CN112101839A

CN112101839A - Method for establishing express delivery time prediction model, prediction method and related equipment

Info

Publication number: CN112101839A
Application number: CN202010660927.3A
Authority: CN
Inventors: 刘成亮; 韦家强
Original assignee: Shanghai Xunmeng Information Technology Co Ltd
Current assignee: Shanghai Xunmeng Information Technology Co Ltd
Priority date: 2020-07-10
Filing date: 2020-07-10
Publication date: 2020-12-18

Abstract

The invention provides a method for establishing an express delivery time prediction model, a prediction method and related equipment, wherein the method for establishing the express delivery time prediction model comprises the following steps: acquiring historical logistics track data; extracting characteristic data of the logistics track from the historical logistics track data, wherein the characteristic data at least comprises track segment characteristics and operation time period characteristics; generating a plurality of track samples, wherein each track sample comprises characteristic data of a logistics track and delivery duration of the logistics track; training a classification model by using the track sample; and taking the trained classification model as the express delivery time prediction model. The method and the system provided by the invention improve the prediction accuracy of the express delivery time.

Description

Method for establishing express delivery time prediction model, prediction method and related equipment

Technical Field

The invention relates to the field of computer application, in particular to a method for establishing an express delivery time prediction model, a prediction method and related equipment.

Background

With the development of network information technology, no matter a logistics company or an e-commerce platform, the function of express delivery estimated delivery time is provided for users generally, so that the users can conveniently arrange signing-in time reasonably, and shopping experience of the users is improved.

At present, the estimated delivery time of the commonly adopted express delivery is adopted, or the average delivery time of the historical track is directly used as the estimated delivery time; or acquiring a current express delivery site, and predicting the express delivery time based on the average delivery time between the site and the delivery position in the historical track.

However, in the two methods, only the departure point and the arrival point are considered, and other influences on the express delivery time are not considered together, so that it is difficult to improve the prediction accuracy of the express delivery time.

Therefore, how to process the logistics track data to increase factors influencing express delivery time prediction and further improve the prediction accuracy of the express delivery time is a technical problem to be solved urgently by technical staff in the field.

Disclosure of Invention

In order to overcome the defects of the related technologies, the invention provides a method for establishing a prediction model of express delivery time, a prediction method and related equipment, and further through processing logistics track data, factors influencing the prediction of the express delivery time are increased, and the prediction accuracy of the express delivery time is further improved.

According to one aspect of the invention, a method for establishing a express delivery time prediction model is provided, and the method comprises the following steps:

acquiring historical logistics track data;

extracting characteristic data of the logistics track from the historical logistics track data, wherein the characteristic data at least comprises track segment characteristics and operation time period characteristics;

generating a plurality of track samples, wherein each track sample comprises characteristic data of a logistics track and delivery duration of the logistics track;

training a classification model by using the track sample; and

and taking the trained classification model as the express delivery time prediction model.

In some embodiments of the present invention, after the obtaining the historical logistics trajectory data and before the extracting the feature data of the logistics trajectory from the historical logistics trajectory data, further includes:

according to at least part of key links of the logistics track, performing data cleaning on the historical logistics track data; and/or

And performing data cleaning on the historical logistics track data according to at least part of abnormal parcel types.

In some embodiments of the invention, the key links include at least a package pick-up link, an origin branch link, a trunk transport link, a destination branch link, and a package delivery link.

In some embodiments of the present invention, the performing data cleansing on the historical logistics track data according to at least part of the key links of the logistics track includes:

according to a parcel collecting link of the logistics track, performing data cleaning on the historical logistics track data; and/or

And performing data cleaning on the historical logistics track data according to a parcel dispatching link of the logistics track.

In some embodiments of the present invention, the package collecting according to the logistics track includes:

cleaning a plurality of collecting records aiming at the same logistics track; and/or

And cleaning the abnormal collecting records aiming at the same logistics track.

In some embodiments of the present invention, in the package delivery according to logistics track link, the performing data cleaning on the historical logistics track data includes:

cleaning a plurality of signing records aiming at the same logistics track; and/or

And cleaning the abnormal signing record aiming at the same logistics track.

