Recognition Method, Storage Medium and Processor for Delivery Address

Abstract

The present invention discloses a method for identifying a delivery address, a storage medium, and a processor. The method includes: obtaining semantic information of an order information to be identified, where the order information at least includes: information used to represent a delivery address recorded in the order; determining an abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; in response to the abnormal probability being greater than a probability threshold, identifying whether the delivery address is in an abnormal state based on a target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or, the status of the target delivery address is a normal state. The present invention solves the technical problem of low accuracy in identifying a delivery address.

Description

Technical Field

[0001] The present invention relates to the field of computers, and in particular, to a method for identifying a delivery address, a storage medium, and a processor. Background Art

[0002] Currently, there are cases where false or adversarial delivery addresses that cannot be delivered are written to achieve the purpose of fraudulently obtaining merchant items or causing losses to the platform's funds.

[0003] To reduce losses caused to buyers or the platform due to the above reasons, in related technologies, delivery addresses are usually identified through some simple rules, resulting in the technical problem of low accuracy in identifying delivery addresses.

[0004] In response to the above problems, no effective solution has been proposed yet. Summary of the Invention

[0005] Embodiments of the present invention provide a method for identifying a delivery address, a storage medium, and a processor to at least solve the technical problem of low accuracy in identifying delivery addresses.

[0006] According to one aspect of the embodiments of the present invention, a method for identifying a delivery address is provided. The method may include: obtaining semantic information of order information to be identified, where the order information at least includes: information used to represent the delivery address recorded in the order; determining an abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; in response to the abnormal probability being greater than a probability threshold, identifying whether the delivery address is in an abnormal state based on a target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or, the state of the target delivery address is a normal state.

[0007] According to one aspect of the embodiments of the present invention, another method for identifying a delivery address is provided. The method may include: obtaining an order to be processed from an e-commerce platform; obtaining semantic information of order information to be identified in the order, where the order information at least includes: information used to represent the delivery address recorded in the order; determining an abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; in response to the abnormal probability being greater than a probability threshold, identifying whether the delivery address is in an abnormal state based on a target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or, the state of the target delivery address is a normal state; in response to the delivery address being in an abnormal state, outputting a prompt message to the e-commerce platform, where the prompt message is used to indicate that the order placement fails.

[0008] According to one aspect of an embodiment of the present invention, another method for identifying a delivery address is provided. The method may include: in response to an input instruction acting on an operation interface, displaying order information to be identified on the operation interface, where the order information at least includes: information used to characterize the delivery address recorded in the order; in response to an identification instruction acting on the operation interface, displaying an identification result of the delivery address on the operation interface, where the identification result is used to indicate whether the delivery address is in an abnormal state, and when the abnormal probability of the delivery address is greater than a probability threshold, it is determined based on a target delivery address, the abnormal probability is determined based on the semantic information of the order information, and is used to indicate the possibility that the delivery address is in an abnormal state, the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state.

[0009] According to one aspect of an embodiment of the present invention, another method for identifying a delivery address is provided. The method may include: obtaining the semantic information of the order information to be identified by calling a first interface, where the first interface includes a first parameter, and the parameter value of the first parameter is the semantic information, and the order information at least includes: information used to characterize the delivery address recorded in the order; determining the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to indicate the possibility that the delivery address is in an abnormal state; in response to the abnormal probability being greater than a probability threshold, determining an identification result based on a target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state, and the identification result is used to indicate whether the delivery address is in an abnormal state; outputting the identification result by calling a second interface, where the second interface includes a second parameter, and the parameter value of the second parameter is the identification result.

[0010] According to one aspect of an embodiment of the present invention, an apparatus for identifying a delivery address is provided. The apparatus may include: a first obtaining unit configured to obtain the semantic information of the order information to be identified, where the order information at least includes: information used to characterize the delivery address recorded in the order; a first determining unit configured to determine the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to indicate the possibility that the delivery address is in an abnormal state; a first identifying unit configured to, in response to the abnormal probability being greater than a probability threshold, identify whether the delivery address is in an abnormal state based on a target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state.

[0011] According to one aspect of an embodiment of the present invention, another identifying device for a delivery address is provided. The device may include: a second obtaining unit, configured to: obtain an order to be processed from an e-commerce platform; a third obtaining unit, configured to obtain semantic information of order information to be identified in the order, where the order information at least includes: information for characterizing the delivery address recorded in the order; a second determining unit, configured to determine an abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; a second identifying unit, configured to, in response to the abnormal probability being greater than a probability threshold, identify whether the delivery address is in an abnormal state based on a target delivery address, where the placing time of the order corresponding to the target delivery address is associated with the placing time of the order corresponding to the delivery address, and / or, the state of the target delivery address is a normal state; a first output unit, configured to, in response to the delivery address being in an abnormal state, output a prompt message to the e-commerce platform, where the prompt message is used to indicate that the order placement fails.

[0012] According to one aspect of an embodiment of the present invention, another identifying device for a delivery address is provided. The device may include: a first display unit, configured to, in response to an input instruction acting on an operation interface, display the order information to be identified on the operation interface, where the order information at least includes: information for characterizing the delivery address recorded in the order; a second display unit, configured to, in response to an identification instruction acting on the operation interface, display an identification result of the delivery address on the operation interface, where the identification result is used to represent whether the delivery address is in an abnormal state, and when the abnormal probability of the delivery address is greater than a probability threshold, is determined based on a target delivery address, the abnormal probability is determined based on the semantic information of the order information, and is used to represent the possibility that the delivery address is in an abnormal state, the placing time of the order corresponding to the target delivery address is associated with the placing time of the order corresponding to the delivery address, and / or, the state of the target delivery address is a normal state.

[0013] According to one aspect of the embodiments of the present invention, another recognition device for a delivery address is provided. The device may include: a fourth acquisition unit configured to acquire semantic information of an order information to be recognized by invoking a first interface, where the first interface includes a first parameter, and the parameter value of the first parameter is the semantic information, and the order information at least includes: information for characterizing the delivery address recorded in the order; a third determination unit configured to determine an exception probability of the delivery address based on the semantic information, where the exception probability is used to represent the possibility that the delivery address is in an abnormal state; a fourth determination unit configured to, in response to the exception probability being greater than a probability threshold, determine a recognition result based on a target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state, and the recognition result is used to represent whether the delivery address is in an abnormal state; a second output unit configured to output the recognition result by invoking a second interface, where the second interface includes a second parameter, and the parameter value of the second parameter is the recognition result.

[0014] According to another aspect of the embodiments of the present invention, a computer-readable storage medium is further provided. The computer-readable storage medium includes a stored program, where, when the program runs, it controls the device where the storage medium is located to execute the recognition method for the delivery address in any one of the above.

[0015] According to another aspect of the embodiments of the present invention, a processor is further provided. The processor is configured to run a program, where, when the program runs, it executes the recognition method for the delivery address in any one of the above.

