CN112070511A

CN112070511A - Method and equipment for detecting unqualified commodities

Info

Publication number: CN112070511A
Application number: CN202010809205.XA
Authority: CN
Inventors: 陈文涛
Original assignee: Shanghai Lianshang Network Technology Co Ltd
Current assignee: Shanghai Lianshang Network Technology Co Ltd
Priority date: 2020-08-12
Filing date: 2020-08-12
Publication date: 2020-12-11

Abstract

The application aims to provide a method and equipment for detecting unqualified commodities, wherein the method comprises the following steps: acquiring connection relation characteristic information corresponding to a knowledge graph, wherein the knowledge graph is constructed according to commodity characteristic information corresponding to a plurality of commodities and object characteristic information corresponding to associated objects associated with the commodity characteristic information, each node in the knowledge graph corresponds to one commodity or one associated object, and the connection relation characteristic information is used for representing the connection relation between each node in the knowledge graph; and obtaining a commodity detection model through training according to the commodity characteristic information, the object characteristic information, the connection relation characteristic information and commodity calibration information corresponding to the commodities, wherein the commodity calibration information is used for calibrating whether each commodity in the commodities is an unqualified commodity.

Description

Method and equipment for detecting unqualified commodities

Technical Field

The present application relates to the field of communications, and in particular, to a technique for detecting an unqualified commodity.

Background

With the progress of science and technology and the development of society, the network is closer to our lives, and the online shopping mode is accepted by most people, but many counterfeit commodities and illegal commodities exist in online shopping, and people are difficult to judge. In the prior art, fake and counterfeit commodities and forbidden commodities can only be marked manually, however, if the description of the commodities changes frequently, the fake and counterfeit commodities and the forbidden commodities are difficult to detect, and the mode needs to spend great labor cost and time cost.

Disclosure of Invention

It is an object of the present application to provide a method and apparatus for detecting defective goods.

According to one aspect of the present application, there is provided a method of detecting a defective article, the method comprising:

acquiring connection relation characteristic information corresponding to a knowledge graph, wherein the knowledge graph is constructed according to commodity characteristic information corresponding to a plurality of commodities and object characteristic information corresponding to associated objects associated with the commodity characteristic information, the knowledge graph comprises a plurality of nodes, each node in the knowledge graph corresponds to one commodity or one associated object, and the connection relation characteristic information is used for representing the connection relation between each node in the knowledge graph;

obtaining a commodity detection model through training according to the commodity feature information, the object feature information, the connection relation feature information and commodity calibration information corresponding to the commodities, wherein the commodity calibration information is used for calibrating whether each commodity in the commodities is an unqualified commodity;

inputting the characteristic information of the target commodity corresponding to the target commodity into the commodity detection model to obtain commodity detection information which is output by the commodity detection model and corresponds to the target commodity, wherein the commodity detection information is used for indicating whether the target commodity is an unqualified commodity.

According to an aspect of the present application, there is provided a network apparatus for detecting a defective commodity, the apparatus including:

the system comprises a module, a module and a module, wherein the module is used for obtaining connection relation characteristic information corresponding to a knowledge graph, the knowledge graph is constructed according to commodity characteristic information corresponding to a plurality of commodities and object characteristic information corresponding to related objects related to the commodity characteristic information, the knowledge graph comprises a plurality of nodes, each node in the knowledge graph corresponds to one commodity or one related object, and the connection relation characteristic information is used for representing the connection relation between each node in the knowledge graph;

a second module, configured to obtain a commodity detection model through training according to the commodity feature information, the object feature information, the connection relationship feature information, and commodity calibration information corresponding to the plurality of commodities, where the commodity calibration information is used to calibrate whether each commodity in the plurality of commodities is an unqualified commodity;

and the three modules are used for inputting the characteristic information of the target commodity corresponding to the target commodity into the commodity detection model to obtain the commodity detection information which is output by the commodity detection model and corresponds to the target commodity, wherein the commodity detection information is used for indicating whether the target commodity is an unqualified commodity.

According to one aspect of the present application, there is provided an apparatus for detecting defective goods, wherein the apparatus comprises:

a processor; and

a memory arranged to store computer executable instructions that, when executed, cause the processor to:

According to one aspect of the application, there is provided a computer-readable medium storing instructions that, when executed, cause a system to:

Compared with the prior art, the method and the device have the advantages that the knowledge graph can be constructed according to the commodity feature information corresponding to the commodities and the object feature information corresponding to the related objects related to the commodity feature information, the commodity detection model is trained according to the commodity feature information, the object feature information, the connection relation feature information corresponding to each node in the knowledge graph and the commodity calibration information, and whether the target commodity is the unqualified commodity can be detected quickly and accurately through the commodity detection model.

Drawings

Other features, objects and advantages of the present application will become more apparent upon reading of the following detailed description of non-limiting embodiments thereof, made with reference to the accompanying drawings in which:

FIG. 1 illustrates a flow chart of a method of detecting a rejected good according to one embodiment of the present application;

FIG. 2 illustrates a flow diagram of a method of constructing a knowledge-graph according to one embodiment of the present application;

FIG. 3 illustrates a flow diagram of a method for obtaining a merchandise detection model through training according to one embodiment of the present application;

FIG. 4 illustrates a network device architecture diagram for detecting defective goods according to one embodiment of the present application;

FIG. 5 illustrates an exemplary system that can be used to implement the various embodiments described in this application.

The same or similar reference numbers in the drawings identify the same or similar elements.

Detailed Description

The present application is described in further detail below with reference to the attached figures.

In a typical configuration of the present application, the terminal, the device serving the network, and the trusted party each include one or more processors (e.g., Central Processing Units (CPUs)), input/output interfaces, network interfaces, and memory.

The Memory may include forms of volatile Memory, Random Access Memory (RAM), and/or non-volatile Memory in a computer-readable medium, such as Read Only Memory (ROM) or Flash Memory. Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, Phase-Change Memory (PCM), Programmable Random Access Memory (PRAM), Static Random-Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), electrically Erasable Programmable Read-Only Memory (EEPROM), flash Memory or other Memory technology, Compact Disc Read-Only Memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device.

The device referred to in this application includes, but is not limited to, a user device, a network device, or a device formed by integrating a user device and a network device through a network. The user equipment includes, but is not limited to, any mobile electronic product, such as a smart phone, a tablet computer, etc., capable of performing human-computer interaction with a user (e.g., human-computer interaction through a touch panel), and the mobile electronic product may employ any operating system, such as an Android operating system, an iOS operating system, etc. The network Device includes an electronic Device capable of automatically performing numerical calculation and information processing according to a preset or stored instruction, and the hardware includes, but is not limited to, a microprocessor, an Application Specific Integrated Circuit (ASIC), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a Digital Signal Processor (DSP), an embedded Device, and the like. The network device includes but is not limited to a computer, a network host, a single network server, a plurality of network server sets or a cloud of a plurality of servers; here, the Cloud is composed of a large number of computers or web servers based on Cloud Computing (Cloud Computing), which is a kind of distributed Computing, one virtual supercomputer consisting of a collection of loosely coupled computers. Including, but not limited to, the internet, a wide area network, a metropolitan area network, a local area network, a VPN network, a wireless Ad Hoc network (Ad Hoc network), etc. Preferably, the device may also be a program running on the user device, the network device, or a device formed by integrating the user device and the network device, the touch terminal, or the network device and the touch terminal through a network.