In some embodiments of the invention, the exception package types include at least returned packages, rejected packages, false packages, and forward packages.

In some embodiments of the invention, the track segment characteristics include at least a time interval between parcels from one site to another.

In some embodiments of the present invention, the time interval between packages in the logistics track from one station to another is calculated based on historical logistics track data including the two stations.

In some embodiments of the invention, the track segment characteristics are encoded at a plurality of temporal granularities.

In some embodiments of the invention, the sites include a collecting site, an originating distribution site, a destination distribution site, and a serving site.

In some embodiments of the invention, the operational time period characteristics include at least one or more of a time period during which the package is picked up, a time period during which the package is checked in, and a time period during which the logistics operation is performed on the package at a site.

In some embodiments of the invention, the period of time during which the logistics operation is performed on the package at a site is calculated based on a plurality of historical logistics trajectory data comprising the site.

In some embodiments of the invention, the operating period characteristics are encoded at a time granularity.

In some embodiments of the invention, the logistics operation includes arrival of the package at the site and/or departure of the package from the site.

In some embodiments of the invention, the characteristic data further comprises a package pull-out location characteristic comprising a package pull-out location, a package delivery location, and a time interval from the package pull-out location to the package delivery location.

In some embodiments of the invention, the characteristic data further comprises a courier company characteristic.

In some embodiments of the invention, the training a classification model using the trajectory samples comprises:

sequencing a plurality of track samples from far to near according to the set logistics operation time;

taking the first N% of track samples as a training set according to the sequencing result to train the classification model, wherein N is a constant which is more than 1 and less than 100;

the last (100-N)% of the trace samples are used as a test set according to the sorting result to test the trained classification model.

In some embodiments of the invention, further comprising:

and evaluating the express delivery time prediction model according to the prediction accuracy and the prediction timeliness.

According to another aspect of the present invention, there is also provided a method for predicting delivery time of an express delivery, including:

acquiring current logistics track data;

extracting characteristic data of the acquired logistics track data;

inputting the extracted feature data into a express delivery time prediction model, the express delivery time prediction model being established via the method of establishing the express delivery time prediction model as described above; and

and forecasting the express delivery time according to the output of the express delivery time forecasting model.

According to another aspect of the present invention, there is also provided an apparatus for creating a prediction model of express delivery time, including:

the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is configured to acquire historical logistics track data;

a first extraction module configured to extract feature data of the logistics track from the historical logistics track data, the feature data including at least track segment features and operation period features;

the generating module is configured to generate a plurality of track samples, and each track sample comprises characteristic data of a logistics track and delivery duration of the logistics track;

a training module configured to train a classification model using the trajectory samples; and

and the establishing module is configured to take the trained classification model as the express delivery time prediction model.

According to another aspect of the present invention, there is also provided an express delivery time prediction apparatus, including:

the second acquisition module is configured to acquire current logistics track data;

the second extraction module is configured to extract the characteristic data of the acquired logistics track data;

an input module configured to input the extracted feature data into an express delivery time prediction model, the express delivery time prediction model being established via a method of establishing an express delivery time prediction model as described above; and

a prediction module configured to predict a time of delivery of the delivery based on an output of the time of delivery prediction model.

According to still another aspect of the present invention, there is also provided an electronic apparatus, including: a processor; a storage medium having stored thereon a computer program which, when executed by the processor, performs the steps as described above.

According to yet another aspect of the present invention, there is also provided a storage medium having stored thereon a computer program which, when executed by a processor, performs the steps as described above.

Compared with the prior art, the invention has the advantages that:

according to the method, factors influencing express delivery time prediction are increased by extracting characteristics at least including track fragment characteristics and operation time period characteristics from historical logistics track data; on the premise of providing a plurality of factors influencing the express delivery time prediction, the classification model is adopted for learning, so that the specific influence of the factors influencing the express delivery time prediction on the express delivery time can be learned, and the express delivery time prediction accuracy is further improved.

Drawings

The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.

Fig. 1 shows a flowchart of a method for establishing a prediction of delivery time of a courier according to an embodiment of the present invention.

Fig. 2 shows a flow chart of cleaning according to at least part of the key links of the logistics track and at least part of the abnormal parcel types according to an embodiment of the invention.