[0016] In the embodiments of the present invention, semantic information of order information to be recognized is acquired, where the order information at least includes: information for characterizing the delivery address recorded in the order; an exception probability of the delivery address is determined based on the semantic information, where the exception probability is used to represent the possibility that the delivery address is in an abnormal state; in response to the exception probability being greater than a probability threshold, it is recognized whether the delivery address is in an abnormal state based on a target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state. That is to say, the present invention determines the exception probability of the delivery address according to the semantic information in the delivery address, and then recognizes the delivery address through the target delivery address to determine whether the delivery address is in an abnormal state. Since the present invention effectively performs recognition at the semantic level of the delivery address, the technical effect of improving the accuracy of recognizing the delivery address is achieved, and thus the technical problem of low accuracy in recognizing the delivery address is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1 It is a hardware structure block diagram of a computer terminal (or mobile device) of a method for identifying a delivery address according to an embodiment of the present invention;

[0019] Figure 2 is a flow chart of a method for identifying a delivery address according to an embodiment of the present invention;

[0020] Figure 3 is a flow chart of another method for identifying a delivery address according to an embodiment of the present invention;

[0021] Figure 4 is a flow chart of another method for identifying a delivery address according to an embodiment of the present invention;

[0022] Figure 5 is a flow chart of another method for identifying a delivery address according to an embodiment of the present invention;

[0023] Figure 6 is a flow chart of a false adversarial address determination method according to an embodiment of the present invention;

[0024] Figure 7 is a schematic diagram of language model prediction according to an embodiment of the present invention;

[0025] Figure 8 is a schematic diagram of language model pre-training according to an embodiment of the present invention;

[0026] Figure 9 is a schematic diagram of shortening the distance between word vectors according to an embodiment of the present invention;

[0027] Figure 10 is a schematic diagram of a delivery address identification device according to an embodiment of the present invention;

[0028] Figure 11 is a schematic diagram of another device for identifying a delivery address according to an embodiment of the present invention;

[0029] Figure 12 is a schematic diagram of another device for identifying a delivery address according to an embodiment of the present invention;

[0030] Figure 13 is a schematic diagram of another device for identifying a delivery address according to an embodiment of the present invention;

[0031] Figure 14 It is a structural block diagram of a computer terminal according to an embodiment of the present invention. Detailed implementation manners

[0032] To enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0033] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention 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 such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0034] First, some nouns or terms that appear during the description of the embodiments of the present application are applicable to the following explanations:

[0035] Real number vector (embed), which can be a real number vector of a fixed length and can be represented by 0-1 coding. By representing Chinese characters with real number vectors, it is possible to compare the vectors with each other;

[0036] Cosine distance (cosine distance), which can be used to evaluate the distance relationship between real number vectors through the cosine distance. The larger the value, the closer the vectors are;

[0037] Statistical language model (n-gram), n-gram model encoding, used to split a sentence or word into a combination of smaller-grained n-grams;

[0038] Fund loss refers to the direct or indirect financial losses suffered by the company or the company's customers due to product design defects, abnormal product implementation, employee operation errors, etc.;

[0039] Countermeasure address, which is a randomly selected address, making it difficult to determine the true address information;

[0040] False address, which is a non-existent and false address.

[0041] Embodiment 1

[0042] According to an embodiment of the present invention, an embodiment of a method for identifying a delivery address is further provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0043] The method embodiment provided in the first embodiment of the present application can be executed on a mobile terminal, a computer terminal, or a similar computing device. Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing a method for identifying a delivery address is shown. As Figure 1 shown, the computer terminal 10 (or mobile device 10) may include one or more processors (the processors may include, but are not limited to, processing devices such as a microprocessor MCU or a programmable logic device FPGA, shown as 102a, 102b,..., 102n in the figure), a memory 104 for storing data, and a transmission module 106 for communication functions. In addition, it may further include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply, and / or a camera. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only illustrative and does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 may further include more or fewer components than Figure 1 shown, or have a different configuration from Figure 1 shown.

[0044] It should be noted that the above one or more processors 102 and / or other circuits for identifying the delivery address can generally be referred to as "circuits for identifying the delivery address" in this article. The circuits for identifying the delivery address can be embodied in whole or in part as software, hardware, firmware, or any arbitrary combination. In addition, the circuits for identifying the delivery address can be a single independent processing module, or be incorporated in whole or in part into any one of the other elements in the computer terminal 10 (or mobile device). As involved in the embodiments of the present application, the circuits for identifying the delivery address act as a processor control (such as the selection of a variable resistor terminal path connected to an interface).

[0045] The memory 104 can be used to store software programs and modules of application software, such as the program instruction / data storage device corresponding to the recognition method of the receiving address in the embodiments of the present invention. The processor 102 executes various functional applications and the recognition of the receiving address by running the software programs and modules stored in the memory 104, that is, implements the recognition method of the receiving address of the above application program. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely disposed relative to the processor 102, and these remote memories can be connected to the computer terminal 10 through a network. Examples of the above network include, but are not limited to, the Internet, enterprise intranets, local area networks, mobile communication networks, and combinations thereof.

[0046] The transmission device 106 is used to receive or send data via a network. Specific examples of the above network may include the wireless network provided by the communication provider of the computer terminal 10. In one instance, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station and thus can communicate with the Internet. In one instance, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0047] The display can be, for example, a touch-screen liquid crystal display (LCD), and the liquid crystal display enables a user to interact with the user interface of the computer terminal 10 (or mobile device).

[0048] It should be noted here that, in some alternative embodiments, the above Figure 1 shown computer device (or mobile device) may include hardware elements (including circuits), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware elements and software elements. It should be pointed out that Figure 1 is only an example of a specific specific instance and is intended to illustrate the types of components that may exist in the above computer device (or mobile device).

[0049] In Figure 1 the shown operating environment, the present application provides a recognition method of a receiving address as Figure 2 shown. It should be noted that the recognition method of the receiving address in this embodiment can be executed by Figure 1 the mobile terminal shown in the embodiment.

[0050] Figure 2 is a flowchart of the recognition method of the receiving address according to Embodiment 1 of the present invention, asFigure 2 As shown, the method may include the following steps:

[0051] Step S202: Obtain the semantic information of the order information to be recognized, where the order information at least includes: the delivery address recorded in the order for characterization.

[0052] In the technical solution provided in step S202 of the present invention, the order information to be recognized is obtained, and the semantic information is determined from the order information, where the semantic information can characterize the delivery address recorded in the order information.

[0053] Optionally, the order information to be recognized is obtained. For example, the order record of the user can be obtained from the user terminal; based on the semantic information of the order record, the delivery address is determined.

[0054] Step S204: Determine the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state.

[0055] In the technical solution provided in step S204 of the present invention, the semantic information can be judged to determine the abnormal probability of the delivery address in the order information to be recognized, where the abnormal probability can be the false probability (p(fake)), which can be used to represent the possibility that the delivery address is in an abnormal state.

[0056] Optionally, the semantic information can be sent to a language model, and the language model judges the semantic information to determine the abnormal probability of the delivery address. It should be noted that no specific limit is imposed on the model for processing the semantic information here.

[0057] Optionally, the language model can determine the abnormal probability at the character level, so as to avoid the problem of the language model failing due to special reasons. For example, the problem of the language model failing caused by the user's countermeasure and bypass.

[0058] Step S206: In response to the abnormal probability being greater than the probability threshold, based on the target delivery address, identify whether the delivery address is in an abnormal state, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or, the status of the target delivery address is in a normal state.

[0059] In the technical solution provided in step S206 of the present invention above, it is determined whether the abnormal probability is greater than the probability threshold. In response to the abnormal probability being greater than the probability threshold, the receiving address is identified based on the target receiving address, and it is determined whether the receiving address is in an abnormal state. Among them, the probability threshold can be a value set according to the actual situation, and the target receiving address can be the receiving address in the normal state, for example, the address in the normal state input recently; it can also be the address information associated with the order placement time of the order corresponding to the receiving address.

[0060] Optionally, the target receiving address can be the address information in the normal state. The receiving address determined based on the semantic information is compared with the target receiving address to determine the abnormal probability of the receiving address. In response to the abnormal probability being greater than the probability threshold set in advance according to the actual requirements, it is determined that the receiving address is in an abnormal state, which can be the address information in the abnormal state, for example, the anti-address information.

[0061] Optionally, the target receiving address can be the address information associated with the order placement time of the order corresponding to the receiving address. The receiving address determined based on the semantic information is compared with the target receiving address to determine the abnormal probability of the receiving address. In response to the abnormal probability being greater than the probability threshold set in advance according to the actual requirements, it is determined that the receiving address is in an abnormal state, which can be a randomly selected address, untrue address information, for example, false address information.