Of course, those skilled in the art will appreciate that the foregoing is by way of example only, and that other existing or future devices, which may be suitable for use in the present application, are also encompassed within the scope of the present application and are hereby incorporated by reference.

In the description of the present application, "a plurality" means two or more unless specifically limited otherwise.

Fig. 1 shows a flowchart of a method for detecting defective goods according to an embodiment of the present application, which includes steps S11, S12, and S13. In step S11, the network device obtains connection relation feature information corresponding to a knowledge graph, where the knowledge graph is constructed according to commodity feature information corresponding to a plurality of commodities and object feature information corresponding to associated objects associated with the commodity feature information, each node in the knowledge graph corresponds to a commodity or an associated object, and the connection relation feature information is used to represent a connection relation between nodes in the knowledge graph; in step S12, the network device obtains a commodity detection model through training according to the commodity feature information, the object feature information, the connection relationship feature information, and commodity calibration information corresponding to the plurality of commodities, where the commodity calibration information is used to calibrate whether each commodity in the plurality of commodities is an unqualified commodity; in step S13, the network device inputs the target product feature information corresponding to the target product into the product detection model, and obtains the product detection information corresponding to the target product and output by the product detection model, where the product detection information is used to indicate whether the target product is an unqualified product.

In step S11, the network device obtains connection characteristic information corresponding to a knowledge graph, where the knowledge graph is constructed according to commodity characteristic information corresponding to a plurality of commodities and object characteristic information corresponding to associated objects associated with the commodity characteristic information, each node in the knowledge graph corresponds to a commodity or an associated object, and the connection characteristic information is used to represent a connection relationship between nodes in the knowledge graph.

In some embodiments, the product characteristic information includes any information related to characteristics of the product, optionally, the product characteristic information includes, but is not limited to, product title description, product classification label, product price, merchant information corresponding to the product, product review information, product delivery location, product sales volume, product evaluation information, product description information, and the like. In some embodiments, the merchandise characteristic information may have one or more associated objects. In some embodiments, the associated object associated with the commodity feature information may be any object included in the commodity feature information, for example, the commodity feature information of the commodity a includes merchant information "merchant B" corresponding to the commodity a, and then the associated object associated with the commodity feature information may be merchant B; for another example, if the product characteristic information of the product a includes product description information "dog likes to use the product" corresponding to the product a, the related object associated with the product characteristic information may be a dog. In some embodiments, an object associated with the semantic content may also be determined as an associated object associated with the product characteristic information according to the semantic content of the product characteristic information, for example, if the semantic content of the product characteristic information of the product a includes "high-end pet food brand", then the objects "cat" and "dog" associated with the semantic content may be determined as the associated object associated with the product characteristic information. In some embodiments, the associated object may be any object in any form, preferably including, but not limited to, a merchant object, a user object, a merchandise object, and the like.

In some embodiments, the object feature information corresponding to the associated object includes any information related to features of the associated object, and when one associated object is a certain commodity, the object feature information corresponding to the associated object is commodity feature information of the commodity; when a related object is a user, object characteristic information corresponding to the related object includes, but is not limited to, historical behavior information of the user for uploading, editing, browsing and purchasing goods, interest tag information of the user, and the like; when a related object is a certain merchant, the object characteristic information corresponding to the related object includes, but is not limited to, other goods sold by the merchant, evaluation information of the merchant, and the like.

In some embodiments, a knowledge graph is constructed by using commodity feature information corresponding to a plurality of commodities and object feature information corresponding to an associated object associated with the commodity feature information, the knowledge graph includes a plurality of nodes, each node corresponds to one commodity or one associated object, and the commodity feature information or the object feature information is an attribute of the corresponding commodity node or the associated object node. In some embodiments, the knowledge graph can reflect a relationship between nodes, in the knowledge graph, a connection between two nodes (i.e., an association between a commodity and a commodity, an association between a commodity and an associated object, and an association between an associated object and an associated object) can be established through respective attributes of the two nodes (i.e., commodity feature information corresponding to a commodity node or object feature information corresponding to an associated object node), and the two nodes can be directly connected or indirectly connected through one or more other nodes.

For example, the node corresponding to the article a is directly connected to the node of the associated object B, and is indirectly connected to the node corresponding to the article C through the node corresponding to the associated object B; for another example, the commodity feature information of the commodity a includes "once purchased by the user U", and the object feature information of the user B includes "once browsed commodity B", so that a direct association between the commodity a and the user U, a direct association between the commodity B and the user U, and an indirect association between the commodity a and the commodity B can be established through the knowledge graph (that is, a node corresponding to the commodity a is directly connected to a node corresponding to the user U, a node corresponding to the commodity B is directly connected to a node corresponding to the user U, and a node corresponding to the commodity a is indirectly connected to a node corresponding to the commodity B through a node corresponding to the user U).

For another example, the commodity feature information of the commodity C includes "once appears in the movie E for multiple times", and the object feature information of the movie E includes "once appears in the commodity D for multiple times", so that a direct association between the commodity C and the movie E, a direct association between the commodity D and the movie E, and an indirect association between the commodity C and the commodity D can be established through the knowledge graph (that is, a node corresponding to the commodity C is directly connected to a node corresponding to the movie E, a node corresponding to the commodity D is directly connected to a node corresponding to the movie E, and a node corresponding to the commodity C is indirectly connected to a node corresponding to the commodity D through a node corresponding to the movie E).

For another example, the product characteristic information of the product a includes "dog '" appears in the product description for a plurality of times, "cat'" appears in the product description for a plurality of times, "common pets both with cats" are included in the object characteristic information of the associated object "dog", and "common pets both with dogs" are included in the object characteristic information of the associated object "cat", so that a direct association between the product a and the associated object "dog" can be established by the knowledge graph, a direct association between the product B and the associated object "cat", a direct association between the associated object "dog" and the associated object "cat", an indirect association between the product a and the product B (that is, a node corresponding to the product a is directly connected to a node corresponding to the associated object "dog", a node corresponding to the product B is directly connected to a node corresponding to the associated object "cat", and a node corresponding to the product a and a node corresponding to the product B dog are directly connected to a node corresponding to the associated object "dog" through a node corresponding to the associated object "dog The nodes corresponding to the joint object cat are indirectly connected).

