CN117112628A - Logistics data updating method and system - Google Patents

Abstract

The invention provides a method and a system for updating logistics data, and relates to the technical field of data processing. In the invention, key information mining processing is carried out on the historical logistics data to be processed so as to output historical key information characteristic representation corresponding to the historical logistics data; carrying out key information mining processing on target logistics data to be processed so as to output target key information characteristic representation corresponding to the target logistics data; performing matching degree calculation processing on the historical key information feature representation and the target key information feature representation to output corresponding feature representation matching degree; and based on the characteristic representation matching degree, carrying out data updating management and control on the historical logistics data and the target logistics data so as to at least save the target logistics data. Based on the method, the reliability of the update of the logistics data can be improved.

Description

A method and system for updating logistics data

Technical field

The present invention relates to the field of data processing technology, and specifically to a method and system for updating logistics data.

Background technique

Logistics data has many application scenarios and high application value. Therefore, logistics data will be collected and stored. However, the amount of logistics data is also very large. Therefore, the stored data also needs to be updated, such as Unnecessary logistics data is discarded, and only part of the logistics data is saved. For example, in the existing technology, in the process of updating and controlling logistics data, the logistics data is generally processed based on the formation time. Specifically, the data with an earlier formation time is generally directly discarded and the current time is retained. Logistics data formed in order to realize the updating of old and new logistics data. In this way, it is easy to have the problem of poor reliability in the update and control of logistics data.

Contents of the invention

In view of this, the object of the present invention is to provide a logistics data updating method and system to improve the reliability of logistics data updating.

In order to achieve the above objects, the embodiments of the present invention adopt the following technical solutions:

A method for updating logistics data, including:

Perform key information mining processing on the historical logistics data to be processed to output the historical key information feature representation corresponding to the historical logistics data. The historical logistics data includes each piece of historical logistics information within the historical time period, and one piece of the historical logistics information Corresponds to a logistics cargo;

Perform key information mining processing on the target logistics data to be processed to output the target key information feature representation corresponding to the target logistics data. The target logistics data includes each piece of target logistics information in the current time period, one piece of the target logistics information. Corresponds to a logistics cargo;

Perform matching degree calculation processing on the historical key information feature representation and the target key information feature representation to output the corresponding feature representation matching degree;

Based on the matching degree of the feature representation, data update management is performed on the historical logistics data and the target logistics data, so as to at least save the target logistics data.

In some preferred embodiments, in the above method for updating logistics data, the step of performing key information mining processing on the historical logistics data to be processed to output the feature representation of historical key information corresponding to the historical logistics data includes:

Utilize the optimized key information mining network to perform key information mining processing on the historical logistics data to be processed to output the historical key information feature representation corresponding to the historical logistics data;

The step of performing key information mining processing on the target logistics data to be processed to output target key information feature representation corresponding to the target logistics data includes:

The optimized key information mining network is used to perform key information mining processing on the target logistics data to be processed, so as to output the target key information feature representation corresponding to the target logistics data.

In some preferred embodiments, in the above-mentioned updating method of logistics data, the updating method further includes: performing network optimization processing to form an optimized key information mining network;

The steps of performing network optimization processing to optimize the key information mining network include:

Extract typical logistics data combinations, and analyze multi-dimensional logistics characterization data corresponding to the typical logistics data combinations. The typical logistics data combinations include at least first typical logistics data and second typical logistics data. The second typical logistics data combinations are The data matching relationship between the logistics data and the first typical logistics data conforms to the preset data matching relationship, and the multi-dimensional logistics characterization data includes at least two dimensions of logistics characterization data;

Perform a hiding operation on the logistics characterization data of the first dimension in the multi-dimensional logistics characterization data to output the corresponding hidden multi-dimensional logistics characterization data;

Through the initial key information mining network, the hidden multi-dimensional logistics representation data is subjected to key information mining processing to output typical key information feature representations and hidden data restoration feature representations corresponding to the typical logistics data combinations;

Based on the first dimension of logistics characterization data, the hidden data restoration feature representation and the typical key information feature representation, the initial key information mining network is used to determine the learning cost value to output the initial key information Mining the corresponding restoration learning cost value and the corresponding key information learning cost value of the network;

According to the restored learning cost value and the key information learning cost value, the initial key information mining network is optimized to form an optimized key information mining network.

In some preferred embodiments, in the above method for updating logistics data, based on the logistics characterization data of the first dimension, the hidden data reduction feature representation and the typical key information feature representation, the initial The key information mining network determines the learning cost value to output the restored learning cost value and the corresponding key information learning cost value corresponding to the initial key information mining network, including:

Based on the first dimension of logistics representation data and the hidden data restoration feature representation, the initial key information mining network is performed to determine the first learning cost value, and the restoration learning code corresponding to the initial key information mining network is output. value;

Based on the characteristic representation of typical key information corresponding to the typical logistics data combination, the initial key information mining network is used to determine the second learning cost value, and the key information learning cost value corresponding to the initial key information mining network is output.