[0062] It should be noted that the embodiments of the present invention are not limited to the above two situations. Here, only examples are given and no specific limitations are made.

[0063] Through steps S202 to S206 of the present application above, the semantic information of the order information to be identified is obtained. Among them, the order information at least includes: the receiving address recorded in the order; the abnormal probability of the receiving address is determined based on the semantic information, where the abnormal probability is used to represent the possibility that the receiving address is in an abnormal state; in response to the abnormal probability being greater than the probability threshold, it is identified whether the receiving address is in an abnormal state based on the target receiving address, where the order placement time of the order corresponding to the target receiving address is associated with the order placement time of the order corresponding to the receiving address, and / or the state of the target receiving address is the normal state. That is to say, the present invention determines the abnormal probability of the receiving address according to the semantic information in the receiving address, and then identifies the receiving address through the target receiving address to determine whether the receiving address is in an abnormal state. Since the present invention effectively identifies at the semantic level of the receiving address, the technical effect of improving the accuracy of identifying the receiving address is achieved, and thus the technical problem of low accuracy of identifying the receiving address is solved.

[0064] The above method of this embodiment will be further introduced below.

[0065] As an alternative implementation, in step S206, based on the target delivery address, identifying whether the delivery address is in an abnormal state includes: based on the vector of the delivery address and the target delivery address, identifying whether the delivery address is in an abnormal state.

[0066] In this embodiment, the vector of the delivery address and the target delivery address are determined. Among them, the vector of the delivery address can be used for address aggregation and filtering, and can be a real number vector (embed); the vector of the delivery address is aggregated or matched and filtered with the target delivery address to determine whether the delivery address is in an abnormal state.

[0067] As an alternative implementation, based on the vector of the delivery address and the target delivery address, identifying whether the delivery address is in an abnormal state includes: aggregating the vector of the delivery address and the vector of the first delivery address to obtain an aggregation result, where the target delivery address includes the first delivery address, and the time difference between the order placement time of the order corresponding to the first delivery address and the order placement time of the order corresponding to the delivery address is within the time threshold, and the aggregation result is used to represent the number of first delivery addresses aggregated with the delivery address; in response to the aggregation result being greater than the aggregation threshold, determining that the delivery address is in an abnormal state.

[0068] In this embodiment, the vector of the delivery address is aggregated with the vector of the first delivery address to obtain an aggregation result, determining whether the aggregation result is greater than the aggregation threshold, and in response to the aggregation result being greater than the aggregation threshold, determining that the delivery address is in an abnormal state. Among them, the first delivery address can be the delivery address corresponding to the order whose order placement time is associated with the order placement time of the order corresponding to the delivery address. The time difference between the order placement time of the order corresponding to the first delivery address and the order placement time of the order corresponding to the delivery address is within the time threshold, and it can be a recently input address or an address aggregated with adversarial traces; the aggregation result is used to represent the number of first delivery addresses aggregated with the delivery address.

[0069] Optionally, after generating the vector of the delivery address, the cosine distance of the vector of the delivery address can be used to aggregate with the vector of the first delivery address (for example, different input addresses of different users who placed orders simultaneously in the recent few hours) to obtain an aggregation result. If the aggregation quantity is greater than the aggregation threshold, it can be determined that the status of this address is abnormal, there is a situation of bypassing risk control through adversarial means, and it is determined as an adversarial address, where the aggregation threshold can be a threshold set according to the actual situation.

[0070] As an alternative implementation, based on the vector of the delivery address and the target delivery address, identify whether the delivery address is in an abnormal state, including: matching the vector of the delivery address with the vector of the second delivery address to obtain a matching result, where the target delivery address includes the second delivery address, the second delivery address is in a normal state, and the matching result is used to represent the similarity between the second delivery address and the delivery address; in response to the matching result being greater than the matching threshold, determine that the delivery address is in a normal state.

[0071] In this embodiment, determine the vector of the delivery address, match the vector of the delivery address with the vector of the second delivery address to obtain a matching result, determine whether the matching result is greater than the matching threshold, and in response to the matching result being greater than the matching threshold, determine that the delivery address is in a normal state, where the normal state may refer to a state where normal delivery can be made; the second delivery address may be an address in the target delivery address that is in a normal state and may be a white address.

[0072] Optionally, the vector of the delivery address can be matched and filtered with the vector of the second delivery address (for example, the white address), that is, matched with all addresses where normal delivery can be made to obtain the similarity between the vector of the delivery address and the vector of the second delivery address. In response to the similarity being greater than the matching threshold, it can be determined that the delivery address is in a normal state and can be normally delivered, where the matching threshold can be a threshold set according to the actual situation.

[0073] As an alternative implementation, generate the vector of the delivery address based on the vector generation model, where the vector generation model is trained based on the words in the first delivery address sample, and the words are represented by the multi-model encoding.

[0074] In this embodiment, the model can be trained based on the words in the first delivery address sample to obtain the vector generation model, where the words in the first delivery address sample can be represented by the multi-model encoding (ngram); the first delivery address sample can be a large amount of address data (for example, a large amount of adversarial data samples).

[0075] Optionally, the context words can be predicted through the words in the first delivery address sample to complete the pre-training of the vector generation model. In the embodiments of the present invention, the representation of the words is not the mapping representation of the words themselves, but is represented by combining the words encoded by the multi-model, thus effectively avoiding possible adversarial problems.

[0076] In the embodiments of the present invention, in order to aggregate the vector of the delivery address with the vector of the first delivery address and match the vector of the delivery address with the vector of the second delivery address, during the training of the model, the cosine distances of addresses pointing to the same location with mutation adversaries or typo changes are made as close as possible, and the cosine distances of addresses not pointing to the same location are made as far apart as possible.

[0077] As an optional implementation manner, determine the context words of the words in the first delivery address sample; train a vector generation model based on the context words.

[0078] In this embodiment, determine the context words of the words in the first delivery address sample, and train a vector generation model based on the context words to effectively model the similarity at the semantic level of the address by training the vectors of character-level addresses.

[0079] The embodiments of the present invention, through a character-level model, compared with a classification model, effectively avoid the problem of accuracy decline caused by insufficient samples and infinite spatial changes.

[0080] As an optional implementation manner, in step S204, determining the anomaly probability of the delivery address based on semantic information includes: obtaining multiple characters of the semantic information; determining the anomaly probability based on the multiple characters.

[0081] In this embodiment, obtain multiple characters of the semantic information, and determine the anomaly probability based on the multiple characters by judging the multiple characters.

[0082] As an optional implementation manner, determining the anomaly probability based on the multiple characters includes: combining the multiple characters to obtain a combination result; determining the anomaly probability based on the probability that the delivery address represented by the combination result is a complete delivery address.

[0083] In this embodiment, obtain multiple characters of the semantic information, combine the multiple characters to obtain a combination result, and determine the anomaly probability based on the probability that the delivery address represented by the combination result is a complete delivery address, where the probability of the complete delivery address may be the probability that the multiple characters can be concatenated to form a complete address.

[0084] Optionally, obtain multiple characters of the semantic information, combine the multiple characters, and determine the anomaly probability by judging the probability value that these characters can be concatenated to form a complete address through the characters.

[0085] As an optional implementation manner, combining the multiple characters to obtain a combination result includes: combining the multiple characters based on a language model to obtain a combination result, where the language model is trained based on the characters in the second delivery address sample.

[0086] In this embodiment, the language model is trained based on the characters in the second receiving address sample to obtain the trained language model. Multiple characters are combined based on the language model to obtain a combination result, where the second receiving address sample can be the real address that has been delivered.

[0087] Optionally, the real address that has been delivered can be used as pre-training data to pre-train the model to obtain a complete language model. In the embodiment of the present invention, the character-level language model avoids the problem of reduced accuracy caused by insufficient samples and infinite spatial variations.