Fig. 3 shows a flow chart of cleaning in a parcel collecting link and a parcel delivering link according to a logistics track according to an embodiment of the invention.

Fig. 4 shows a flow chart for cleansing a plurality of blanket records and abnormal blanket records according to an embodiment of the invention.

FIG. 5 illustrates a flow diagram for scrubbing multiple check-in records and exception preceding records, according to an embodiment of the present invention.

FIG. 6 is a flow diagram illustrating sample processing prior to classification model training in accordance with an embodiment of the present invention.

Fig. 7 is a flowchart illustrating an express delivery time prediction method according to an embodiment of the present invention.

Fig. 8 is a block diagram illustrating an apparatus for modeling express delivery time prediction according to an embodiment of the present invention.

Fig. 9 is a block diagram showing an express delivery time prediction apparatus according to an embodiment of the present invention.

Fig. 10 schematically illustrates a computer-readable storage medium in an exemplary embodiment of the invention.

Fig. 11 schematically illustrates an electronic device in an exemplary embodiment of the invention.

Detailed Description

Example embodiments will now be described more fully with reference to the accompanying drawings. Example embodiments may, however, be embodied in many different forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

Furthermore, the drawings are merely schematic illustrations of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repetitive description will be omitted. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor devices and/or microcontroller devices.

The flow charts shown in the drawings are merely illustrative and do not necessarily include all of the steps. For example, some steps may be decomposed, and some steps may be combined or partially combined, so that the actual execution sequence may be changed according to the actual situation.

In various embodiments of the present invention, the present invention is applied to a logistics platform or an e-commerce platform, but the present invention is not limited thereto, and the present invention may also be applied to any other platform that needs to predict delivery time of a courier.

Fig. 1 shows a flowchart of a method for establishing a express delivery time prediction model according to an embodiment of the present invention. The method for establishing the express delivery time prediction model comprises the following steps:

step S110: and acquiring historical logistics track data.

Specifically, the historical logistics track data can be obtained from a database of the logistics platform. When the method is applied to the e-commerce platform, the historical track data of each logistics platform can be obtained from the databases of a plurality of logistics platforms.

Step S120: extracting characteristic data of the logistics track from the historical logistics track data, wherein the characteristic data at least comprises track segment characteristics and operation time period characteristics.

Specifically, in the process of the logistics track, the package passes through a plurality of stations of the collecting station, the originating distribution station, the destination distribution station and the dispatching station. However, the historical unit quantity on the complete track from the collecting site to the sending site is sparse, and the rule of the logistics track is difficult to learn deeply. Therefore, the logistics track segment can be divided by the sites in step S120, so that the track segment characteristics at least include the time interval from one site to another. For example, the time interval from the pick-up site to the originating distribution site (originating branch), the time interval from the originating distribution site to the destination distribution site (trunk), and the time interval from the destination distribution site to the dispatch site (destination branch). The invention is not limited in this regard and the track segment characteristics may also include the time interval of a flow wrapping between more stations when more stations are included in the flow track.

Further, the time interval between packages from one station to another station in the logistics track can be calculated according to a plurality of historical logistics track data comprising the two stations. For example, according to a plurality of historical logistics track data including the two stations, the time interval of circulation of the parcels between the two stations in the plurality of historical logistics track data is obtained, and the average value, the median and the like of the plurality of time intervals are calculated mathematically to obtain the time interval of the parcels from one station to the other station in the logistics track.

In the foregoing embodiment, since track segment characteristics include at least the time interval of a parcel from one site to another, to represent the time interval, the track segment characteristics are encoded at a plurality of time granularities. Specifically, the time granularity may include, for example, weeks, days, hours, minutes, and the like. The track segment is encoded according to a plurality of time granularities, for example, two granularities of "day" and "hour" are encoded, and a three-integer number is used for encoding, wherein the hundred digits represent the number of days apart, the ten digits and the ones represent the hour time period for the track segment to complete, and when the time interval from one station to another station is encoded as "315", the current track segment is represented and ends 15 points after 3 days (which may represent that the current track segment ends 15 points after 3 days from the beginning of the track segment, or may represent that the current track segment ends 15 points after 3 days from the beginning of the complete track segment, and may be set as required, and the present invention is not limited thereto). The above is merely an example of the encoding manner for describing the time interval between parcels from one station to another station, and the present invention is not limited thereto. In this embodiment, the time interval between the package from one site to another site in the track segment feature further includes the site numbers of the one site and the another site.