In some preferred embodiments, in the above logistics data updating method, the initial key information mining network is used to determine the second learning cost value based on the typical key information feature representation corresponding to the typical logistics data combination. Operation, the step of outputting the key information learning cost value corresponding to the initial key information mining network includes:

From the second typical logistics data included in the typical logistics data combination, relevant typical logistics data and non-related typical logistics data are determined, the relevant typical logistics data and the first typical logistics data have a matching relationship, and the non-related typical logistics data The relevant typical logistics data and the first typical logistics data do not have a matching relationship; and, based on the characteristic representation of the typical key information, the matching degree between the first typical logistics data and the relevant typical logistics data is analyzed, to output the corresponding relevant dimension matching degree, and to analyze the matching degree between the first typical logistics data and the non-relevant typical logistics data to output the corresponding non-relevant dimension matching degree; and, based on the correlation Dimension matching degree and non-relevant dimension matching degree, analyze the key information learning cost value corresponding to the initial key information mining network;

Based on the first dimension of logistics characterization data and the hidden data restoration feature representation, the initial key information mining network is used to determine the first learning cost value, and the corresponding restoration of the initial key information mining network is output. Steps to learn value include:

Reduction processing of characterization data is performed based on the hidden data reduction feature representation to output corresponding reduced logistics characterization data; and, based on the first-dimensional logistics characterization data and the reduced logistics characterization data, the initial key information is The mining network determines the first learning cost value and outputs the restored learning cost value corresponding to the initial key information mining network.

In some preferred embodiments, in the above method for updating logistics data, the hidden multi-dimensional logistics representation data is subjected to key information mining processing through an initial key information mining network to output the typical logistics data. The steps of combining the corresponding typical key information feature representation and the hidden data restoration feature representation include:

The hidden multi-dimensional logistics representation data is integrated into the distribution domain to output the initial feature representation corresponding to each dimension located in the same distribution domain;

Through the initial key information mining network, the initial feature representation is aggregated to output the typical key information feature representation corresponding to each typical logistics data in the typical logistics data combination;

According to the characteristic representation of typical key information, the hidden data in the hidden multi-dimensional logistics representation data is restored, and the restored characteristic representation of the hidden data is output.

In some preferred embodiments, in the above method for updating logistics data, the historical logistics data and the target logistics data are subject to data update control based on the feature representation matching degree, so as to at least update the target logistics data. The steps to save logistics data include:

Extract the pre-configured reference matching degree;

Perform size comparison processing on the reference matching degree and the feature representation matching degree;

When the characteristic representation matching degree is greater than or equal to the reference matching degree, both the historical logistics data and the target logistics data are saved;

When the feature representation matching degree is less than the reference matching degree, the historical logistics data is discarded and the target logistics data is saved.

Embodiments of the present invention also provide a logistics data updating system, including:

The first key information mining module is used to perform key information mining processing on the historical logistics data to be processed to output the historical key information feature representation corresponding to the historical logistics data. The historical logistics data includes each piece of history within the historical time period. Logistics information, one piece of historical logistics information corresponds to one logistics cargo;

The second key information mining module is used to perform key information mining processing on the target logistics data to be processed, so as to output the target key information feature representation corresponding to the target logistics data. The target logistics data includes each target in the current time period. Logistics information, one piece of target logistics information corresponds to one logistics cargo;

A matching degree calculation module, configured to perform matching degree calculation processing on the historical key information feature representation and the target key information feature representation, to output the corresponding feature representation matching degree;

A data update management and control module is configured to perform data update management and control on the historical logistics data and the target logistics data based on the characteristic representation matching degree, so as to at least save the target logistics data.

In some preferred embodiments, in the above-mentioned logistics data updating system, the first key information mining module is specifically used to:

Utilize the optimized key information mining network to perform key information mining processing on the historical logistics data to be processed to output the historical key information feature representation corresponding to the historical logistics data;

The second key information mining module is specifically used for:

The optimized key information mining network is used to perform key information mining processing on the target logistics data to be processed, so as to output the target key information feature representation corresponding to the target logistics data.

In some preferred embodiments, in the above-mentioned logistics data update system, the data update management and control module is specifically used to:

Extract the pre-configured reference matching degree;

Perform size comparison processing on the reference matching degree and the feature representation matching degree;

When the characteristic representation matching degree is greater than or equal to the reference matching degree, both the historical logistics data and the target logistics data are saved;

When the feature representation matching degree is less than the reference matching degree, the historical logistics data is discarded and the target logistics data is saved.