In some embodiments, the connection relationship may be a direct connection relationship between two nodes, for example, the node a and the node B are directly connected, or the connection relationship may also be an indirect connection relationship between two nodes, where the connection relationship includes the hop count corresponding to the connection between two nodes, for example, if the node a and the node B are directly connected, the hop count from the node a to the node B is 1, and if the node a is indirectly connected to the node C through the node B, the hop count from the node a to the node C is 2. In some embodiments, each node may have a connection relationship with only one node, or may have a connection relationship with multiple nodes at the same time. In some embodiments, only one connection relationship may exist between two nodes, or multiple connection relationships may exist simultaneously.

In some embodiments, the connection relationship characteristic information may be a set of multiple connection relationships between respective nodes in the knowledge-graph. For example, the commodity feature information of commodity a includes "once purchased by user U1", the object feature information of user U1 includes "once purchased commodity B", the knowledge graph includes three nodes corresponding to commodity a, user U1 and commodity B, respectively, and connection relationship feature information corresponding to the three nodes is obtained based on the knowledge graph, the connection relationship feature information is used to indicate that commodity a and user U1 have a direct connection relationship of "purchased", user U1 and commodity B have a direct connection relationship of "purchase", and commodity a and commodity B have an indirect connection relationship (an indirect connection relationship means that two nodes are not directly connected, but are connected through one or more other nodes, as in this example, commodity a and commodity B are connected through user U1).

In some embodiments, the connection relationship between the two nodes is directional, such as the connection relationship "purchased" from the commodity a to the user U1 in the above example is from the node corresponding to the commodity a to the node corresponding to the user U1, and the connection relationship "purchased" from the user U1 to the commodity B is from the node corresponding to the user U1 to the node corresponding to the commodity B.

In step S12, the network device obtains a commodity detection model through training according to the commodity feature information, the object feature information, the connection relationship feature information, and commodity calibration information corresponding to the plurality of commodities, where the commodity calibration information is used to calibrate whether each commodity in the plurality of commodities is an unqualified commodity.

In some embodiments, the non-conforming goods include, but are not limited to, counterfeit goods, illegal goods, and the like. In some embodiments, input commodity calibration information corresponding to the commodity is received. In some embodiments, at least one calibrated commodity having a connection relation with the commodity is obtained from the knowledge graph, and the commodity is calibrated and corresponding commodity calibration information is obtained according to similarity information between the commodity feature information corresponding to the commodity and the commodity feature information corresponding to the at least one calibrated commodity. In some embodiments, the commodity calibration information is manually tagged to the commodity by the model training personnel.

In some embodiments, if at least one calibrated commodity in a connected relationship with the commodity in the knowledge graph has similar commodity characteristic information such as similar price, similar purchase amount, similar browsing amount, and the like with the commodity, whether the commodity is an unqualified commodity can be calibrated according to whether the calibrated commodity is an unqualified commodity, for example, if the calibrated commodity is an unqualified commodity, the commodity can be calibrated as also an unqualified commodity, and for example, if the calibrated commodity is a qualified commodity, the commodity can also be calibrated as well as a qualified commodity.

In some embodiments, a plurality of commodity samples and commodity calibration information corresponding to each commodity sample are collected, object feature information corresponding to an associated object associated with each commodity sample and connection relationship feature information obtained through a knowledge graph are obtained, a commodity detection model is obtained through training based on the data, a commodity detection model can be generated through training based on the data, a current commodity detection model can be updated through training based on the data, and the updated current commodity detection model is used as the commodity detection model. In some embodiments, the input of the commodity detection model is commodity characteristic information corresponding to a certain commodity and commodity detection information indicating whether the commodity is a disqualified commodity is output.

In step S13, the network device inputs the target product feature information corresponding to the target product into the product detection model, and obtains the product detection information corresponding to the target product and output by the product detection model, where the product detection information is used to indicate whether the target product is an unqualified product. In some embodiments, the article detection information includes indication information indicating whether the target article is a non-conforming article, such as "1" indicating that the article is a conforming article and "0" indicating that the article is a non-conforming article. In some embodiments, the article detection information includes probability information that the target article is a non-conforming article, such as the article detection information indicating that the target article has a 70% probability of being a non-conforming article, such as the article detection information indicating that the target article has a 30% probability of being a conforming article.

According to the commodity detection model and the commodity detection method, the knowledge graph can be constructed according to commodity feature information corresponding to a plurality of commodities and object feature information corresponding to the associated objects associated with the commodity feature information, and then the commodity detection model is trained according to the commodity feature information, the object feature information, connection relation feature information corresponding to each node in the knowledge graph and commodity calibration information.

In some embodiments, the step S11 is preceded by a step S14 (not shown). In step S14, the network device constructs the knowledge graph based on the product feature information corresponding to the plurality of products and the object feature information corresponding to the related object associated with the product feature information.

In some embodiments, the obtaining a commodity detection model by training includes at least one of:

1) generating the commodity detection model through training

In some embodiments, a commodity detection model is generated through training based on commodity feature information corresponding to a plurality of commodities required for building a knowledge graph, object feature information corresponding to an associated object associated with the commodity feature information, and connection relation feature information obtained from the knowledge graph.

2) Updating the current commodity detection model through training, and taking the updated current commodity detection model as the commodity detection model

In some embodiments, in response to a newly added commodity event for a constructed current knowledge graph, at least one newly added commodity corresponding to the newly added commodity event, commodity feature information corresponding to the at least one newly added commodity, and object feature information corresponding to an associated object associated with the commodity feature information are obtained, and the current knowledge graph is reconstructed, or in response to a newly added commodity feature information event for an existing commodity node in the constructed current knowledge graph, at least one existing commodity corresponding to the newly added commodity feature information event, newly added commodity feature information corresponding to the at least one existing commodity, and object feature information corresponding to an associated object associated with the newly added commodity feature information are obtained, and the current knowledge graph is reconstructed. In some embodiments, the latest connection relation feature information is obtained from the reconstructed latest knowledge graph, and the currently obtained commodity detection model is updated through training based on the data, and the updated current commodity detection model is used as the commodity detection model. In some embodiments, the add-on merchandise event may be initiated manually or automatically when a predetermined condition is met. In some embodiments, the currently obtained merchandise detection model is continuously updated, so as to continuously improve the detection accuracy and the detection efficiency of the merchandise detection model.

