feature vector to obtain a similarity calculation result; The similarity calculation result is used as the similarity between the target behavior feature and the sample behavior feature.

3. The method according to claim 2, characterized in that The target behavior features are at least two, and the calculating the similarity between the first feature vector and the second feature vector to obtain a similarity calculation result includes: Calculating the similarity between the first feature vector and the second feature vector corresponding to each target behavior feature to obtain a similarity calculation result corresponding to each target behavior feature; Determine the target weight corresponding to each target behavior feature; Determine a comprehensive similarity calculation result based on the similarity calculation result corresponding to each target behavior feature and the target weight; The using the similarity calculation result as the similarity between the target behavior feature and the sample behavior feature includes: The comprehensive similarity calculation result is used as the similarity between the target behavior feature and the sample behavior feature.

4. The method according to claim 3, characterized in that The determining of the comprehensive similarity calculation result based on the similarity calculation result corresponding to each target behavior feature and the target weight includes: Calculate the product of the similarity calculation result corresponding to each target behavior feature and the target weight to obtain the similarity calculation result of each target behavior feature; The sum of the similarity calculation results of each target behavior feature is calculated to obtain the comprehensive similarity calculation result.

5. The method according to claim 3, characterized in that: The determining the target weight corresponding to each target behavior feature includes: Obtaining a feature set corresponding to each target behavior feature within a preset time period; Based on each feature in each feature set, determine a feature standard deviation and a feature average corresponding to each feature set; Determine the coefficient of variation of each target behavior feature based on the feature standard deviation and feature average corresponding to each feature set; Based on the coefficient of variation of each target behavior feature, a target weight corresponding to each target behavior feature is determined.

6. The method according to claim 5, characterized in that The determining, based on the coefficient of variation of each target behavior feature, a target weight corresponding to each target behavior feature comprises: Based on the coefficient of variation of each target behavior feature, calculating the sum of the coefficients of variation of each target behavior feature; Calculating the ratio of the coefficient of variation of each target behavior feature to the sum of the coefficients of variation to obtain the coefficient of variation ratio of each target behavior feature; The coefficient of variation ratio of each target behavior feature is used as the target weight corresponding to each target behavior feature.

7. The method according to any one of claims 1 to 6, characterized in that: There are at least two sample objects, and determining the association information between the target object and the second service type based on the similarity between the target behavior feature and the sample behavior feature includes: Calculating the similarity between the target behavior feature and each sample behavior feature to obtain a similarity calculation result corresponding to each sample object; Based on the similarity calculation results corresponding to the sample objects, the association information between the target object and the second business type is determined.

8. The method according to claim 7, characterized in that The determining, based on the similarity calculation results corresponding to the sample objects, the association information between the target object and the second business type includes: If the similarity calculation results corresponding to the sample objects are all greater than the preset similarity threshold, it is determined that the target object has a high correlation with the second business type; If the similarity calculation result corresponding to any sample object is less than or equal to the preset similarity threshold, it is determined that the correlation between the target object and the second business type is low.

9. The method according to any one of claims 1 to 6, characterized in that: Before obtaining the sample behavior characteristics of the sample object of the second business type of the target application, the method further includes: Acquire a candidate object set in a second business type of the target application; Acquire business data of a second business type of each candidate object in the candidate object set; The candidate objects whose business data are within the target interval are determined as the sample objects.

10. The method according to claim 1, characterized in that The method further comprises: Obtain the resources of the target object and the number of concerned objects; Based on the resources of the target object, a target domain of the target object is determined.

11. The method according to claim 1, characterized in that: The method further comprises: Based on the association information, sharing information is sent to the target object, where the sharing information includes an identifier of the second service type.

12. An object recognition device, characterized in that: include: A target behavior feature acquisition module is configured to execute acquisition of target behavior features of a target object in a target application; The target behavior feature is a feature related to the operation behavior of the target object in at least one first business type of the target application; the target behavior feature includes at least one of the following: the behavior duration of the target object, the number of multimedia information released, the exposure duration of multimedia information, the delivery value of virtual resources, the recovery value of virtual resources, the growth number of attention objects, and the number of occurrences of target content; A sample object determination module is configured to determine candidate objects of the business data within a target interval as objects to be screened, wherein the number of the objects to be screened is at least two; the candidate objects are objects of a second business type of the target application; and based on the target field of the target object and the number of objects of interest, determining, from at least two objects to be screened, sample objects that are identical to the target field and match the number of objects of interest; A sample behavior feature acquisition module is configured to execute acquisition of sample behavior features of sample objects of a second business type of the target application; The sample behavior features and the target behavior features are features of the same type; the sample behavior features and the target behavior features are of the same number; A normalization processing module is configured to perform normalization processing on the corresponding values ​​in the target behavior feature and the sample behavior feature according to the tenth and ninetieth digits to obtain a normalized target behavior feature and a normalized sample behavior feature; Determining the similarity between the normalized target behavior feature and the normalized sample behavior feature as the similarity between the target behavior feature and the sample behavior feature; A correlation information determination module, configured to determine the correlation information between the target object and the second business type based on the similarity between the target behavior feature and the sample behavior feature; The object recognition result determination module is configured to obtain an object recognition result of the target object based on the association degree information.