The logistics data updating method and system provided by the embodiments of the present invention can perform key information mining processing on the historical logistics data to be processed to output the historical key information feature representation corresponding to the historical logistics data; perform key information mining on the target logistics data to be processed. The information mining process is to output the target key information feature representation corresponding to the target logistics data; the matching degree calculation process is performed on the historical key information feature representation and the target key information feature representation to output the corresponding features to represent the matching degree; based on the feature representation matching degree, Perform data update control on historical logistics data and target logistics data to at least save the target logistics data. Based on the above content, since the feature representation matching degree of historical logistics data and target logistics data is calculated, data update management and control of historical logistics data and target logistics data can be carried out based on the obtained feature representation matching degree. In this way, compared with Conventional technical solutions for data update management and control based directly on formation time can improve the reliability of logistics data updates, thus improving the shortcomings of existing technologies.

In order to make the above-mentioned objects, features and advantages of the present invention more obvious and understandable, preferred embodiments are given below and described in detail with reference to the accompanying drawings.

Description of drawings

Figure 1 is a structural block diagram of a logistics data update platform provided by an embodiment of the present invention.

FIG. 2 is a schematic flowchart of the steps included in the logistics data updating method provided by the embodiment of the present invention.

Figure 3 is a schematic diagram of each module included in the logistics data update system provided by the embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments These are only some of the embodiments of the present invention, not all of them. The components of the embodiments of the invention generally described and illustrated in the figures herein may be arranged and designed in a variety of different configurations.

Therefore, the following detailed description of the embodiments of the invention provided in the appended drawings is not intended to limit the scope of the claimed invention, but rather to represent selected embodiments of the invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of protection of the present invention.

As shown in Figure 1, an embodiment of the present invention provides a logistics data update platform. Wherein, the logistics data updating platform may include a memory and a processor.

In detail, the memory and the processor are electrically connected directly or indirectly to realize data transmission or interaction. For example, they can be electrically connected to each other through one or more communication buses or signal lines. The memory may store at least one software function module (computer program) that may exist in the form of software or firmware. The processor may be used to execute an executable computer program stored in the memory, thereby implementing the logistics data updating method provided by embodiments of the present invention (as described later).

Alternatively, in some implementations, the memory may be, but is not limited to, random access memory (Random Access Memory, RAM), read only memory (Read Only Memory, ROM), programmable read only memory ( Programmable Read-Only Memory (PROM), Erasable ProgrammableRead-Only Memory (EPROM), Electrically Erasable ProgrammableRead-Only Memory (EEPROM), etc.

Alternatively, in some implementations, the processor may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), a system on a chip (System on Chip). Chip, SoC), etc.; it can also be a digital signal processor (DSP), application specific integrated circuit (ASIC), field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

Alternatively, in some implementations, the logistics data updating platform may be a server with data processing capabilities.

In conjunction with Figure 2, an embodiment of the present invention also provides a method for updating logistics data, which can be applied to the above-mentioned logistics data update platform. Wherein, the method steps defined in the process related to the logistics data updating method can be implemented by the logistics data updating platform.

The specific process shown in Figure 2 will be elaborated below.

Step S110: Perform key information mining processing on the historical logistics data to be processed to output the historical key information feature representation corresponding to the historical logistics data.

In the embodiment of the present invention, the logistics data update platform can perform key information mining processing on the historical logistics data to be processed, so as to output the historical key information feature representation corresponding to the historical logistics data. The historical logistics data includes each piece of historical logistics information within the historical time period, such as historical logistics information 1, historical logistics information 2, historical logistics information 3, historical logistics information 4, historical logistics information 5, historical logistics information 6, historical logistics Information 7, historical logistics information 8, historical logistics information 9, etc., one piece of historical logistics information corresponds to one logistics cargo. For example, the historical logistics information may include cargo description information of the logistics goods, logistics sending place information, logistics receiving place information, corresponding user information, etc.

Step S120: Perform key information mining processing on the target logistics data to be processed to output target key information feature representation corresponding to the target logistics data.

In the embodiment of the present invention, the logistics data update platform can perform key information mining processing on the target logistics data to be processed, so as to output the target key information feature representation corresponding to the target logistics data. The target logistics data includes each piece of target logistics information in the current time period, such as target logistics information 1, target logistics information 2, target logistics information 3, target logistics information 4, target logistics information 5, target logistics information 6, target logistics Information 7, target logistics information 8, target logistics information 9, etc., one piece of target logistics information corresponds to one logistics cargo. For example, the target logistics information may include cargo description information of the logistics goods, logistics sending location information, logistics receiving location information, corresponding user information, etc.

Step S130: Perform matching degree calculation processing on the historical key information feature representation and the target key information feature representation to output a corresponding feature representation matching degree.