[0088] The embodiment of the present invention also provides another method for identifying a receiving address.

[0089] Figure 3 It is a flowchart of another method for identifying a receiving address according to an embodiment of the present invention. As Figure 3 shown, the method may include the following steps.

[0090] Step S302, obtain the order to be processed from the e-commerce platform.

[0091] Step S304, obtain the semantic information of the order information to be recognized in the order, where the order information at least includes: the receiving address recorded in the order.

[0092] Step S306, determine the abnormal probability of the receiving address based on the semantic information, where the abnormal probability is used to represent the possibility that the receiving address is in an abnormal state.

[0093] Step S308, in response to the abnormal probability being greater than the probability threshold, identify whether the receiving address is in an abnormal state based on the target receiving address, where the order placement time of the order corresponding to the target receiving address is associated with the order placement time of the order corresponding to the receiving address, and / or the status of the target receiving address is a normal state.

[0094] Step S310, in response to the receiving address being in an abnormal state, output a prompt message to the e-commerce platform, where the prompt message is used to indicate that the order placement fails.

[0095] In the technical solution provided in step S310 of the present invention, determine whether the abnormal probability of receiving is greater than the probability threshold. In response to the abnormal probability being greater than the probability threshold, then identify whether the receiving address is in an abnormal state based on the target receiving address. In response to the receiving address being in an abnormal state, output a prompt message to the e-commerce platform, where the prompt message can be used to indicate that the order placement fails and can be a component displayed on the e-commerce platform with the message "Order failed" displayed on it.

[0096] The embodiment of the present invention also provides another method for identifying a receiving address.

[0097] Figure 4 It is a flowchart of another method for identifying a delivery address according to an embodiment of the present invention. As Figure 4 shown, the method may include the following steps.

[0098] Step S402, in response to an input instruction on the operation interface, display order information to be identified on the operation interface, where the order information at least includes: information used to represent the delivery address recorded in the order.

[0099] In the technical solution provided in step S402 of the present invention above, the input operation instruction can be triggered by the user and is used to display the order information to be identified on the operation interface. Thus, this embodiment responds to the input operation instruction on the interaction interface and displays the order information to be identified, where the order information at least includes: information used to represent the delivery address recorded in the order.

[0100] Step S404, in response to an identification instruction on the operation interface, display the identification result of the delivery address on the operation interface, where the identification result is used to indicate whether the delivery address is in an abnormal state, and when the abnormal probability of the delivery address is greater than the probability threshold, it is determined based on the target delivery address. The abnormal probability is determined based on the semantic information of the order information and is used to represent the possibility that the delivery address is in an abnormal state. The order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state.

[0101] In the technical solution provided in step S404 of the present invention above, in response to the identification instruction on the operation interface of the interaction interface, the delivery address is identified, the possibility that the delivery address is in an abnormal state is determined based on the semantic information of the order information and the target delivery address to obtain the identification result, and the identification result is displayed on the interaction interface, where the identification result is used to indicate whether the delivery address is in an abnormal state, and it can be determined that the delivery address is abnormal when the abnormal probability of the delivery address is greater than the probability threshold; the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state.

[0102] An embodiment of the present invention also provides another method for identifying a delivery address.

[0103] Figure 5 It is a flowchart of another method for identifying a delivery address according to an embodiment of the present invention. As Figure 5 shown, the method may include the following steps.

[0104] Step S502, obtain the semantic information of the order information to be recognized by calling the first interface. The first interface includes a first parameter, and the parameter value of the first parameter is the semantic information. The order information at least includes: the receiving address recorded in the order for representation.

[0105] In the technical solution provided in step S502 of the present invention above, the first interface can be an interface for data interaction between the server and the client. The client can transmit the semantic information of at least one order information to be recognized into the first interface as a first parameter of the first interface, so as to achieve the purpose of uploading the semantic information of the order information to be recognized to the server.

[0106] Step S504, determine the abnormal probability of the receiving address based on the semantic information, where the abnormal probability is used to represent the possibility that the receiving address is in an abnormal state.

[0107] Step S506, in response to the abnormal probability being greater than the probability threshold, determine the recognition result based on the target receiving address. The order placement time of the order corresponding to the target receiving address is associated with the order placement time of the order corresponding to the receiving address, and / or the status of the target receiving address is a normal state. The recognition result is used to represent whether the receiving address is in an abnormal state.

[0108] Step S508, output the recognition result by calling the second interface. The second interface includes a second parameter, and the parameter value of the second parameter is the recognition result.

[0109] In the technical solution provided in step S508 of the present invention above, the second interface can be an interface for data interaction between the server and the client. The server can transmit the recognition result into the second interface as a parameter of the second interface to achieve the purpose of sending the recognition result to the client. Optionally, the platform outputs the recognition result by calling the second interface, where the second interface is used to deploy the recognition result through the Internet and access the system to be measured, so as to output the recognition result.

[0110] In the embodiment of the present invention, by determining the abnormal probability of the receiving address according to the semantic information in the receiving address, and then identifying the receiving address through the target receiving address to determine whether the receiving address is in an abnormal state. Since the present invention effectively performs recognition at the semantic level of the receiving address, the technical effect of improving the accuracy of recognizing the receiving address is achieved, and thus the technical problem of low accuracy in recognizing the receiving address is solved.

[0111] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0112] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to enable a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present invention.

[0113] Embodiment 2

[0114] The following further introduces the preferred implementation manner of the above method of this embodiment, and specifically illustrates it with a method for determining false and adversarial addresses based on semantics and aggregation.

[0115] Currently, in e-commerce, there are bad behaviors of writing false or adversarial receiving addresses that cannot be delivered by logistics. For example, in red envelope fraud and scalper ticket purchasing, the above behaviors cause serious harm to the seller. For example, if the seller fails to notice the false address and chooses to ship directly, it will cause the goods to be unable to be delivered locally and can only be destroyed on the spot, resulting in losses to the seller. At the same time, it will also bypass the risk control strategy of the e-commerce platform, causing losses to the funds of the e-commerce platform. For example, by filling in an adversarial receiving address to bypass the risk control strategy to receive red envelopes, it causes losses to the marketing funds of the platform. Therefore, since the receiving address is the last link in placing an order, it plays an important role in effectively identifying false addresses and adversarial addresses and cooperating with risk control processing.

[0116] On some foreign e-commerce platforms, only some simple rules are set for identification, and at the same time, a relatively large disposal granularity is relied on to identify address information. However, this method does not fundamentally solve the problem of low accuracy in identifying false addresses and adversarial addresses through algorithms; on some domestic e-commerce platforms, generally, users are required to select streets and add map positioning, etc. However, this method cannot be used in the case of incomplete address data, and there is a problem with the scope of use.

[0117] In related technologies, for the identification of false addresses, a classification model is usually used for differentiation. However, due to the infinite spatial variation of false addresses, it is easy to find error points in the classification model and carry out adversarial attacks; moreover, it is very difficult for the classification model to collect all samples completely. Therefore, there is still a technical problem of low accuracy in identifying the receiving address. For the identification of adversarial addresses, the address normalization method or the minhash method is usually used. However, for address normalization, the construction and mining of normalized addresses are required, which is a large workload and cannot respond to the continuously changing new addresses and new forms of changes. Although the minhash method is relatively simple to use, it only considers character-level similarity and ignores semantic-level similarity, resulting in many cases where although it can be recognized as similar adversarial addresses, the minhash method is bypassed, and there is still a technical problem of low accuracy in identifying the receiving address.

[0118] The embodiment of the present invention proposes to effectively identify false addresses at the semantic level of a single address by establishing an effective matching between a semantic discrimination model and an address library. At the same time, through address representation based on adversarial training, addresses with adversarial traces are effectively aggregated, thereby improving the identification effect of adversarial addresses.