In some embodiments, the method further comprises: and the network equipment acquires the commodity calibration information corresponding to each commodity in the plurality of commodities. In some embodiments, the network device obtains the product calibration information corresponding to each product sent by the other devices. In some embodiments, the network appliance receives commodity calibration information entered by a model trainer or other operator for each commodity. In some embodiments, at least one calibrated commodity having a connection relationship with the commodity is obtained from the knowledge graph, and if the similarity between the commodity feature information corresponding to the commodity and the commodity feature information corresponding to the at least one calibrated commodity meets a predetermined similarity threshold, the commodity calibration information corresponding to the commodity is determined according to the commodity calibration information of the at least one calibrated commodity. In some embodiments, the commodity calibration information corresponding to the commodity may be input from other devices, or may be directly input by a human. For example, a model trainer manually determines commodity calibration information corresponding to a commodity and inputs the commodity calibration information to a network device, and the network device receives the commodity calibration information corresponding to the commodity input by the model trainer.

In some embodiments, the obtaining of the product calibration information corresponding to the product includes: obtaining at least one calibrated commodity which has a connection relation with the commodity from the knowledge graph; and if the similarity between the commodity characteristic information corresponding to the commodity and the commodity characteristic information corresponding to the at least one calibrated commodity meets a preset similarity threshold, determining the commodity calibration information corresponding to the commodity according to the commodity calibration information of the at least one calibrated commodity.

In some embodiments, at least one calibrated commodity having a direct connection to the commodity is obtained from the knowledge-graph. In some embodiments, at least one calibrated commodity having an indirect connection to the commodity and having a hop count between the commodity and the calibrated commodity less than a predetermined hop count threshold is obtained from the knowledge-graph. In some embodiments, if only one calibrated commodity exists in the knowledge graph and the similarity between the commodity feature information corresponding to the commodity and the commodity feature information corresponding to the calibrated commodity meets a predetermined similarity threshold, the commodity calibration information corresponding to the calibrated commodity may be directly determined as the commodity calibration information corresponding to the commodity. In some embodiments, the similarity includes, but is not limited to, a commodity price similarity, a commodity purchase quantity similarity, a commodity browsing quantity similarity, and the like.

In some embodiments, if a plurality of calibrated commodities which have a connection relationship with the commodity exist in the knowledge graph, determining whether calibrated commodities of which the similarity with the commodity meets a predetermined similarity threshold exist in the plurality of calibrated commodities, if so, directly determining the calibrated commodities of which the similarity with the commodity meets the predetermined similarity threshold as target calibrated commodities, or selecting calibrated commodities with higher similarity from among the calibrated commodities of which the similarity with the commodity meets the predetermined similarity threshold as target calibrated commodities, if only one target calibrated commodity exists, directly determining commodity calibration information corresponding to the target calibrated commodity as commodity calibration information corresponding to the commodity, if a plurality of target calibrated commodities exist and the commodity calibration information of the plurality of target calibrated commodities is consistent, the commodity calibration information corresponding to the commodity can be directly used as the commodity calibration information corresponding to the commodity, if a plurality of target calibrated commodities exist and the commodity calibration information of the plurality of target calibrated commodities is inconsistent, the commodity calibration information corresponding to the commodity can be determined according to the ratio between the unqualified commodity and the qualified commodity in the plurality of target calibrated commodities (for example, when the ratio is greater than a predetermined ratio threshold value, the commodity calibration information corresponding to the commodity is determined to indicate that the commodity is unqualified commodity, otherwise, the commodity calibration information corresponding to the commodity is determined to indicate that the commodity is qualified commodity), or if the plurality of target marked commodities only have unqualified commodities, directly determining the commodity marking information corresponding to the commodities to indicate that the commodities are unqualified commodities, or, if only qualified commodities exist in the plurality of target marked commodities, directly determining the commodity marking information corresponding to the commodities to indicate that the commodities are qualified commodities.

In some embodiments, the step S14 is preceded by: the network equipment acquires commodity feature information corresponding to each commodity in the plurality of commodities, determines a related object related to the commodity feature information according to the commodity feature information, and acquires object feature information corresponding to the related object. In some embodiments, for each commodity, after the commodity-related information corresponding to the commodity is collected locally or on a network, the commodity-related information is any information related to the commodity, and then the commodity feature information corresponding to the commodity may be determined directly according to the commodity-related information, or the commodity feature information corresponding to the commodity may be determined after feature extraction is performed on the commodity-related information.

In some embodiments, for each commodity, one or more objects are extracted from the commodity characteristic information corresponding to the commodity, all or part of the one or more objects are determined as associated objects associated with the commodity characteristic information, and optionally, at least one object having a certain relation with the commodity is selected from the one or more objects, wherein the object having a certain relation with the commodity may be an object having a semantic containing or contained relation with the commodity (such as "pet" and "dog"), an object capable of being used in a set with the commodity (such as "charger" and "charging wire"), and the like.

In some embodiments, according to the semantic content of the product characteristic information, an object associated with the semantic content is determined as an associated object associated with the product characteristic information, for example, the semantic content of the product characteristic information of the product a includes "high-end pet food brand", then the objects "cat" and "dog" associated with the semantic content may be determined as associated objects associated with the product characteristic information, that is, the determined associated objects are not objects directly included in the product characteristic information, thereby enabling more comprehensive association.

In some embodiments, for the associated object, after the object related information corresponding to the associated object is collected locally or on a network, the object related information is any information related to the associated object, and then the object feature information corresponding to the associated object may be determined directly according to the object related information, or the object feature information corresponding to the associated object may be determined after feature extraction is performed on the object related information. In some embodiments, if the associated object is a commodity, the commodity feature information corresponding to the commodity may be directly used as the corresponding object feature information.

In some embodiments, for each of the plurality of commodities, the commodity feature information corresponding to the commodity includes one or more objects, where determining, according to the commodity feature information, the associated object associated with the commodity feature information includes: and taking at least one object in the one or more objects as a related object related to the commodity characteristic information. In some embodiments, the one or more objects may be directly used as the associated objects associated with the characteristic information of the product. In some embodiments, if the product characteristic information corresponding to the product includes a plurality of objects, at least one object may be determined from the plurality of objects as the associated object associated with the product characteristic information.

In some embodiments, the regarding at least one of the one or more objects as an associated object associated with the characteristic information of the commodity includes: and determining at least one object from the one or more objects, and using the at least one object as a related object associated with the commodity characteristic information. In some embodiments, at least one object, of the one or more objects, with which the degree of association with the commodity is greater than a predetermined degree of association is selected as the associated object with which the characteristic information of the commodity is associated. In some embodiments, the object with the highest access rate or click rate is determined from the one or more objects as the associated object associated with the characteristic information of the commodity. In some embodiments, at least one object is determined from the one or more objects according to a specific gravity of each of the one or more objects in the merchandise characteristic information, wherein the specific gravity of each of the at least one object in the merchandise characteristic information satisfies a predetermined specific gravity threshold.