Further, track segment characteristics may not be so limited, and track segment characteristics may also include, for example, the transport (e.g., shipping, trucking, train transportation, etc.) of the track segment (wrapping from one station to another). Other features of relevance in the track segment are also within the scope of the present invention.

In particular, considering that the time periods (such as morning, afternoon, evening, etc.) of the operation of the packages in each link are different, the transportation of the packages is also influenced. For example, packages that are pulled in the morning are typically not entered into the originating distribution site until that day, and packages that are pulled in the afternoon or evening are typically not entered into the originating distribution site until that day. Thus, the characteristic data of the logistics trajectory can also include the operating period characteristics.

The operational time period characteristics may include one or more of a time period during which the package is picked up, a time period during which the package is checked in, and a time period during which the logistics operation is performed on the package at a site. The logistics operation may be the arrival of the package at the site. The logistics operation may also be for parcels to leave the site. Further, the logistics operations may include arrival of the package at the site and departure of the package from the site. Thus, the operational time of a package at the site is represented by the time period the package arrived at and/or departed from the site.

Further, the period of time for performing the logistics operation on the package at a site is calculated according to a plurality of pieces of historical logistics track data containing the site. For example, the clustering may be performed according to different periods of time when the logistics operation is performed on the parcel at a site in the plurality of pieces of historical logistics track data including the site, and the period of time when the clustering number is the largest may be used as the period of time when the logistics operation is performed on the parcel at the site. For another example, the average value, median and other data of different time periods may be calculated as the time period for performing the logistics operation on the parcel at the one site according to different time periods for performing the logistics operation on the parcel at the one site in the plurality of pieces of historical logistics track data including the site. For another example, the calculated time interval from one site to another site of the package in the track segment characteristics may be used to calculate the time interval of the logistics operation performed on the package at one site according to the calculated time interval. Further, the time acquisition period in the logistics track can be calculated by any one of the two methods (aggregation and mean/median calculation) or the actual time acquisition period in the historical logistics track data can be used as the time acquisition period in the characteristic. The sign-in time period in the logistics track can be calculated in any one of the three manners (aggregation, average/median calculation, and sequential superposition determination of self-acquisition stations according to the calculated time intervals), or the actual sign-in time period in the historical logistics track data is used as the sign-in time period in the characteristic.

In the foregoing embodiments, in order to represent the time period of the operation, the operation period characteristics may be encoded in one time granularity, and the present invention is not limited thereto. Specifically, the time granularity may include, for example, weeks, days, hours, minutes, and the like. The time period of operation may be encoded in "hour" granularity. For example, a two-bit integer number is used for encoding to indicate the hour period corresponding to the operation. When the time period for an operation is encoded as "15," it means that the current operation (e.g., pull, arrive at a site, leave a site, sign-in) is operating at 15 o' clock.

The above is merely an illustrative description of the operational time period features provided by the present invention and the present invention is not so limited.

In some embodiments of the invention, considering that the pick-up and delivery location also affects the delivery time of the express, the feature data of the logistics track may further include the pick-up and delivery location feature, so as to enrich the dimension of the feature vector. The package pick-up and delivery site characteristics comprise a package pick-up site, a delivery site and a time interval from the package pick-up site to the delivery site. The package collecting place and the package sending place can be represented by cities, regions or customized regional division. Therefore, factors of the pick-up place and the delivery place can be added into the feature vector.

In the following, the city is taken as an example of representing a pickup location and a delivery location, the pickup city code can be of an integer data type, and each code uniquely corresponds to one city, which is referred to as a pickup city. The tile city codes may also be of an integer data type, with each code uniquely corresponding to a city, referred to herein as a tile city. The time interval for package flows between solicited cities may be encoded at multiple time granularities. For example, the codes are encoded in two time granularities of "days" and "hours", and one whole number is used for encoding, wherein the hundreds number represents the hour time period from the package city to the delivery city in days, tens number and units (the encoding mode can be similar to the time interval from one site to another site of the package).