In the embodiment of the present invention, the logistics data update platform can perform matching degree calculation processing, such as cosine similarity calculation, on the historical key information feature representation and the target key information feature representation to output the corresponding feature representation. suitability.

Step S140: Based on the feature representation matching degree, perform data update management and control on the historical logistics data and the target logistics data, so as to at least save the target logistics data.

In an embodiment of the present invention, the logistics data update platform can perform data update management and control on the historical logistics data and the target logistics data based on the characteristic representation matching degree, so as to at least save the target logistics data. .

Based on the above content, that is, step S110, step S120, step S130 and step S140, since the historical logistics data and the target logistics data are used to calculate the feature representation matching degree, the historical logistics data and the target logistics data can be compared based on the obtained feature representation matching degree. The target logistics data is subject to data update management and control. In this way, compared with the conventional technical solution of data update control based directly on the formation time, the reliability of logistics data update can be improved, thereby improving the deficiencies in the existing technology.

Optionally, in some embodiments, the above-mentioned step S110, that is, the step of performing key information mining processing on the historical logistics data to be processed to output the historical key information feature representation corresponding to the historical logistics data, may be further Includes some of the following:

The optimized key information mining network is used to perform key information mining processing on the historical logistics data to be processed, so as to output the historical key information feature representation corresponding to the historical logistics data.

Optionally, in some embodiments, the above-mentioned step S120, that is, the step of performing key information mining processing on the target logistics data to be processed to output the target key information feature representation corresponding to the target logistics data, may be further Includes some of the following:

The optimized key information mining network is used to perform key information mining processing on the target logistics data to be processed, so as to output the target key information feature representation corresponding to the target logistics data.

Optionally, in some embodiments, the logistics data updating method may also include the following steps, such as performing network optimization processing to form an optimized key information mining network. Based on this, the steps of performing network optimization processing to form an optimized key information mining network may further include the following contents:

Extract typical logistics data combinations, and analyze multi-dimensional logistics characterization data corresponding to the typical logistics data combinations. The typical logistics data combinations include at least first typical logistics data and second typical logistics data. The second typical logistics data combinations are The data matching relationship between the logistics data and the first typical logistics data conforms to the preset data matching relationship, and the multi-dimensional logistics characterization data includes at least two dimensions of logistics characterization data, such as image dimensions and text dimensions;

Perform a hiding operation on the logistics characterization data of the first dimension in the multi-dimensional logistics characterization data to output the corresponding hidden multi-dimensional logistics characterization data. The first dimension may be any one of the at least two dimensions. Dimension, there is no specific limit on this. For example, when the first dimension belongs to the image dimension, part of the image corresponding to the logistics representation data of the first dimension can be hidden, such as replacing it with a frame of all-white image, etc. ;

Through the initial key information mining network, the hidden multi-dimensional logistics representation data is subjected to key information mining processing to output typical key information feature representations and hidden data restoration feature representations corresponding to the typical logistics data combinations;

Based on the first dimension of logistics characterization data, the hidden data restoration feature representation and the typical key information feature representation, the initial key information mining network is used to determine the learning cost value to output the initial key information Mining the corresponding restoration learning cost value and the corresponding key information learning cost value of the network;

According to the restored learning cost value and the key information learning cost value, the initial key information mining network is optimized to form an optimized key information mining network. For example, the restored learning cost value and the key information learning cost value can be The key information learning cost values are weighted and summed to output a total learning cost value, and then the initial key information mining network is optimized based on the total learning cost value to form an optimized key information mining network.

Optionally, in some embodiments, the hidden multi-dimensional logistics representation data is subjected to key information mining processing through an initial key information mining network to output typical key information corresponding to the typical logistics data combination. Feature representation and hidden data The steps to restore feature representation can further include the following:

The hidden multi-dimensional logistics characterization data is integrated into the distribution domain to output an initial feature representation corresponding to each dimension in the same distribution domain. For example, the hidden multi-dimensional logistics characterization data can be They are respectively mapped to the same distribution domain. In this way, the initial feature representation corresponding to each dimension in the same distribution domain can be obtained. The distribution domain can also be understood as a feature space or a data set, etc.;

Through the initial key information mining network, the initial feature representation is aggregated to output the typical key information feature representation corresponding to each typical logistics data in the typical logistics data combination;

According to the typical key information feature representation, the hidden data in the hidden multi-dimensional logistics representation data is restored, and the hidden data restoration feature representation is output. That is to say, based on the typical key information feature representation, the The feature representation corresponding to the first dimension is subjected to correlation analysis and prediction to output the corresponding hidden data to restore the feature representation.