In some embodiments, said determining at least one object from said one or more objects comprises: and determining at least one object from the one or more objects according to the specific gravity of each object in the commodity characteristic information, wherein the specific gravity of each object in the at least one object in the commodity characteristic information meets a preset specific gravity threshold value. In some embodiments, the specific gravity of an object in the commodity feature information is used for representing the importance degree of the object in the commodity feature information, and the importance degree can reflect the influence degree of the object on the use or sale of the commodity to a certain degree. In some embodiments, at least one object with a corresponding specific gravity greater than or equal to a predetermined specific gravity threshold is determined from the one or more objects according to the specific gravity of each object in the characteristic information of the commodity.

In some embodiments, the method further comprises: and the network equipment determines the proportion of each object in the commodity characteristic information according to the occurrence frequency of the object in the commodity characteristic information. In some embodiments, the higher the number of occurrences of an object in the product characteristic information, the higher the specific gravity of the object in the product characteristic information, and vice versa. In some embodiments, the specific gravity of the object in the product characteristic information is further adjusted by combining the appearance position of the object in the product characteristic information, for example, if an object appears in the product characteristic information multiple times and most of the object appears in the product description information of the product, the specific gravity of the object in the product characteristic information is increased, and optionally, different weighting coefficients can be set for different appearance positions, so as to adjust the specific gravity of the object in the product characteristic information.

In some embodiments, the method further comprises: and the network equipment determines the specific gravity of each object in the commodity characteristic information according to the semantic importance degree of the object in the commodity characteristic information. In some embodiments, the semantic importance level can reflect the association between the object and the commodity to a certain extent, and the higher the semantic importance level is, the higher the association between the object and the commodity is. In some embodiments, the higher the semantic importance of an object in the commodity feature information, the higher the weight of the object in the commodity feature information, and vice versa.

In some embodiments, the determining, according to the article characteristic information, the associated object associated with the article characteristic information includes: obtaining semantic content of the commodity feature information; and according to the semantic content, determining an object associated with the semantic content as an associated object associated with the commodity feature information. In some embodiments, the semantic content of the commodity feature information is obtained by performing semantic analysis on the commodity feature information. In some embodiments, one or more keywords in the semantic content are obtained, and an object associated with the one or more keywords is determined as an associated object associated with the commodity characteristic information; for example, if the semantic content of the product characteristic information of the product a includes "high-end pet food brand", the keyword "pet" in the semantic content may be obtained, and the objects "cat" and "dog" associated with the keyword "pet" are determined as the associated objects associated with the product characteristic information.

In some embodiments, the method further comprises: for each associated object, the network equipment determines a secondary associated object associated with the associated object according to the object characteristic information corresponding to the associated object, and obtains the object characteristic information corresponding to the secondary associated object; wherein the step S11 includes: and constructing a knowledge graph according to the commodity feature information corresponding to a plurality of commodities, the object feature information corresponding to the related object associated with the commodity feature information and the object feature information corresponding to the secondary related object associated with the related object. In some embodiments, the secondary associated objects are obtained by performing one or more association operations on associated objects associated with the commodity characteristic information, and one associated object may be associated with one or more secondary associated objects. In some embodiments, a secondary associated object associated with the associated object may be determined based on a predetermined association relationship. The implementation manner of obtaining the object feature information corresponding to the secondary associated object is the same as or similar to the implementation manner of obtaining the object feature information corresponding to the associated object, and is not described herein again. In some embodiments, a knowledge graph is constructed according to commodity feature information corresponding to a plurality of commodities, object feature information corresponding to an associated object associated with the commodity feature information, and object feature information corresponding to a secondary associated object associated with the associated object, and the constructed knowledge graph includes a plurality of nodes, each node corresponds to one commodity or one associated object or one secondary associated object, so that a node corresponding to a commodity may be indirectly connected to one secondary associated object corresponding to the associated object through one associated object corresponding to the commodity, and may also be indirectly connected to other secondary associated objects corresponding to the associated object through one associated object corresponding to the commodity and at least one secondary associated object corresponding to the associated object. In some embodiments, the secondary associated object corresponding to the associated object may be any object included in the object characteristic information corresponding to the associated object. In some embodiments, an object associated with the semantic content may be determined as an associated object associated with the associated object according to the semantic content of the object feature information corresponding to the associated object. In some embodiments, the secondary associated object may be any object in any form, preferably including, but not limited to, a merchant object, a user object, a merchandise object, and the like. It should be noted that any explanatory explanation regarding the related objects in the foregoing embodiments can be applied to the secondary related objects.

In some embodiments, the determining the secondary associated object corresponding to the associated object according to the object characteristic information corresponding to the associated object includes step S15 (not shown), step S16 (not shown), and step S17 (not shown). In step S15, the network device performs a correlation operation on the correlation object according to the object feature information corresponding to the correlation object, to obtain one or more secondary correlation objects; in step S16, the network device performs the association operation again for each secondary associated object obtained by the association operation this time according to the object feature information corresponding to the secondary associated object, so as to obtain one or more secondary associated objects; in step S17, the network device repeats said step S16 until the stop association condition is satisfied. In some embodiments, the association operation refers to an operation for determining a secondary association object of an object (which may be an association object corresponding to the characteristic information of the product or an obtained secondary association object), where one or more secondary association objects may be obtained by performing one association operation. In some embodiments, after performing the association operation with respect to the object characteristic information corresponding to the associated object and obtaining one or more secondary associated objects, at least one association operation may be performed again to obtain more secondary associated objects, for example, the associated object corresponding to the article characteristic information of the article a is "dog", then, the association operation is executed to the object characteristic information of the associated object to obtain the secondary associated object cat corresponding to the dog, and then, performing association operation on the object characteristic information corresponding to the cat to obtain a secondary associated object movie M corresponding to the cat, and then, executing the association operation to the object characteristic information corresponding to the movie M to obtain a secondary association object English short corresponding to the movie M, and the like until the association stopping condition is met, so that more comprehensive secondary associated objects can be obtained. In some embodiments, the stop association condition includes any condition that triggers a stop of the execution of the association operation.

In some embodiments, the disassociation condition comprises any one of: the execution times of the associated operation reach a preset time threshold; the obtained number of secondary associated objects reaches a predetermined number threshold. In some embodiments, the number threshold may be set based on experience, optionally, the number threshold may be adjusted based on user feedback information. In some embodiments, after the last association operation, the number of obtained secondary associated objects (that is, the total number of secondary associated objects obtained by multiple association operations that have been performed) is less than a predetermined number threshold, and after the present association operation, the number of obtained secondary associated objects may be equal to or exceed the predetermined number threshold, and then the stop association condition is considered to be satisfied.

In some embodiments, the method further comprises: and if the commodity detection information indicates that the target commodity is an unqualified commodity, the network equipment outputs at least one calibrated unqualified commodity which has a connection relation with the target commodity. In some embodiments, the at least one rejected as calibrated commodity is obtained from a knowledge-graph, and the rejected as calibrated commodity may be directly connected with the target commodity or indirectly connected with the target commodity. For example, if the product detection information output after the product a is input to the product detection model indicates that the product a is a defective product, calibrated defective products B and C directly connected to the product a are output. Optionally, at least one calibrated qualified commodity having a connection relation with the target commodity can be output for further comparison or processing.