[0119] The above method of this embodiment will be further introduced below.

[0120] Figure 6 It is a flowchart of a method for determining false adversarial addresses according to an embodiment of the present invention. As Figure 6 shown, this embodiment includes the following steps.

[0121] Step S601, input an address.

[0122] Input the address to be identified and identify the address.

[0123] Step S602, determine the falsity of the semantics.

[0124] Since directly making a preliminary determination through semantic literals is likely to miss some adversarial addresses that look very similar to normal addresses and mis-hit some relatively rare white addresses, therefore, first judge the falsity of the input address according to semantics. For example, to identify false addresses of the garbled type, a perception score is implemented for addresses that rarely appear abnormally.

[0125] In this embodiment, the determination of the falsity of the semantics can be implemented through a language model. Figure 7 It is a schematic diagram of a language model prediction according to an embodiment of the present invention. As Figure 7 shown, the language model judges the probability value that these characters can be concatenated to form a complete address through the input characters, and obtains a prediction result.

[0126] Optionally, during the training phase, all delivered real addresses are used as pre-training data to pre-train the language model to obtain a complete language model.

[0127] Optionally, after the model is trained, the model can be used to predict the input address, the entire address can be sent to the model for calculation, and the probability of the input address being false can be determined by a formula.

[0128] Optionally, the difference between the embodiment of the present invention and the ordinary language model is that the language model of the embodiment of the present invention makes judgments at the character level to effectively avoid the problem of language model failure caused by users' attempts to combat bypass, and at the same time smoothes the length to ensure addresses of different lengths. Therefore, the language model of the embodiment of the present invention can judge addresses above the scores given by the model as false addresses, thereby more effectively identifying false addresses of the garbled type.

[0129] Step S603, generating a real number vector (embed) of addresses.

[0130] In this embodiment, a real number vector of addresses is generated in order to aggregate recent input addresses and perform matching and filtering with white addresses in subsequent processes.

[0131] Optionally, by performing expected training on the language model, the cosine distances of addresses pointing to the same place with variation or typographical errors are made as close as possible, and the cosine distances of addresses not pointing to the same place are made as far as possible.

[0132] Optionally, by utilizing a large amount of address data, the current word is used to predict the context word for pre-training to complete the training of the model, wherein the word representation is not represented by the mapping of the word itself, but is formed by combining the model encoding (ngram) to effectively avoid possible adversarial problems.

[0133] Optionally, Figure 8 is a schematic diagram of a language model pre-training according to an embodiment of the present invention, such as Figure 8 As shown, the model code 11, model code 12, model code 31 and model code 32 of the current word are input, the language model processes the input data, generates a real number vector of the address, aggregates the real number vector of the generated address, and obtains the final predicted address.

[0134] Optionally, Figure 9 is a schematic diagram of shortening the distance between word vectors according to an embodiment of the present invention. Figure 9As shown, through contrastive learning, the similarity between vectors is determined. Adversarial knowledge is used to generate mutated address samples as positive examples, reducing the vector distance (embedding distance) between adversarial words and normal words. At the same time, data from non - same locations in the map data is taken as negative examples to increase the vector distance between adversarial words and normal words.

[0135] Step S604: Aggregate with the recently input address and perform matching and filtering on the white - listed address.

[0136] In this embodiment, after generating real - valued vectors (embeddings), the cosine distance of the real - valued vectors can be used to aggregate with different input addresses of different users who placed orders simultaneously in the recent few hours. If it is found that the aggregation quantity is greater than a certain threshold, it can be determined that there is a situation of bypassing risk control by adversarial means for this address, and it is determined as an adversarial address. It can be matched and filtered with the white - listed addresses, and matched with all addresses that can be normally shipped. If the similarity is found to be greater than a certain threshold, it can be determined that this address can be normally shipped.

[0137] Step S605: Determine the false adversarial address.

[0138] In this embodiment, through semantic falsity determination, input address aggregation, and white - listed address matching and filtering, the determination result of the false adversarial address is determined, and a reliable risk control conclusion is obtained.

[0139] Optionally, for false addresses, in the embodiments of the present invention, through a character - level language model, compared with the classification model, Table 1 shows the accuracy comparison results at the same recall rate according to the embodiments of the present invention. As shown in Table 1, the embodiments of the present invention greatly improve the accuracy of non - false address recognition and effectively avoid the decrease in accuracy caused by insufficient samples and infinite spatial changes.

[0140] Table 1 shows the accuracy comparison results at the same recall rate according to the embodiments of the present invention

[0141] False address comparison Accuracy Classification method 81.45％ This method 94.69％

[0142] Optionally, for adversarial addresses, in the embodiments of the present invention, by training character - level embeddings and using the similarity at the semantic level of the modeled address, rather than simply comparing literal similarities, Table 2 shows the comparison results of the similarities according to the embodiments of the present invention. As shown in Table 2, through this determination result of the false adversarial address, compared with not using this result, at the same accuracy, the recall rate can be increased by about 20%.

[0143] Table 2 shows the comparison results of the similarities according to the embodiments of the present invention

[0144] False address comparison AUC MinHash method 0.67 This method 0.76

[0145] In an embodiment of the present invention, by determining the abnormal probability of the delivery address according to the semantic information in the delivery address, and then identifying the delivery address through the target delivery address to determine whether the delivery address is in an abnormal state. Since the present invention effectively identifies at the semantic level of the delivery address, the technical effect of improving the accuracy of identifying the delivery address is achieved, and thus the technical problem of low accuracy in identifying the delivery address is solved.

[0146] Embodiment 3

[0147] According to an embodiment of the present invention, there is also provided an identification device for a delivery address for implementing the Figure 2 identification method of the delivery address shown above.

[0148] Figure 10 FIG. is a schematic diagram of an identification device for a delivery address according to an embodiment of the present invention. As Figure 10 shown, the identification device 1000 for the delivery address may include: a first acquisition unit 1002, a first determination unit 1004, and a first identification unit 1006.

[0149] The first acquisition unit 1002 is configured to acquire the semantic information of the order information to be identified, where the order information at least includes: the delivery address recorded in the order.

[0150] The first determination unit 1004 is configured to determine the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state.

[0151] The first identification unit 1006 is configured to, in response to the abnormal probability being greater than the probability threshold, identify whether the delivery address is in an abnormal state based on the target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state.

[0152] It should be noted here that the above-mentioned first acquisition unit 1002, first determination unit 1004, and first identification unit 1006 correspond to steps S202 to S206 in Embodiment 1. The instances and application scenarios implemented by the three units and the corresponding steps are the same, but are not limited to the content disclosed in the above-mentioned Embodiment 1. It should be noted that the above units, as part of the device, can run in the computer terminal 10 provided in Embodiment 1.

[0153] According to an embodiment of the present invention, there is also provided an identification device for a delivery address for implementing the Figure 3 identification method of the delivery address shown above.

[0154] Figure 11Schematic diagram of another receiving address recognition device according to an embodiment of the present invention, as Figure 11 shown, the receiving address recognition device 1100 may include: a second acquisition unit 1102, a third acquisition unit 1104, a second determination unit 1106, a second recognition unit 1108, and a first output unit 1110.

[0155] The second acquisition unit 1102 is configured to: acquire an order to be processed from an e-commerce platform.

[0156] The third acquisition unit 1104 is configured to acquire semantic information of order information to be recognized in the order, where the order information at least includes: information for characterizing the receiving address recorded in the order.

[0157] The second determination unit 1106 is configured to determine an abnormal probability of the receiving address based on the semantic information, where the abnormal probability is used to represent the possibility that the receiving address is in an abnormal state.