Optionally, in some embodiments, the initial feature representation is aggregated through an initial key information mining network to output typical keys corresponding to each typical logistics data in the typical logistics data combination. The steps of information feature representation may further include the following:

Build an initial key information mining network, the initial key information mining network including a first key information mining unit and a second key information mining unit;

Through the first key information mining unit, the initial feature representation is cascaded and combined to output the corresponding cascaded initial feature representation (such as initial feature representation 1-initial feature representation 2-initial feature representation 3-initial feature representation 4), then perform key information mining processing on the cascaded initial feature representation, such as convolution operation or filtering processing, and output the first typical key information feature representation;

Through the second key information mining unit, each of the initial feature representations is subjected to key information mining processing, such as convolution operation or filtering processing, and a local key information feature representation is output, and then each of the local key information feature representations is Perform cascade combination (such as local key information feature representation 1-local key information feature representation 2-local key information feature representation 3-local key information feature representation 4), and output the second typical key information feature representation;

Perform a cascade combination of the first typical key information feature representation and the second typical key information feature representation (such as the first typical key information feature representation - the second typical key information feature representation) to output the typical logistics data Typical key information feature representation corresponding to the combination.

Optionally, in some embodiments, the initial key information mining network is learned based on the first dimension of logistics characterization data, the hidden data restoration feature representation and the typical key information feature representation. The step of determining the cost value to output the restoration learning cost value and the corresponding key information learning cost value corresponding to the initial key information mining network may further include the following contents:

Based on the first dimension of logistics representation data and the hidden data restoration feature representation, the initial key information mining network is performed to determine the first learning cost value, and the restoration learning code corresponding to the initial key information mining network is output. value;

Based on the characteristic representation of typical key information corresponding to the typical logistics data combination, the initial key information mining network is used to determine the second learning cost value, and the key information learning cost value corresponding to the initial key information mining network is output.

Optionally, in some implementations, based on the characteristic representation of typical key information corresponding to the typical logistics data combination, the initial key information mining network is used to determine the second learning cost value, and the initial key information mining network is outputted. The steps of learning the value of key information corresponding to the key information mining network can further include the following contents:

From the second typical logistics data included in the typical logistics data combination, relevant typical logistics data and non-related typical logistics data are determined, and the relevant typical logistics data and the first typical logistics data have a matching relationship (such as between the data The matching degree between is greater than the first preset matching degree, the first preset matching degree can be 0.6, 0.8, 0.9, etc.), the non-correlated typical logistics data and the first typical logistics data do not have a matching relationship ( If the matching degree between data is less than the second preset matching degree, the second preset matching degree may be 0.2, 0.3, 0.4, etc.);

According to the typical key information feature representation, the matching degree between the first typical logistics data and the related typical logistics data is analyzed to output the corresponding related dimension matching degree, and the first typical logistics data is analyzed The matching degree between the data and the non-related typical logistics data to output the corresponding non-related dimension matching degree;

Based on the relevant dimension matching degree and the non-relevant dimension matching degree, the key information learning cost corresponding to the initial key information mining network is analyzed (for example, the relevant dimension matching degree and the corresponding actual matching can be degree to calculate the error, and calculate the error between the non-correlated dimension matching degree and the corresponding actual matching degree, and then fuse the errors of the two dimensions to obtain the key information learning cost, or no longer have actual matching degree In the case of matching degree, the positive correlation value of the absolute difference between the relevant dimension matching degree and 1 can be used as the corresponding error, and the positive correlation value of the absolute difference between the non-correlated dimension matching degree and 0 can be used as the corresponding error. value as the corresponding error, and then fuse the errors in the two dimensions to obtain the key information learning cost value).

Optionally, in some embodiments, based on the first dimension of logistics characterization data and the hidden data restoration feature representation, the initial key information mining network is used to determine the first learning cost value, The step of outputting the restoration learning cost value corresponding to the initial key information mining network may further include the following content:

Perform restoration processing of characterization data based on the hidden data restoration feature representation. For example, the process of key information mining can be reversed to output corresponding restored logistics characterization data;

According to the first-dimensional logistics characterization data and the restored logistics characterization data, the initial key information mining network is used to determine the first learning cost value, and the restored learning cost value corresponding to the initial key information mining network is output. ; That is to say, the difference between the first dimension of logistics representation data and the reduced logistics representation data can be calculated to obtain the reduction learning cost value corresponding to the initial key information mining network.

Optionally, in some embodiments, the above-mentioned step S140, that is, performing data update management and control on the historical logistics data and the target logistics data based on the characteristic representation matching degree, so as to at least update the target logistics data. The steps for saving logistics data can further include the following:

Extract the pre-configured reference matching degree. The specific value of the reference matching degree is not limited and can be selected and configured according to actual needs, such as 0.3, 0.5, 0.7, etc.;

Perform size comparison processing on the reference matching degree and the feature representation matching degree;

When the characteristic representation matching degree is greater than or equal to the reference matching degree, both the historical logistics data and the target logistics data are saved;

When the feature representation matching degree is less than the reference matching degree, the historical logistics data is discarded and the target logistics data is saved.