In some embodiments, if there are a plurality of calibrated unqualified commodities which have connection relations with the target commodity; wherein the method further comprises: and the network equipment determines at least one unqualified marked commodity from the plurality of unqualified marked commodities, wherein the connection hop count corresponding to the connection relation between each unqualified marked commodity in the at least one unqualified marked commodity and the target commodity is less than or equal to a preset hop count threshold value. In some embodiments, a plurality of unqualified marked commodities which have connection relations with the target commodity are obtained from the knowledge-graph, a connection hop count corresponding to the connection relation between the target commodity and each unqualified commodity is obtained, and then at least one unqualified commodity with the connection hop count smaller than or equal to a preset hop count threshold value is selected from the plurality of unqualified commodities. In some embodiments, the hop count threshold may be set based on experience, and optionally may be adjusted based on feedback information for merchandise detection information.

FIG. 2 shows a flowchart of a method for constructing a knowledge-graph according to one embodiment of the present application, the method including step S21. In step S21, the network device constructs a knowledge graph according to commodity feature information corresponding to a plurality of commodities and object feature information corresponding to an associated object associated with the commodity feature information, where the knowledge graph includes a plurality of nodes, and each node in the knowledge graph corresponds to one commodity or one associated object. The related operations in this embodiment have been described in detail in the foregoing embodiments, and are not described herein again.

FIG. 3 is a flowchart illustrating a method for obtaining a merchandise detection model through training according to an embodiment of the present application, wherein the method includes steps S31 and S32. In step S31, the network device obtains connection characteristic information corresponding to a knowledge graph, where the knowledge graph is constructed according to commodity characteristic information corresponding to a plurality of commodities and object characteristic information corresponding to associated objects associated with the commodity characteristic information, the knowledge graph includes a plurality of nodes, each node in the knowledge graph corresponds to a commodity or an associated object, and the connection characteristic information is used to represent a connection relationship between nodes in the knowledge graph; in step S32, the network device obtains a commodity detection model through training according to the commodity feature information, the object feature information, the connection relationship feature information, and commodity calibration information corresponding to the plurality of commodities, where the commodity calibration information is used to calibrate whether each commodity in the plurality of commodities is an unqualified commodity. The related operations in this embodiment have been described in detail in the foregoing embodiments, and are not described herein again.

Fig. 4 shows a block diagram of a network device for detecting defective goods according to an embodiment of the present application, which includes a one-module 11, a two-module 12, and a three-module 13. A module 11, configured to obtain connection relationship feature information corresponding to a knowledge graph, where the knowledge graph is constructed according to commodity feature information corresponding to a plurality of commodities and object feature information corresponding to an associated object associated with the commodity feature information, the knowledge graph includes a plurality of nodes, each node in the knowledge graph corresponds to one commodity or one associated object, and the connection relationship feature information is used to represent a connection relationship between nodes in the knowledge graph; a second module 12, configured to obtain a commodity detection model through training according to the commodity feature information, the object feature information, the connection relationship feature information, and commodity calibration information corresponding to the multiple commodities, where the commodity calibration information is used to calibrate whether each commodity in the multiple commodities is an unqualified commodity; and a third module 13, configured to input target commodity feature information corresponding to a target commodity into the commodity detection model, so as to obtain commodity detection information output by the commodity detection model and corresponding to the target commodity, where the commodity detection information is used to indicate whether the target commodity is an unqualified commodity.

The module 11 is configured to obtain connection relationship feature information corresponding to a knowledge graph, where the knowledge graph is constructed according to commodity feature information corresponding to a plurality of commodities and object feature information corresponding to an associated object associated with the commodity feature information, the knowledge graph includes a plurality of nodes, each node in the knowledge graph corresponds to one commodity or one associated object, and the connection relationship feature information is used to represent a connection relationship between each node in the knowledge graph.

A second module 12, configured to obtain a commodity detection model through training according to the commodity feature information, the object feature information, the connection relationship feature information, and commodity calibration information corresponding to the multiple commodities, where the commodity calibration information is used to calibrate whether each commodity in the multiple commodities is an unqualified commodity.

And a third module 13, configured to input target commodity feature information corresponding to a target commodity into the commodity detection model, so as to obtain commodity detection information output by the commodity detection model and corresponding to the target commodity, where the commodity detection information is used to indicate whether the target commodity is an unqualified commodity. In some embodiments, the article detection information includes indication information indicating whether the target article is a non-conforming article, such as "1" indicating that the article is a conforming article and "0" indicating that the article is a non-conforming article. In some embodiments, the article detection information includes probability information that the target article is a non-conforming article, such as the article detection information indicating that the target article has a 70% probability of being a non-conforming article, such as the article detection information indicating that the target article has a 30% probability of being a conforming article.

In some embodiments, the apparatus further comprises a quad-module 14 (not shown). A fourth module 14, configured to construct the knowledge graph according to the commodity feature information corresponding to the multiple commodities and the object feature information corresponding to the associated object associated with the commodity feature information. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