In some embodiments of the invention, the characteristics of the courier company can be added to the characteristic data, considering that different courier companies have differences in transportation network, vehicle frequency and operation time design. Thus, the characteristic data may also include courier company characteristics. Express company features may be coded with integer data types for uniquely referring to an express company.

Step S130: generating a plurality of track samples, wherein each track sample comprises characteristic data of the logistics track and the delivery time of the logistics track.

Step S140: and training a classification model by using the track sample.

Specifically, the present invention may adopt classification models such as XGBoost, GB (Gradient Boosting), gbdt (Gradient Boosting Decision tree), etc., and the present invention is not limited thereto. Various machine learning models, deep learning models, etc. with classification functions are within the scope of the present invention.

Step S150: and taking the trained classification model as the express delivery time prediction model.

In the method for establishing the express delivery time prediction model, factors influencing express delivery time prediction are increased by extracting characteristics at least comprising track segments and characteristics of operation time periods from historical logistics track data; on the premise of providing a plurality of factors influencing the express delivery time prediction, the classification model is adopted for learning, so that the specific influence of the factors influencing the express delivery time prediction on the express delivery time can be learned, and the express delivery time prediction accuracy is further improved.

Referring now to fig. 2, fig. 2 illustrates a flow diagram for cleaning at least some critical sections of a logistic track and at least some abnormal parcel types, according to an embodiment of the present invention.

In this embodiment, after the historical logistics track data is acquired in step S110 in fig. 1, and before the feature data of the logistics track is extracted from the historical logistics track data in step S120, step S101 and step S102 shown in fig. 2 may be further included:

step S101: and performing data cleaning on the historical logistics track data according to at least part of key links of the logistics track.

Step S102: and performing data cleaning on the historical logistics track data according to at least part of abnormal parcel types.

Specifically, although fig. 2 shows step S101 and step S102, the execution order of step S101 and step S102 may be performed in reverse order, simultaneously, separately, and step S101 and step S102 may be performed separately, and such variations are within the scope of the present invention.

Specifically, the key links in step S101 at least include a package collecting link, an originating branch link, a trunk transportation link, a destination branch link, and a package delivering link. The package collecting link mainly refers to the process that an express company obtains packages from users and starts logistics transportation. In the package collecting link, factors influencing the final delivery time can comprise different express companies, package collecting time points and package collecting cities. The initial branch link mainly refers to the process from the package delivery from the collecting site to the initial distribution site. In the originating branch link, the factors affecting the final delivery time may include the originating branch transit time length. The trunk transportation link mainly refers to the process of transporting the packages from the originating distribution site to the destination distribution site. In the trunk transportation link, the factors affecting the final delivery time may include the time point when the parcel arrives at the originating distribution site and the length of the trunk transportation time. The destination branch link mainly refers to the process of transporting the packages from the destination distribution station to the piece dispatching station. In the destination branch link, factors influencing the final delivery time can include the time point when the package reaches the destination distribution station and the destination branch transportation time. The package delivery link mainly refers to that the package is sent from a delivery site and finally delivered to a user, and the package signing-in process is completed. In the parcel delivery link, factors affecting the final delivery time may include the time point when the parcel arrives at the delivery site, the city in which the parcel is delivered, and the like. The foregoing is merely a schematic illustration of the various key elements provided by the present invention. Further, the feature data in the step 5120 may also be extracted according to factors affecting the final delivery time in each key link, which is not described herein again.

Specifically, step 5101 will be described with reference to fig. 3-5, which are not repeated herein.