Optionally, in other embodiments, the above-mentioned step S140, that is, performing data update management and control on the historical logistics data and the target logistics data based on the characteristic representation matching degree, so as to at least update the historical logistics data and the target logistics data. The step of saving the target logistics data may further include the following contents:

Extract the pre-configured reference matching degree. The specific value of the reference matching degree is not limited and can be selected and configured according to actual needs, such as 0.3, 0.5, 0.7, etc.;

Perform size comparison processing on the reference matching degree and the feature representation matching degree;

When the characteristic representation matching degree is greater than or equal to the reference matching degree, both the historical logistics data and the target logistics data are saved;

When the matching degree of the feature representation is less than the reference matching degree, part of the data in the historical logistics data is discarded, part of the historical logistics data that has not been discarded is saved, and the target logistics data is processed save.

In conjunction with Figure 3, an embodiment of the present invention also provides a logistics data update system, which can be applied to the above-mentioned logistics data update platform. Wherein, the logistics data updating system may include:

The first key information mining module is used to perform key information mining processing on the historical logistics data to be processed to output the historical key information feature representation corresponding to the historical logistics data. The historical logistics data includes each piece of history within the historical time period. Logistics information, one piece of historical logistics information corresponds to one logistics cargo;

The second key information mining module is used to perform key information mining processing on the target logistics data to be processed, so as to output the target key information feature representation corresponding to the target logistics data. The target logistics data includes each target in the current time period. Logistics information, one piece of target logistics information corresponds to one logistics cargo;

A matching degree calculation module, configured to perform matching degree calculation processing on the historical key information feature representation and the target key information feature representation, to output the corresponding feature representation matching degree;

A data update management and control module is configured to perform data update management and control on the historical logistics data and the target logistics data based on the characteristic representation matching degree, so as to at least save the target logistics data.

Optionally, in some implementations, the first key information mining module is specifically configured to use an optimized key information mining network to perform key information mining processing on the historical logistics data to be processed, so as to output the historical logistics data corresponding to Representation of historical key information features;

Optionally, in some implementations, the second key information mining module is specifically configured to: use the optimized key information mining network to perform key information mining processing on the target logistics data to be processed, so as to output the target logistics The target key information features corresponding to the data are represented.

Optionally, in some implementations, the logistics data updating system may also include other software functional modules, and the other software functional modules may be used to: perform network optimization processing to form an optimized key information mining network.

Optionally, in some implementations, the other software function modules can be specifically used to: extract typical logistics data combinations, and analyze the multi-dimensional logistics characterization data corresponding to the typical logistics data combinations. The typical logistics data combination at least includes first typical logistics data and second typical logistics data. The data matching relationship between the second typical logistics data and the first typical logistics data conforms to the preset data matching relationship. The multi-dimensional The logistics characterization data includes at least two dimensions of logistics characterization data; the first dimension of the logistics characterization data in the multi-dimensional logistics characterization data is hidden to output the corresponding hidden multi-dimensional logistics characterization data; through the initial key information Mining the network, the hidden multi-dimensional logistics representation data is subjected to key information mining processing to output the typical key information feature representation corresponding to the typical logistics data combination and the hidden data restoration feature representation; logistics based on the first dimension The characterization data, the hidden data restoration feature representation and the typical key information feature representation are used to determine the learning cost value of the initial key information mining network to output the restoration learning cost value corresponding to the initial key information mining network. and the corresponding key information learning cost value; based on the restored learning cost value and the key information learning cost value, the initial key information mining network is optimized to form an optimized key information mining network.

Optionally, in some implementations, the data update management and control module is specifically used to:

Extract a preconfigured reference matching degree; compare the reference matching degree and the feature representation matching degree; when the feature representation matching degree is greater than or equal to the reference matching degree, compare the history Both the logistics data and the target logistics data are saved; when the characteristic representation matching degree is less than the reference matching degree, the historical logistics data is discarded and the target logistics data is saved.

In summary, the present invention provides a method and system for updating logistics data, which can perform key information mining processing on the historical logistics data to be processed, so as to output the historical key information feature representation corresponding to the historical logistics data; the target logistics to be processed The data is processed for key information mining to output the target key information feature representation corresponding to the target logistics data; the matching degree calculation process is performed on the historical key information feature representation and the target key information feature representation to output the corresponding features to represent the matching degree; based on the feature representation Matching degree, perform data update control on historical logistics data and target logistics data, so as to at least save the target logistics data. Based on the above content, since the feature representation matching degree of historical logistics data and target logistics data is calculated, data update management and control of historical logistics data and target logistics data can be carried out based on the obtained feature representation matching degree. In this way, compared with Conventional technical solutions for data update management and control based directly on formation time can improve the reliability of logistics data updates, thus improving the shortcomings of existing technologies.