1) generating the commodity detection model through training

Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, the apparatus is further configured to: and for each commodity in the plurality of commodities, obtaining commodity calibration information corresponding to the commodity. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, the obtaining of the product calibration information corresponding to the product includes: obtaining at least one calibrated commodity which has a connection relation with the commodity from the knowledge graph; and if the similarity between the commodity characteristic information corresponding to the commodity and the commodity characteristic information corresponding to the at least one calibrated commodity meets a preset similarity threshold, determining the commodity calibration information corresponding to the commodity according to the commodity calibration information of the at least one calibrated commodity. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, the apparatus is further configured to: for each commodity in the plurality of commodities, acquiring commodity feature information corresponding to the commodity, determining a related object related to the commodity feature information according to the commodity feature information, and acquiring object feature information corresponding to the related object. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, for each of the plurality of commodities, the commodity feature information corresponding to the commodity includes one or more objects, where determining, according to the commodity feature information, the associated object associated with the commodity feature information includes: and taking at least one object in the one or more objects as a related object related to the commodity characteristic information. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, the regarding at least one of the one or more objects as an associated object associated with the characteristic information of the commodity includes: and determining at least one object from the one or more objects, and using the at least one object as a related object associated with the commodity characteristic information. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, said determining at least one object from said one or more objects comprises: and determining at least one object from the one or more objects according to the specific gravity of each object in the commodity characteristic information, wherein the specific gravity of each object in the at least one object in the commodity characteristic information meets a preset specific gravity threshold value. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, the apparatus is further configured to: and determining the specific gravity of each object in the commodity characteristic information according to the occurrence frequency of the object in the commodity characteristic information. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, the apparatus is further configured to: and determining the specific gravity of each object in the commodity feature information according to the semantic importance degree of each object in the commodity feature information. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, the determining, according to the article characteristic information, the associated object associated with the article characteristic information includes: obtaining semantic content of the commodity feature information; and according to the semantic content, determining an object associated with the semantic content as an associated object associated with the commodity feature information. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, the apparatus is further configured to: for each associated object, determining a secondary associated object associated with the associated object according to the object characteristic information corresponding to the associated object, and obtaining the object characteristic information corresponding to the secondary associated object; wherein the step S11 includes: and constructing a knowledge graph according to the commodity feature information corresponding to a plurality of commodities, the object feature information corresponding to the related object associated with the commodity feature information and the object feature information corresponding to the secondary related object associated with the related object. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, the determining the secondary related object corresponding to the related object according to the object characteristic information corresponding to the related object includes a five-module 15 (not shown), a six-module 16 (not shown), and a seven-module 17 (not shown). A fifth module 15, configured to perform a correlation operation on the correlation object according to object feature information corresponding to the correlation object, so as to obtain one or more secondary correlation objects; a sixth module 16, configured to execute the association operation again according to the object feature information corresponding to the secondary association object for each secondary association object obtained by the association operation of this time, so as to obtain one or more secondary association objects; a seventh module 17, configured to trigger the sixth module 16 to repeatedly perform the operation until the stop association condition is satisfied. Here, the specific implementation of the five-module 15, the six-module 16 and the seven-module 17 is the same as or similar to the embodiment related to steps S15, S16 and S17 in fig. 1, and therefore, the detailed description thereof is omitted, and the detailed implementation is incorporated herein by reference.

In some embodiments, the disassociation condition comprises any one of: the execution times of the associated operation reach a preset time threshold; the obtained number of secondary associated objects reaches a predetermined number threshold. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, the apparatus is further configured to: and if the commodity detection information indicates that the target commodity is an unqualified commodity, outputting at least one calibrated unqualified commodity which has a connection relation with the target commodity. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, if there are a plurality of calibrated unqualified commodities which have connection relations with the target commodity; wherein the device is further configured to: determining at least one unqualified marked commodity from the plurality of unqualified marked commodities, wherein the connection hop count corresponding to the connection relationship between each unqualified marked commodity in the at least one unqualified marked commodity and the target commodity is less than or equal to a preset hop count threshold value. Here, the related operations are the same as or similar to those of the embodiment shown in fig. 1, and therefore are not described again, and are included herein by reference.

In some embodiments, as shown in FIG. 5, the system 300 can be implemented as any of the devices in the various embodiments described. In some embodiments, system 300 may include one or more computer-readable media (e.g., system memory or NVM/storage 320) having instructions and one or more processors (e.g., processor(s) 305) coupled with the one or more computer-readable media and configured to execute the instructions to implement modules to perform the actions described herein.

For one embodiment, system control module 310 may include any suitable interface controllers to provide any suitable interface to at least one of processor(s) 305 and/or any suitable device or component in communication with system control module 310.

The system control module 310 may include a memory controller module 330 to provide an interface to the system memory 315. Memory controller module 330 may be a hardware module, a software module, and/or a firmware module.

System memory 315 may be used, for example, to load and store data and/or instructions for system 300. For one embodiment, system memory 315 may include any suitable volatile memory, such as suitable DRAM. In some embodiments, the system memory 315 may include a double data rate type four synchronous dynamic random access memory (DDR4 SDRAM).

For one embodiment, system control module 310 may include one or more input/output (I/O) controllers to provide an interface to NVM/storage 320 and communication interface(s) 325.

For example, NVM/storage 320 may be used to store data and/or instructions. NVM/storage 320 may include any suitable non-volatile memory (e.g., flash memory) and/or may include any suitable non-volatile storage device(s) (e.g., one or more Hard Disk Drives (HDDs), one or more Compact Disc (CD) drives, and/or one or more Digital Versatile Disc (DVD) drives).

NVM/storage 320 may include storage resources that are physically part of the device on which system 300 is installed or may be accessed by the device and not necessarily part of the device. For example, NVM/storage 320 may be accessible over a network via communication interface(s) 325.

Communication interface(s) 325 may provide an interface for system 300 to communicate over one or more networks and/or with any other suitable device. System 300 may wirelessly communicate with one or more components of a wireless network according to any of one or more wireless network standards and/or protocols.

For one embodiment, at least one of the processor(s) 305 may be packaged together with logic for one or more controller(s) (e.g., memory controller module 330) of the system control module 310. For one embodiment, at least one of the processor(s) 305 may be packaged together with logic for one or more controller(s) of the system control module 310 to form a System In Package (SiP). For one embodiment, at least one of the processor(s) 305 may be integrated on the same die with logic for one or more controller(s) of the system control module 310. For one embodiment, at least one of the processor(s) 305 may be integrated on the same die with logic for one or more controller(s) of the system control module 310 to form a system on a chip (SoC).

In various embodiments, system 300 may be, but is not limited to being: a server, a workstation, a desktop computing device, or a mobile computing device (e.g., a laptop computing device, a holding computing device, a tablet, a netbook, etc.). In various embodiments, system 300 may have more or fewer components and/or different architectures. For example, in some embodiments, system 300 includes one or more cameras, a keyboard, a Liquid Crystal Display (LCD) screen (including a touch screen display), a non-volatile memory port, multiple antennas, a graphics chip, an Application Specific Integrated Circuit (ASIC), and speakers.

The present application also provides a computer readable storage medium having stored thereon computer code which, when executed, performs a method as in any one of the preceding.

The present application also provides a computer program product, which when executed by a computer device, performs the method of any of the preceding claims.

The present application further provides a computer device, comprising:

one or more processors;

a memory for storing one or more computer programs;

the one or more computer programs, when executed by the one or more processors, cause the one or more processors to implement the method of any preceding claim.

It should be noted that the present application may be implemented in software and/or a combination of software and hardware, for example, implemented using Application Specific Integrated Circuits (ASICs), general purpose computers or any other similar hardware devices. In one embodiment, the software programs of the present application may be executed by a processor to implement the steps or functions described above. Likewise, the software programs (including associated data structures) of the present application may be stored in a computer readable recording medium, such as RAM memory, magnetic or optical drive or diskette and the like. Additionally, some of the steps or functions of the present application may be implemented in hardware, for example, as circuitry that cooperates with the processor to perform various steps or functions.

In addition, some of the present application may be implemented as a computer program product, such as computer program instructions, which when executed by a computer, may invoke or provide methods and/or techniques in accordance with the present application through the operation of the computer. Those skilled in the art will appreciate that the form in which the computer program instructions reside on a computer-readable medium includes, but is not limited to, source files, executable files, installation package files, and the like, and that the manner in which the computer program instructions are executed by a computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Computer-readable media herein can be any available computer-readable storage media or communication media that can be accessed by a computer.