In step 5102, exception packages may include return packages, reject packages, dummy packages, forward packages, and the like. Specifically, the returned package is the package of the logistics track generated by returning the package to the merchant after the user signs the package. And performing data cleaning on the historical logistics track data according to the returned packages by rejecting the logistics track data corresponding to all the returned packages. The rejection package is a package which rejects the execution of the signing operation after the user receives the package. And removing logistics track data corresponding to all the label rejecting packages to perform data cleaning on the historical logistics track data according to the label rejecting packages. The false package is a package of a logistics track generated by means of counterfeiting and falsifying for avoiding punishment of an e-commerce platform merchant. And data cleaning of the historical logistics track data according to the false packages can be realized by removing the logistics track data corresponding to all the false packages. The express company transfers the package to other express companies to transport the package to generate the package of the logistics track aiming at the remote area which can not be covered by part of the package. The data cleaning of the historical logistics track data according to the transfer order package can be realized by removing the logistics track data corresponding to all the transfer order packages. Therefore, data cleaning is carried out on the historical logistics track data according to the one or more abnormal parcel types, so that the influence of abnormal tracks of the abnormal parcels on the prediction of the express delivery time of the normal parcels is avoided, and the prediction accuracy of the express delivery time is improved.

Referring now to fig. 3, fig. 3 is a flow chart illustrating a package pick-up and delivery for cleaning according to a logistics track, according to an embodiment of the present invention. Fig. 3 realizes step S101 shown in fig. 2 through step S111 and step S112.

Step S111: and performing data cleaning on the historical logistics track data according to a package collecting link of the logistics track.

Step S112: and performing data cleaning on the historical logistics track data according to a parcel dispatching link of the logistics track.

Specifically, although fig. 3 shows step S111 and step S112, the execution order of step S111 and step S112 may be performed in reverse order, simultaneously, separately, and step S111 and step S112 may be performed separately, and these variations are within the scope of the present invention. A specific implementation of step S111 can be explained with reference to fig. 4. The specific implementation of step S112 can be explained with reference to fig. 5.

Referring now to fig. 4, fig. 4 illustrates a flow chart for cleansing a plurality of package records and an abnormal package record according to an embodiment of the present invention. Fig. 4 shows the following steps:

step S121: and cleaning a plurality of collecting records aiming at the same logistics track.

Step S122: and cleaning the abnormal collecting records aiming at the same logistics track.

Specifically, when the package courier scans repeatedly, a plurality of package operation records are generated for the same freight note, and the time points are different, a plurality of package records are generated. Thus, step S121 may be cleaned as follows: and aiming at the same logistics track, only one acquisition record at the latest operation time point is reserved for a plurality of acquisition records, and the rest acquisition records are removed.

Specifically, when the operation of the package courier is abnormal, an abnormal track segment with a part of time points before package is generated, and an abnormal package record is generated. The method can eliminate all track segments of which the operation time points are positioned before the acquisition operation aiming at the same logistics track and the freight note with the abnormal acquisition record so as to clean the abnormal acquisition record.

Specifically, although fig. 4 shows step S121 and step S122, the execution order of step S121 and step S122 may be performed in reverse order, simultaneously, separately, and step S121 and step S122 may be performed separately, and these variations are within the scope of the present invention.

Referring now to FIG. 5, FIG. 5 illustrates a flow diagram for purging a plurality of check-in records and exception aforesaid records, in accordance with an embodiment of the present invention. Fig. 5 shows the following steps:

step S131: and cleaning a plurality of signing records aiming at the same logistics track.

Step S132: and cleaning the abnormal signing record aiming at the same logistics track.

Specifically, when the dispatch courier repeatedly scans to generate multiple signing operation records for the same freight order and the time points are different, multiple signing records are generated. Thus, step S131 may be cleaned as follows: and aiming at the same logistics track, only one signing record at the latest operation time point is reserved for the freight notes of the plurality of signing records, and the rest signing records are removed.

Specifically, when the dispatch courier operates abnormally, an abnormal track segment with a part of time points after signing is generated, and an abnormal signing record is generated. All track fragments of which the operation time points are positioned after the signing operation can be removed aiming at the freight notes with the abnormal signing records in the same logistics track, so that the abnormal signing records can be cleaned.

Specifically, although fig. 5 shows step S131 and step S132, the execution order of step S131 and step S132 may be performed in reverse order, simultaneously, separately, and step S131 and step S132 may be performed separately, and these variations are within the scope of the present invention.

Therefore, through the cleaning of the historical logistics track data, the characteristic data obtained in the step S120 can be more accurate, and the interference of abnormal data is reduced.