[0158] The second recognition unit 1108 is configured to, in response to the abnormal probability being greater than a probability threshold, recognize whether the receiving address is in an abnormal state based on a target receiving address, where the order placement time of the order corresponding to the target receiving address is associated with the order placement time of the order corresponding to the receiving address, and / or, the state of the target receiving address is a normal state.

[0159] The first output unit 1110 is configured to, in response to the receiving address being in an abnormal state, output a prompt message to the e-commerce platform, where the prompt message is used to indicate that the order placement fails.

[0160] It should be noted here that the above-mentioned second acquisition unit 1102, third acquisition unit 1104, second determination unit 1106, second recognition unit 1108, and first output unit 1110 correspond to steps S302 to S310 in Embodiment 1. The functions of the five units are the same as those of the corresponding steps in terms of implementation examples and application scenarios, but are not limited to the content disclosed in the above-mentioned Embodiment 1. It should be noted that the above units, as part of the device, can run in the computer terminal 10 provided in Embodiment 1.

[0161] According to an embodiment of the present invention, there is also provided a receiving address recognition device for implementing the above Figure 4 shown receiving address recognition method.

[0162] Figure 12 Schematic diagram of another receiving address recognition device according to an embodiment of the present invention. As Figure 12 shown, the receiving address recognition device 1200 may include: a first display unit 1202 and a second display unit 1204.

[0163] The first display unit 1202 is configured to respond to an input instruction acting on the operation interface and display the order information to be recognized on the operation interface, where the order information at least includes: the receiving address recorded in the order for characterization.

[0164] The second display unit 1204 is configured to respond to a recognition instruction acting on the operation interface and display the recognition result of the receiving address on the operation interface, where the recognition result is used to indicate whether the receiving address is in an abnormal state, and when the abnormal probability of the receiving address is greater than the probability threshold, it is determined based on the target receiving address. The abnormal probability is determined based on the semantic information of the order information and is used to indicate the possibility that the receiving address is in an abnormal state. The order placement time of the order corresponding to the target receiving address is associated with the order placement time of the order corresponding to the receiving address, and / or the status of the target receiving address is a normal state

[0165] It should be noted here that the above first display unit 1202 and second display unit 1204 correspond to steps S402 to S404 in Embodiment 1. The two units have the same implemented examples and application scenarios as the corresponding steps, but are not limited to the content disclosed in the above Embodiment 1. It should be noted that the above units, as part of the device, can run in the computer terminal 10 provided in Embodiment 1.

[0166] According to an embodiment of the present invention, there is also provided a receiving address recognition device for implementing the above Figure 5 shown receiving address recognition method.

[0167] Figure 13 is a schematic diagram of another receiving address recognition device according to an embodiment of the present invention. As Figure 13 shown, the receiving address recognition device 1300 may include: a fourth acquisition unit 1302, a third determination unit 1304, a fourth determination unit 1306, and a second output unit 1308.

[0168] The fourth acquisition unit 1302 is configured to obtain the semantic information of the order information to be recognized by calling the first interface, where the first interface includes a first parameter, and the parameter value of the first parameter is the semantic information. The order information at least includes: the receiving address recorded in the order for characterization.

[0169] The third determination unit 1304 is configured to determine the abnormal probability of the receiving address based on the semantic information, where the abnormal probability is used to indicate the possibility that the receiving address is in an abnormal state.

[0170] A fourth determination unit 1306, configured to determine an identification result based on a target delivery address in response to an exception probability being greater than a probability threshold, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal status, and the identification result is used to indicate whether the delivery address is in an abnormal state.

[0171] A second output unit 1308, configured to output the identification result by invoking a second interface, where the second interface includes a second parameter, and the parameter value of the second parameter is the identification result.

[0172] It should be noted here that the above-mentioned fourth acquisition unit 1302, third determination unit 1304, fourth determination unit 1306, and second output unit 1308 correspond to steps S502 to S508 in Embodiment 1. The instances and application scenarios implemented by the four units and the corresponding steps are the same, but are not limited to the content disclosed in the above-mentioned Embodiment 1. It should be noted that the above units, as part of the device, can run in the computer terminal 10 provided in Embodiment 1.

[0173] In the delivery address identification device of this embodiment, by determining the exception probability of the delivery address according to the semantic information in the delivery address, and then identifying the delivery address through the target delivery address to determine whether the delivery address is in an abnormal state. Since the present invention effectively performs identification at the semantic level of the delivery address, the technical effect of improving the accuracy of identifying the delivery address is achieved, and thus the technical problem of low accuracy in identifying the delivery address is solved.

[0174] Embodiment 4

[0175] An embodiment of the present invention may provide a delivery address identification system, which may include a computer terminal, and the computer terminal may be any one of the computer terminal devices in a computer terminal group. Optionally, in this embodiment, the above-mentioned computer terminal may also be replaced with a terminal device such as a mobile terminal.

[0176] Optionally, in this embodiment, the above-mentioned computer terminal may be located in at least one of multiple network devices in a computer network.

[0177] In this embodiment, the above computer terminal may execute the program code of the following steps in the method for identifying a delivery address: obtain the semantic information of the order information to be identified, where the order information at least includes: the delivery address recorded in the order; determine the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; in response to the abnormal probability being greater than the probability threshold, identify whether the delivery address is in an abnormal state based on the target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state.

[0178] Optionally, Figure 14 is a structural block diagram of a computer terminal according to an embodiment of the present invention. As Figure 14 shown, the computer terminal A may include: one or more (only one is shown in the figure) processors 1402, a memory 1404, and a transmission device 1406.

[0179] Among them, the memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the method and device for identifying a delivery address in the embodiment of the present invention. The processor executes various functional applications and the identification of the delivery address by running the software programs and modules stored in the memory, that is, implements the above method for identifying a delivery address. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory may further include a memory remotely set relative to the processor, and these remote memories may be connected to the computer terminal A through a network. Examples of the above network include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0180] The processor may call the information and application programs stored in the memory through the transmission device to execute the following steps: obtain the semantic information of the order information to be identified, where the order information at least includes: the delivery address recorded in the order; determine the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; in response to the abnormal probability being greater than the probability threshold, identify whether the delivery address is in an abnormal state based on the target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state.

[0181] Optionally, the above processor may also execute the program code of the following steps: identify whether the delivery address is in an abnormal state based on the vector of the delivery address and the target delivery address.

[0182] Optionally, the above-mentioned processor may also execute the program code of the following steps: Aggregate the vector of the delivery address and the vector of the first delivery address to obtain an aggregation result, where the target delivery address includes the first delivery address, and the time difference between the order placement time of the order corresponding to the first delivery address and the order placement time of the order corresponding to the delivery address is within the time threshold, and the aggregation result is used to represent the number of first delivery addresses aggregated with the delivery address; In response to the aggregation result being greater than the aggregation threshold, determine that the delivery address is in an abnormal state.

[0183] Optionally, the above-mentioned processor may also execute the program code of the following steps: Match the vector of the delivery address and the vector of the second delivery address to obtain a matching result, where the target delivery address includes the second delivery address, and the second delivery address is in a normal state, and the matching result is used to represent the similarity between the second delivery address and the delivery address; In response to the matching result being greater than the matching threshold, determine that the delivery address is in a normal state.

[0184] Optionally, the above-mentioned processor may also execute the program code of the following steps: Generate a vector of the delivery address based on a vector generation model, where the vector generation model is trained based on the words in the first delivery address sample, and the words are represented by a multi-model encoding.

[0185] Optionally, the above-mentioned processor may also execute the program code of the following steps: Determine the context words of the words in the first delivery address sample; Train a vector generation model based on the context words.

[0186] Optionally, the above-mentioned processor may also execute the program code of the following steps: Determine the abnormal probability of the delivery address based on semantic information, including: Obtain multiple characters of the semantic information; Determine the abnormal probability based on the multiple characters.