Communication media includes media by which communication signals, including, for example, computer readable instructions, data structures, program modules, or other data, are transmitted from one system to another. Communication media may include conductive transmission media such as cables and wires (e.g., fiber optics, coaxial, etc.) and wireless (non-conductive transmission) media capable of propagating energy waves such as acoustic, electromagnetic, RF, microwave, and infrared. Computer readable instructions, data structures, program modules, or other data may be embodied in a modulated data signal, for example, in a wireless medium such as a carrier wave or similar mechanism such as is embodied as part of spread spectrum techniques. The term "modulated data signal" means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. The modulation may be analog, digital or hybrid modulation techniques.

By way of example, and not limitation, computer-readable storage media may include volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules or other data. For example, computer-readable storage media include, but are not limited to, volatile memory such as random access memory (RAM, DRAM, SRAM); and non-volatile memory such as flash memory, various read-only memories (ROM, PROM, EPROM, EEPROM), magnetic and ferromagnetic/ferroelectric memories (MRAM, FeRAM); and magnetic and optical storage devices (hard disk, tape, CD, DVD); or other now known media or later developed that can store computer-readable information/data for use by a computer system.

An embodiment according to the present application comprises an apparatus comprising a memory for storing computer program instructions and a processor for executing the program instructions, wherein the computer program instructions, when executed by the processor, trigger the apparatus to perform a method and/or a solution according to the aforementioned embodiments of the present application.

It will be evident to those skilled in the art that the present application is not limited to the details of the foregoing illustrative embodiments, and that the present application may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned. Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the apparatus claims may also be implemented by one unit or means in software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.

Claims

1. A method of detecting a rejected good, wherein the method comprises:

2. The method of claim 1, wherein the obtaining of the connection relation feature information corresponding to the knowledge-graph further comprises:

and constructing the knowledge graph according to the commodity feature information corresponding to the commodities and the object feature information corresponding to the associated object associated with the commodity feature information.

3. The method of claim 2, wherein the constructing the knowledge graph according to the commodity feature information corresponding to the plurality of commodities and the object feature information corresponding to the associated object associated with the commodity feature information further comprises:

for each commodity in the plurality of commodities, acquiring commodity feature information corresponding to the commodity, determining a related object related to the commodity feature information according to the commodity feature information, and acquiring object feature information corresponding to the related object.

4. The method of claim 3, wherein, for each of the plurality of commodities, the commodity characteristic information corresponding to the commodity comprises one or more objects;

wherein, the determining the associated object associated with the commodity feature information according to the commodity feature information includes:

and taking at least one object in the one or more objects as a related object related to the commodity characteristic information.

5. The method according to claim 4, wherein the taking at least one of the one or more objects as the associated object associated with the commodity feature information comprises:

and determining at least one object from the one or more objects, and using the at least one object as a related object associated with the commodity characteristic information.

6. The method of claim 5, wherein the determining at least one object from the one or more objects comprises:

and determining at least one object from the one or more objects according to the specific gravity of each object in the commodity characteristic information, wherein the specific gravity of each object in the at least one object in the commodity characteristic information meets a preset specific gravity threshold value.

7. The method of claim 6, wherein the method further comprises:

and determining the specific gravity of each object in the commodity characteristic information according to the occurrence frequency of the object in the commodity characteristic information.

8. The method of claim 6, wherein the method further comprises:

and determining the specific gravity of each object in the commodity feature information according to the semantic importance degree of each object in the commodity feature information.

9. The method according to claim 3, wherein the determining the associated object associated with the commodity characteristic information according to the commodity characteristic information comprises:

obtaining semantic content of the commodity feature information;

and according to the semantic content, determining an object associated with the semantic content as an associated object associated with the commodity feature information.

10. The method of claim 2, wherein the method further comprises:

for each associated object, determining a secondary associated object associated with the associated object according to the object characteristic information corresponding to the associated object, and obtaining the object characteristic information corresponding to the secondary associated object;

the constructing the knowledge graph according to the commodity feature information corresponding to the plurality of commodities and the object feature information corresponding to the associated object associated with the commodity feature information includes:

and constructing the knowledge graph according to the commodity feature information corresponding to a plurality of commodities, the object feature information corresponding to the associated object associated with the commodity feature information and the object feature information corresponding to the secondary associated object associated with the associated object.

11. The method according to claim 10, wherein determining the secondary associated object corresponding to the associated object according to the object feature information corresponding to the associated object comprises:

according to the object characteristic information corresponding to the associated object, performing associated operation on the associated object to obtain one or more secondary associated objects;

for each secondary associated object obtained by the current associated operation, executing the associated operation again according to the object characteristic information corresponding to the secondary associated object to obtain one or more secondary associated objects;

and repeating the step of executing the association operation again until the association stopping condition is met.

12. The method of claim 11, wherein the stop association condition comprises any one of:

the execution times of the associated operation reach a preset time threshold;

the obtained number of secondary associated objects reaches a predetermined number threshold.

13. The method of claim 1, wherein the obtaining a commodity detection model through training comprises at least one of:

generating the commodity detection model through training;

and updating the current commodity detection model through training, and taking the updated current commodity detection model as the commodity detection model.

14. The method of claim 1, wherein the method further comprises:

and for each commodity in the plurality of commodities, obtaining commodity calibration information corresponding to the commodity.

15. The method of claim 14, wherein the obtaining of the product calibration information corresponding to the product comprises:

obtaining at least one calibrated commodity which has a connection relation with the commodity from the knowledge graph;

and if the similarity between the commodity characteristic information corresponding to the commodity and the commodity characteristic information corresponding to the at least one calibrated commodity meets a preset similarity threshold, determining the commodity calibration information corresponding to the commodity according to the commodity calibration information of the at least one calibrated commodity.

16. The method of claim 1, wherein the method further comprises:

and if the commodity detection information indicates that the target commodity is an unqualified commodity, outputting at least one calibrated unqualified commodity which has a connection relation with the target commodity.

17. The method of claim 16, wherein if there are a plurality of the unqualified goods which have connection relations with the target goods;

wherein the method further comprises:

determining at least one unqualified marked commodity from the plurality of unqualified marked commodities, wherein the connection hop count corresponding to the connection relationship between each unqualified marked commodity in the at least one unqualified marked commodity and the target commodity is less than or equal to a preset hop count threshold value.

18. An apparatus for detecting defective goods, wherein the apparatus comprises:

a processor; and

a memory arranged to store computer executable instructions that, when executed, cause the processor to perform the method of any of claims 1 to 17.

19. A computer-readable medium storing instructions that, when executed, cause a system to perform the operations of any of the methods of claims 1-17.