Turning next to FIG. 6, FIG. 6 illustrates a flow diagram of sample processing prior to classification model training in accordance with an embodiment of the present invention. Fig. 6 shows a specific implementation of step S140 in fig. 1. Fig. 6 shows the following steps in total:

step S141: and sequencing the plurality of track samples from far to near according to the set logistics operation time.

The set logistics operation time can be a pickup time, a dispatch time or an operation time at any site.

Step S142: and taking the first N% of track samples as a training set according to the sequencing result to train the classification model, wherein N is a constant which is more than 1 and less than 100.

Step S143: the last (100-N)% of the trace samples are used as a test set according to the sorting result to test the trained classification model.

In some embodiments, N is a constant less than 50 (e.g., N is 10, 20, 30, etc.). Therefore, the trajectory samples with more data and longer time are trained, and the trajectory samples with less data and shorter time are tested, so that the classification model with higher accuracy and timeliness can be obtained.

Specifically, when the test set is used for evaluation, the express delivery time prediction model can be evaluated according to the prediction accuracy and the prediction timeliness. The prediction accuracy is the proportion of the number of samples which are accurately predicted in the test set to the total number of samples in the test set. The prediction timeliness rate is the proportion of the number of samples whose predicted value (predicted arrival time) is not earlier than the actual value (actual arrival time) in the test set to the total number of samples in the test set. Therefore, the trained express delivery time prediction model is evaluated through the prediction accuracy and the prediction timeliness, so that the express delivery time prediction model can be evaluated, the express delivery time prediction model can be optimized based on the evaluation result, and the prediction accuracy and the prediction timeliness of the express delivery time prediction model are improved.

The above are merely a plurality of specific implementations of the present invention, and each implementation may be implemented independently or in combination, and the present invention is not limited thereto.

In another aspect of the invention, a method for predicting delivery time of an express delivery is also provided. As shown in fig. 7, the express delivery time prediction method includes the following steps:

step S210: and acquiring current logistics track data.

Step S220: and extracting the characteristic data of the acquired logistics track data.

Specifically, the extraction of the feature data in step S220 is similar to the extraction of the feature data in step S120 in fig. 1, for example, according to the content included in the feature data, the actual value of each feature data is extracted from the current logistics trajectory data, which is not described herein again.

Step S230: the extracted feature data is input into an express delivery time prediction model that is built via the method of building an express delivery time prediction model as described above.

Specifically, the express delivery time prediction Model may be stored in a PMML (Predictive Model Markup Language) format. The express delivery time prediction model may be an online model to allow real-time overlap. The express delivery time prediction model can also be a model under the line so as to accelerate the prediction time.

Step S240: and forecasting the express delivery time according to the output of the express delivery time forecasting model.

According to the express delivery time prediction method provided by the invention, the factors influencing the express delivery time prediction are increased by extracting the characteristics at least comprising track fragments and the characteristics of operation time periods from the historical logistics track data; on the premise of providing a plurality of factors influencing the express delivery time prediction, the classification model is adopted for learning, so that the specific influence of the factors influencing the express delivery time prediction on the express delivery time can be learned, and the express delivery time prediction accuracy is further improved.

Referring now to fig. 8, fig. 8 is a block diagram illustrating an apparatus for modeling express delivery time prediction according to an embodiment of the present invention. The device 300 for predicting the delivery time of a built express delivery comprises a generating module 210, a first sending module 220, a first obtaining module 310, a first extracting module 320, a generating module 330, a training module 340 and a building module 350.

The first acquisition module 310 is configured to acquire historical logistics trajectory data.

The first extraction module 320 is configured to extract feature data of the logistics trajectory from the historical logistics trajectory data, the feature data including at least a trajectory segment feature and an operation period feature.

The generating module 330 is configured to generate a plurality of trajectory samples, each of which includes characteristic data of a logistics trajectory and a delivery duration of the logistics trajectory.

The training module 340 is configured to train a classification model using the trace samples.

The building module 350 is configured to use the trained classification model as the express delivery time prediction model.