In the several embodiments provided by the embodiments of the present invention, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device and method embodiments described above are only illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show possible implementation architectures of the devices, methods and computer program products according to multiple embodiments of the present invention. Functionality and operation. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code that contains one or more components for implementing the specified logical function(s). Executable instructions. It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two consecutive blocks may actually execute substantially in parallel, or they may sometimes execute in the reverse order, depending on the functionality involved. It will also be noted that each block of the block diagram and/or flowchart illustration, and combinations of blocks in the block diagram and/or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts. , or can be implemented using a combination of specialized hardware and computer instructions.

In addition, each functional module in various embodiments of the present invention can be integrated together to form an independent part, each module can exist alone, or two or more modules can be integrated to form an independent part.

If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention essentially or the part that contributes to the existing technology or the 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, including Several instructions are used to cause a computer device (which may be a personal computer, electronic device, or network device, etc.) to execute all or part of the steps of the method described in various embodiments of the present invention. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. . It should be noted that, as used herein, the terms "include", "comprises" or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article or device that includes a series of elements not only includes those elements, It also includes other elements not expressly listed or inherent in the process, method, article or equipment. Without further limitation, an element defined by the statement "comprises a..." does not exclude the presence of additional identical elements in a process, method, article, or apparatus that includes the stated element.

The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims (10)

1. A method for updating logistic data, comprising:

performing key information mining processing on historical logistics data to be processed so as to output historical key information characteristic representation corresponding to the historical logistics data, wherein the historical logistics data comprises each piece of historical logistics information in a historical time period, and one piece of historical logistics information corresponds to one logistics cargo;

performing key information mining processing on target logistics data to be processed to output target key information characteristic representation corresponding to the target logistics data, wherein the target logistics data comprises target logistics information of each item in the current time period, and one piece of target logistics information corresponds to one logistics cargo;

performing matching degree calculation processing on the historical key information feature representation and the target key information feature representation to output corresponding feature representation matching degree;

And based on the characteristic representation matching degree, carrying out data updating management and control on the historical logistics data and the target logistics data so as to at least save the target logistics data.

2. The method for updating logistics data according to claim 1, wherein the step of performing key information mining processing on the historical logistics data to be processed to output a historical key information feature representation corresponding to the historical logistics data comprises the steps of:

performing key information mining processing on the historical logistics data to be processed by using an optimized key information mining network so as to output a historical key information characteristic representation corresponding to the historical logistics data;

the step of performing key information mining processing on target logistics data to be processed to output target key information characteristic representation corresponding to the target logistics data comprises the following steps:

and carrying out key information mining processing on target logistics data to be processed by utilizing the optimized key information mining network so as to output target key information characteristic representation corresponding to the target logistics data.

3. The method for updating physical distribution data according to claim 2, wherein the method for updating further comprises: performing network optimization processing to form an optimized key information mining network;

The step of performing network optimization processing to form an optimized key information mining network comprises the following steps:

extracting a typical logistics data combination, and analyzing multi-dimensional logistics characterization data corresponding to the typical logistics data combination, wherein the typical logistics data combination at least comprises first typical logistics data and second typical logistics data, a data matching relationship between the second typical logistics data and the first typical logistics data accords with a preset data matching relationship, and the multi-dimensional logistics characterization data comprises logistics characterization data with at least two dimensions;

hiding the logistics characterization data of the first dimension in the multi-dimensional logistics characterization data to output corresponding hidden multi-dimensional logistics characterization data;

performing key information mining processing on the hidden multidimensional logistics characterization data through an initial key information mining network to output typical key information characteristic representation and hidden data restoration characteristic representation corresponding to the typical logistics data combination;

based on the logistics characterization data, the hidden data reduction feature representation and the typical key information feature representation of the first dimension, determining a learning cost value of the initial key information mining network so as to output a reduction learning cost value corresponding to the initial key information mining network and a corresponding key information learning cost value;

And carrying out optimization processing on the initial key information mining network according to the reduction learning cost value and the key information learning cost value so as to form an optimized key information mining network.

4. The method for updating logistics data according to claim 3, wherein the step of performing a learning cost value determining operation on the initial key information mining network based on the logistics characterization data, the hidden data restoration feature representation and the typical key information feature representation in the first dimension to output a restoration learning cost value corresponding to the initial key information mining network and a corresponding key information learning cost value comprises:

based on the logistics characterization data of the first dimension and the hidden data restoration characteristic representation, carrying out first learning cost value determination operation on the initial key information mining network, and outputting restoration learning cost values corresponding to the initial key information mining network;

and according to the typical key information characteristic representation corresponding to the typical logistics data combination, carrying out the second learning cost value determining operation on the initial key information mining network, and outputting the key information learning cost value corresponding to the initial key information mining network.