[0187] Optionally, the above-mentioned processor may also execute the program code of the following steps: Combine the multiple characters to obtain a combination result; Determine the abnormal probability based on the probability that the delivery address represented by the combination result is a complete delivery address.

[0188] Optionally, the above-mentioned processor may also execute the program code of the following steps: Combine the multiple characters based on a language model to obtain a combination result, where the language model is trained based on the characters in the second delivery address sample.

[0189] As an optional example, the processor may call the information and application programs stored in the memory through a transmission device to perform the following steps: obtain an order to be processed from an e-commerce platform; obtain the semantic information of the order information to be recognized in the order, where the order information at least includes: information used to represent the delivery address recorded in the order; determine the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; in response to the abnormal probability being greater than a probability threshold, determine whether the delivery address is in an abnormal state based on the target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state; in response to the delivery address being in an abnormal state, output a prompt message to the e-commerce platform, where the prompt message is used to indicate that the order placement fails.

[0190] As an optional example, the processor may call the information and application programs stored in the memory through a transmission device to perform the following steps: in response to an input instruction acting on an operation interface, display the order information to be recognized on the operation interface, where the order information at least includes: information used to represent the delivery address recorded in the order; in response to a recognition instruction acting on the operation interface, display the recognition result of the delivery address on the operation interface, where the recognition result is used to represent whether the delivery address is in an abnormal state, and when the abnormal probability of the delivery address is greater than the probability threshold, it is determined based on the target delivery address, the abnormal probability is determined based on the semantic information of the order information, and is used to represent the possibility that the delivery address is in an abnormal state, the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state.

[0191] As an optional example, the processor may call the information and application programs stored in the memory through a transmission device to perform the following steps: obtain the semantic information of the order information to be recognized by calling a first interface, where the first interface includes a first parameter, and the parameter value of the first parameter is the semantic information, the order information at least includes: information used to represent the delivery address recorded in the order; determine the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; in response to the abnormal probability being greater than a probability threshold, determine a recognition result based on the target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is a normal state, and the recognition result is used to represent whether the delivery address is in an abnormal state; output the recognition result by calling a second interface, where the second interface includes a second parameter, and the parameter value of the second parameter is the recognition result.

[0192] An embodiment of the present invention provides a method for identifying a delivery address. By determining the abnormal probability of the delivery address according to the semantic information in the delivery address, and then identifying the delivery address through the target delivery address to determine whether the delivery address is in an abnormal state. Since the present invention effectively identifies at the semantic level of the delivery address, the technical effect of improving the accuracy of identifying the delivery address is achieved, and thus the technical problem of low accuracy in identifying the delivery address is solved.

[0193] Those of ordinary skill in the art can understand that Figure 14 The structure shown is only for illustration. Computer terminal A can also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a palm computer, and a mobile Internet device (Mobile Internet Devices, MID), a PAD and other terminal devices. Figure 14 It does not limit the structure of the above computer terminal A. For example, computer terminal A may further include more or fewer components (such as a network interface, a display device, etc.) than those shown Figure 14 in the figure, or have a different configuration from that shown Figure 14 in the figure.

[0194] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the relevant hardware of the terminal device through a program. The program can be stored in a computer-readable storage medium. The storage medium can include: a flash drive, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disc, etc.

[0195] Embodiment 5

[0196] An embodiment of the present invention further provides a computer-readable storage medium. Optionally, in this embodiment, the above computer-readable storage medium can be used to store the program code executed by the method for identifying a delivery address provided in the first embodiment above.

[0197] Optionally, in this embodiment, the above storage medium can be located in any one of the computer terminals in a computer terminal group in a computer network, or in any one of the mobile terminals in a mobile terminal group.

[0198] Optionally, in this embodiment, the storage medium is configured to store program code for performing the following steps: obtaining semantic information of an order information to be recognized, where the order information at least includes: a receiving address recorded in the order; determining an abnormal probability of the receiving address based on the semantic information, where the abnormal probability is used to represent the possibility that the receiving address is in an abnormal state; in response to the abnormal probability being greater than a probability threshold, identifying whether the receiving address is in an abnormal state based on a target receiving address, where the placing time of the order corresponding to the target receiving address is associated with the placing time of the order corresponding to the receiving address, and / or, the state of the target receiving address is a normal state.

[0199] Optionally, the above computer-readable storage medium may also execute program code for performing the following steps: determining a protocol format of a target communication protocol; determining a target control signal associated with the target communication protocol based on the protocol format.

[0200] Optionally, the above computer-readable storage medium may also execute program code for performing the following steps: identifying whether the receiving address is in an abnormal state based on the vector of the receiving address and the target receiving address.

[0201] Optionally, the above computer-readable storage medium may also execute program code for performing the following steps: aggregating the vector of the receiving address and the vector of a first receiving address to obtain an aggregation result, where the target receiving address includes the first receiving address, and the difference between the placing time of the order corresponding to the first receiving address and the placing time of the order corresponding to the receiving address is within a time threshold, and the aggregation result is used to represent the number of first receiving addresses aggregated with the receiving address; in response to the aggregation result being greater than an aggregation threshold, determining that the receiving address is in an abnormal state.

[0202] Optionally, the above computer-readable storage medium may also execute program code for performing the following steps: matching the vector of the receiving address and the vector of a second receiving address to obtain a matching result, where the target receiving address includes the second receiving address, and the second receiving address is in a normal state, and the matching result is used to represent the similarity between the second receiving address and the receiving address; in response to the matching result being greater than a matching threshold, determining that the receiving address is in a normal state.

[0203] Optionally, the above computer-readable storage medium may also execute program code for performing the following steps: generating a vector of the receiving address based on a vector generation model, where the vector generation model is trained based on words in a first receiving address sample, and the words are encoded and represented by a multi-model.

[0204] Optionally, the above computer-readable storage medium may also execute program code for performing the following steps: determining context words of words in the first receiving address sample; training a vector generation model based on the context words.

[0205] Optionally, the above computer-readable storage medium may also execute program code for the following steps: determining an abnormal probability of a delivery address based on semantic information, including: obtaining a plurality of characters of the semantic information; determining the abnormal probability based on the plurality of characters.

[0206] Optionally, the above computer-readable storage medium may also execute program code for the following steps: combining the plurality of characters to obtain a combination result; determining the abnormal probability based on the probability that the delivery address represented by the combination result is a complete delivery address.

[0207] Optionally, the above computer-readable storage medium may also execute program code for the following steps: combining the plurality of characters based on a language model to obtain a combination result, where the language model is trained based on the characters in the second delivery address sample.

[0208] As an optional example, the computer-readable storage medium is configured to store program code for performing the following steps: obtaining an order to be processed from an e-commerce platform; obtaining semantic information of order information to be recognized in the order, where the order information at least includes: information used to represent the delivery address recorded in the order; determining an abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; in response to the abnormal probability being greater than a probability threshold, determining whether the delivery address is in an abnormal state based on a target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or, the status of the target delivery address is a normal state; in response to the delivery address being in an abnormal state, outputting a prompt message to the e-commerce platform, where the prompt message is used to indicate that the order placement fails.

[0209] As an optional example, the computer-readable storage medium is configured to store program code for performing the following steps: in response to an input instruction on an operation interface, displaying the order information to be recognized on the operation interface, where the order information at least includes: information used to represent the delivery address recorded in the order; in response to a recognition instruction on the operation interface, displaying a recognition result of the delivery address on the operation interface, where the recognition result is used to represent whether the delivery address is in an abnormal state, and when the abnormal probability of the delivery address is greater than the probability threshold, it is determined based on the target delivery address, the abnormal probability is determined based on the semantic information of the order information, and is used to represent the possibility that the delivery address is in an abnormal state, the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or, the status of the target delivery address is a normal state.