13. The device according to claim 12, characterized in that The device also includes: A vector conversion module, configured to convert the normalized target behavior feature into a first feature vector and convert the normalized sample behavior feature into a second feature vector; a similarity calculation result determination module, configured to calculate the similarity between the first feature vector and the second feature vector to obtain a similarity calculation result; The behavior feature similarity determination module is configured to use the similarity calculation result as the similarity between the target behavior feature and the sample behavior feature.

14. The device according to claim 13, characterized in that The target behavior features are at least two, and the similarity calculation result determination module includes: A similarity calculation result determination unit is configured to calculate the similarity between the first feature vector and the second feature vector corresponding to each target behavior feature to obtain a similarity calculation result corresponding to each target behavior feature; A target weight determination unit, configured to determine the target weight corresponding to each target behavior feature; A comprehensive similarity calculation result determination unit is configured to determine a comprehensive similarity calculation result based on the similarity calculation result corresponding to each target behavior feature and the target weight; The behavior feature similarity determination module includes: The behavior feature similarity determination unit is configured to use the comprehensive similarity calculation result as the similarity between the target behavior feature and the sample behavior feature.

15. The device according to claim 14, characterized in that The comprehensive similarity calculation result determination unit comprises: A product calculation subunit is configured to calculate the product of the similarity calculation result corresponding to each target behavior feature and the target weight to obtain the similarity calculation result of each target behavior feature; The comprehensive similarity calculation result determination subunit is configured to calculate the sum of the similarity calculation results of each target behavior feature to obtain the comprehensive similarity calculation result.

16. The device according to claim 14, characterized in that The target weight determination unit comprises: A feature set acquisition subunit is configured to acquire a feature set corresponding to each target behavior feature within a preset time period; A feature average value determination subunit is configured to determine a feature standard deviation and a feature average value corresponding to each feature set based on each feature in each feature set; A coefficient of variation determination subunit, configured to determine the coefficient of variation of each target behavior feature based on a feature standard deviation and a feature average corresponding to each feature set; The target weight determination subunit is configured to determine the target weight corresponding to each target behavior feature based on the coefficient of variation of each target behavior feature.

17. The device according to claim 16, characterized in that The target weight determination subunit comprises: A coefficient of variation and calculation subunit, configured to calculate the sum of the coefficients of variation of each target behavior feature based on the coefficient of variation of each target behavior feature; A coefficient of variation ratio determination subunit is configured to calculate the ratio of the coefficient of variation of each target behavior feature to the sum of the coefficients of variation to obtain the coefficient of variation ratio of each target behavior feature; The feature weight determination subunit is configured to use the coefficient of variation ratio of each target behavior feature as the target weight corresponding to each target behavior feature.

18. The device according to any one of claims 12 to 17, characterized in that: There are at least two sample objects, and the association information determination module includes: A similarity calculation result determination unit is configured to calculate the similarity between the target behavior feature and each sample behavior feature to obtain a similarity calculation result corresponding to each sample object; The association information determining unit is configured to determine the association information between the target object and the second business type based on the similarity calculation results corresponding to the sample objects.

19. The device according to claim 18, characterized in that The association information determination unit includes: A first relevance determination subunit is configured to determine that the relevance between the target object and the second business type is high if the similarity calculation results corresponding to the sample objects are all greater than a preset similarity threshold; The second association determination subunit is configured to determine that the association between the target object and the second business type is low if a similarity calculation result corresponding to any sample object is less than or equal to the preset similarity threshold.

20. The device according to any one of claims 12 to 17, characterized in that The device also includes: A candidate object set acquisition module, configured to execute acquisition of a candidate object set in a second business type of a target application; A business data acquisition module, configured to execute acquisition of business data of a second business type of each candidate object in the candidate object set; The sample object determination module is configured to determine the candidate object of the business data within the target interval as the sample object.

21. The device according to claim 12, characterized in that The device also includes: An object quantity acquisition unit, configured to execute acquisition of resources of the target object and the quantity of the objects of interest; The target domain determining unit is configured to determine the target domain of the target object based on the resources of the target object.

22. The device according to claim 12, characterized in that The device also includes: The sharing information sending module is configured to send sharing information to the target object based on the association information, where the sharing information includes an identifier of the second business type.

23. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the object recognition method according to any one of claims 1 to 11.

24. A computer-readable storage medium, characterized in that: When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the object recognition method according to any one of claims 1 to 11.

25. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the object recognition method according to any one of claims 1 to 11 is implemented.

Images