5. The method for updating physical distribution data according to claim 4, wherein the step of performing a second learning cost value determination operation on the initial critical information mining network according to the representative critical information feature representation corresponding to the representative physical distribution data combination, and outputting the critical information learning cost value corresponding to the initial critical information mining network includes:

determining relevant typical logistics data and irrelevant typical logistics data from second typical logistics data included in the typical logistics data combination, wherein the relevant typical logistics data and the first typical logistics data have a matching relationship, and the irrelevant typical logistics data and the first typical logistics data do not have a matching relationship; according to the typical key information characteristic representation, analyzing the matching degree between the first typical logistics data and the related typical logistics data to output corresponding related dimension matching degree, and analyzing the matching degree between the first typical logistics data and the non-related typical logistics data to output corresponding non-related dimension matching degree; based on the relevant dimension matching degree and the irrelevant dimension matching degree, analyzing the key information learning cost value corresponding to the initial key information mining network;

The step of restoring the characteristic representation based on the logistics characterization data and the hidden data in the first dimension, performing a first learning cost value determining operation on the initial key information mining network, and outputting a restoring learning cost value corresponding to the initial key information mining network includes:

performing restoration processing of the characterization data based on the hidden data restoration characteristic representation to output corresponding restored logistics characterization data; and according to the logistics characterization data and the restored logistics characterization data of the first dimension, determining a first learning cost value of the initial key information mining network, and outputting a restored learning cost value corresponding to the initial key information mining network.

6. The method for updating logistics data according to claim 3, wherein the step of performing key information mining processing on the hidden multidimensional logistics characterization data through an initial key information mining network to output a typical key information feature representation and a hidden data restoration feature representation corresponding to the typical logistics data combination comprises the steps of:

carrying out integration processing of the distribution domain on the hidden multidimensional stream characterization data so as to output initial feature representations corresponding to each dimension in the same distribution domain;

The initial characteristic representation is subjected to aggregation processing through an initial key information mining network so as to output a typical key information characteristic representation corresponding to each typical logistics data in the typical logistics data combination;

and according to the typical key information characteristic representation, carrying out reduction processing on hidden data in the hidden multidimensional stream characterization data, and outputting hidden data reduction characteristic representation.

7. The method for updating physical distribution data according to any one of claims 1 to 6, wherein the step of performing data updating management on the historical physical distribution data and the target physical distribution data based on the feature representation matching degree to save at least the target physical distribution data comprises the steps of:

extracting a pre-configured reference matching degree;

comparing the reference matching degree with the characteristic representation matching degree;

storing the historical logistics data and the target logistics data under the condition that the characteristic representation matching degree is larger than or equal to the reference matching degree;

and discarding the historical logistics data and storing the target logistics data under the condition that the characteristic representation matching degree is smaller than the reference matching degree.

8. A system for updating logistic data, comprising:

the first key information mining module is used for carrying out key information mining processing on historical logistics data to be processed so as to output historical key information characteristic representation corresponding to the historical logistics data, wherein the historical logistics data comprises each piece of historical logistics information in a historical time period, and one piece of historical logistics information corresponds to one logistics cargo;

the second key information mining module is used for carrying out key information mining processing on target logistics data to be processed so as to output target key information characteristic representation corresponding to the target logistics data, wherein the target logistics data comprises each item of target logistics information in the current time period, and one piece of target logistics information corresponds to one piece of logistics goods;

the matching degree calculation module is used for carrying out matching degree calculation processing on the historical key information feature representation and the target key information feature representation so as to output corresponding feature representation matching degree;

and the data updating management and control module is used for carrying out data updating management and control on the historical logistics data and the target logistics data based on the characteristic representation matching degree so as to at least save the target logistics data.

9. The system for updating logistics data according to claim 8, wherein the first key information mining module is specifically configured to:

performing key information mining processing on the historical logistics data to be processed by using an optimized key information mining network so as to output a historical key information characteristic representation corresponding to the historical logistics data;

the second key information mining module is specifically configured to:

and carrying out key information mining processing on target logistics data to be processed by utilizing the optimized key information mining network so as to output target key information characteristic representation corresponding to the target logistics data.

10. The system for updating physical distribution data according to claim 8, wherein the data updating management module is specifically configured to:

extracting a pre-configured reference matching degree;

comparing the reference matching degree with the characteristic representation matching degree;

storing the historical logistics data and the target logistics data under the condition that the characteristic representation matching degree is larger than or equal to the reference matching degree;

and discarding the historical logistics data and storing the target logistics data under the condition that the characteristic representation matching degree is smaller than the reference matching degree.