[0210] As an alternative example, a computer-readable storage medium is configured to store program code for performing the following steps: obtaining semantic information of an order to be recognized by invoking a first interface, where the first interface includes a first parameter, and the parameter value of the first parameter is the semantic information, and the order information at least includes: an address for receiving goods recorded in the order; determining an abnormal probability of the address for receiving goods based on the semantic information, where the abnormal probability is used to represent the possibility that the address for receiving goods is in an abnormal state; in response to the abnormal probability being greater than a probability threshold, determining an identification result based on a target address for receiving goods, where the order placement time of the order corresponding to the target address for receiving goods is associated with the order placement time of the order corresponding to the address for receiving goods, and / or, the status of the target address for receiving goods is a normal state, and the identification result is used to represent whether the address for receiving goods is in an abnormal state; outputting the identification result by invoking a second interface, where the second interface includes a second parameter, and the parameter value of the second parameter is the identification result.

[0211] The serial numbers of the above-described embodiments of the present invention are only for description and do not represent the superiority or inferiority of the embodiments.

[0212] In the above embodiments of the present invention, the descriptions of the various embodiments have their own emphases. For parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0213] In several embodiments provided in the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces, and the indirect coupling or communication connection of units or modules can be in an electrical or other form.

[0214] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0215] In addition, in each embodiment of the present invention, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0216] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), mobile hard disks, magnetic disks, or optical discs that can store program codes.

[0217] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can still be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A method for identifying a delivery address, characterized in that, Including: Obtain the semantic information of the order information to be recognized, where the order information at least includes: the delivery address recorded in the order; Determine the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; In response to the abnormal probability being greater than the probability threshold, identify whether the delivery address is in the abnormal state based on the target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the status of the target delivery address is in a normal state; Among them, identifying whether the delivery address is in the abnormal state based on the target delivery address includes: aggregating the delivery address with the first delivery address in the target delivery address to obtain an aggregation result; based on the aggregation result, identifying whether the delivery address is in the abnormal state, where the difference between the order placement time of the order corresponding to the first delivery address and the order placement time of the order corresponding to the delivery address is within the time threshold; or Matching the delivery address with the second delivery address in the target delivery address to obtain a matching result; based on the matching result, identifying whether the delivery address is in the abnormal state, where the second delivery address is in a normal state.

2. The method according to claim 1, characterized in that Aggregating the delivery address with the first delivery address in the target delivery address to obtain the aggregation result includes: Aggregating the vector of the delivery address and the vector of the first delivery address to obtain the aggregation result, where the aggregation result is used to represent the number of the first delivery addresses aggregated with the delivery address.

3. The method according to claim 2, wherein Identifying whether the delivery address is in the abnormal state based on the aggregation result includes: In response to the aggregation result being greater than the aggregation threshold, determine that the delivery address is in the abnormal state.

4. The method according to claim 1, characterized in that, Matching the delivery address with the second delivery address in the target delivery address to obtain the matching result includes: Matching the vector of the delivery address and the vector of the second delivery address to obtain the matching result, where the matching result is used to represent the similarity between the second delivery address and the delivery address; Identifying whether the delivery address is in the abnormal state based on the matching result includes: in response to the matching result being greater than the matching threshold, determine that the delivery address is in the normal state.

5. The method according to claim 1, wherein The method further includes: Generating the vector of the delivery address based on a vector generation model, where the vector generation model is trained based on the words in the first delivery address sample, and the words are represented by a multi-model encoding.

6. The method according to claim 5, wherein The method further includes: Determine the context words of the words in the first delivery address sample; Train the vector generation model based on the context words.

7. The method according to claim 1, wherein Determining the abnormal probability of the delivery address based on the semantic information includes: Obtain multiple characters of the semantic information; Determine the abnormal probability based on the multiple characters.

8. The method according to claim 7, wherein Determining the abnormal probability based on the multiple characters includes: Combine the multiple characters to obtain a combination result; Determine the abnormal probability based on the probability that the delivery address represented by the combination result is a complete delivery address.

9. The method according to claim 8, wherein Combining the multiple characters to obtain a combination result includes: Combining the multiple characters based on a language model to obtain the combination result, where the language model is trained based on the characters in the second delivery address sample.

10. A method for identifying a delivery address, characterized in that, Including: Obtain an order to be processed from an e-commerce platform; Obtain the semantic information of the order information to be recognized in the order, where the order information at least includes: the delivery address recorded in the order; Determine the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to represent the possibility that the delivery address is in an abnormal state; In response to the abnormal probability being greater than a probability threshold, identify whether the delivery address is in the abnormal state based on a target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or, the status of the target delivery address is a normal state; In response to the delivery address being in the abnormal state, output a prompt message to the e-commerce platform, where the prompt message is used to indicate that the order placement fails; Among them, identifying whether the delivery address is in the abnormal state based on the target delivery address includes: aggregating the delivery address with the first delivery address in the target delivery address to obtain an aggregation result; identifying whether the delivery address is in the abnormal state based on the aggregation result, where the difference between the order placement time of the order corresponding to the first delivery address and the order placement time of the order corresponding to the delivery address is within a time threshold; or, Matching the delivery address with the second delivery address in the target delivery address to obtain a matching result; identifying whether the delivery address is in the abnormal state based on the matching result, where the second delivery address is in a normal state.

11. A method for identifying a delivery address, characterized in that, Including: In response to an input instruction on the operation interface, display the order information to be recognized on the operation interface, where the order information at least includes: the delivery address recorded in the order; In response to an identification instruction acting on the operation interface, display the identification result of the delivery address on the operation interface, where the identification result is used to indicate whether the delivery address is in an abnormal state, and when the abnormal probability of the delivery address is greater than a probability threshold, it is obtained based on an aggregation result or a matching result. The aggregation result is obtained by aggregating the delivery address with a first delivery address among target delivery addresses, and the matching result is obtained by matching the delivery address with a second delivery address among the target delivery addresses. The time difference between the order placement time of the order corresponding to the first delivery address and the order placement time of the order corresponding to the delivery address is within a time threshold, the second delivery address is in a normal state, the abnormal probability is determined based on the semantic information of the order information, and is used to indicate the possibility that the delivery address is in the abnormal state. The order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the state of the target delivery address is in a normal state.

12. A method for identifying a delivery address, characterized in that, including: Obtain the semantic information of the order information to be identified by calling a first interface, where the first interface includes a first parameter, and the parameter value of the first parameter is the semantic information. The order information at least includes: the delivery address recorded in the order; Determine the abnormal probability of the delivery address based on the semantic information, where the abnormal probability is used to indicate the possibility that the delivery address is in an abnormal state; In response to the abnormal probability being greater than the probability threshold, determine the identification result based on the target delivery address, where the order placement time of the order corresponding to the target delivery address is associated with the order placement time of the order corresponding to the delivery address, and / or the state of the target delivery address is in a normal state. The identification result is used to indicate whether the delivery address is in the abnormal state; Output the identification result by calling a second interface, where the second interface includes a second parameter, and the parameter value of the second parameter is the identification result; where determining the identification result based on the target delivery address includes: aggregating the delivery address with a first delivery address among the target delivery addresses to obtain an aggregation result; based on the aggregation result, identify whether the delivery address is in the abnormal state, where the time difference between the order placement time of the order corresponding to the first delivery address and the order placement time of the order corresponding to the delivery address is within a time threshold; or, matching the delivery address with a second delivery address among the target delivery addresses to obtain a matching result; based on the matching result, identify whether the delivery address is in the abnormal state, where the second delivery address is in a normal state.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, where when the program is run by a processor, it controls the device where the computer-readable storage medium is located to execute the method according to any one of claims 1 to 12.

14. A processor, characterized in that, The processor is used to run a program, where when the program runs, it executes the method according to any one of claims 1 to 12.

Images