Fig. 8 is a schematic diagram illustrating the device 300 for creating a prediction model of express delivery time according to the present invention, and the splitting, merging, and adding of modules are within the scope of the present invention without departing from the spirit of the present invention. The device 300 for establishing the express delivery time prediction model according to the present invention may be implemented by software, hardware, firmware, plug-in, or any combination thereof, which is not limited thereto.

Referring now to fig. 9, fig. 9 is a block diagram illustrating an express delivery time prediction apparatus according to an embodiment of the present invention. The express delivery time prediction apparatus 400 includes a second obtaining module 410, a second extracting module 420, an inputting module 430, and a predicting module 440

The second obtaining module 410 is configured to obtain current logistics trajectory data.

The second extraction module 420 is configured to extract feature data of the acquired logistics trajectory data.

The input module 430 is configured to input the extracted feature data into an express delivery time prediction model that is built via a method of building an express delivery time prediction model as described above.

The prediction module 440 is configured to predict a time of delivery of the courier based on an output of the courier delivery time prediction model.

Fig. 9 is a schematic diagram illustrating an express delivery time prediction apparatus 400 provided by the present invention, and the module splitting, merging and adding are within the protection scope of the present invention without departing from the concept of the present invention. The express delivery time prediction apparatus 400 provided by the present invention may be implemented by software, hardware, firmware, plug-in, or any combination thereof, which is not limited to the present invention.

In the device 300 and the device 400 for predicting express delivery time of the exemplary embodiment of the present invention, factors affecting the prediction of express delivery time are increased by extracting characteristics including at least track segments and operation time period characteristics from the historical logistics track data; on the premise of providing a plurality of factors influencing the express delivery time prediction, the classification model is adopted for learning, so that the specific influence of the factors influencing the express delivery time prediction on the express delivery time can be learned, and the express delivery time prediction accuracy is further improved.

In an exemplary embodiment of the present invention, a computer-readable storage medium is further provided, on which a computer program is stored, which when executed by, for example, a processor, may implement the steps of the processing method for canceling an order as described in any one of the above embodiments. In some possible embodiments, the various aspects of the invention may also be implemented in the form of a program product comprising program code means for causing a terminal device to carry out the steps according to various exemplary embodiments of the invention described in the above-mentioned method of processing cancel order section of this description, when said program product is run on said terminal device.

Referring to fig. 10, a program product 700 for implementing the above method according to an embodiment of the present invention is described, which may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited in this regard and, in the present document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The computer readable storage medium may include a propagated data signal with readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A readable storage medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the tenant computing device, partly on the tenant device, as a stand-alone software package, partly on the tenant computing device and partly on a remote computing device, or entirely on the remote computing device or server. In the case of remote computing devices, the remote computing devices may be connected to the tenant computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).

In an exemplary embodiment of the invention, there is also provided an electronic device that may include a processor and a memory for storing executable instructions of the processor. Wherein the processor is configured to execute the steps of the processing method for canceling an order in any one of the above embodiments via executing the executable instructions.

As will be appreciated by one skilled in the art, aspects of the present invention may be embodied as a system, method or program product. Thus, various aspects of the invention may be embodied in the form of: an entirely hardware embodiment, an entirely software embodiment (including firmware, microcode, etc.) or an embodiment combining hardware and software aspects that may all generally be referred to herein as a "circuit," module "or" system.

An electronic device 500 according to this embodiment of the invention is described below with reference to fig. 11. The electronic device 500 shown in fig. 11 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present invention.

As shown in fig. 11, the electronic device 500 is embodied in the form of a general purpose computing device. The components of the electronic device 500 may include, but are not limited to: at least one processing unit 510, at least one memory unit 520, a bus 530 that couples various system components including the memory unit 520 and the processing unit 510, a display unit 540, and the like.

Wherein the storage unit stores program code executable by the processing unit 510 to cause the processing unit 510 to perform steps according to various exemplary embodiments of the present invention described in the above-mentioned processing method of canceling an order section of the present specification. For example, the processing unit 510 may perform the steps as shown in any one or more of fig. 1-7.

The memory unit 520 may include a readable medium in the form of a volatile memory unit, such as a random access memory unit (RAM)5201 and/or a cache memory unit 5202, and may further include a read only memory unit (ROM) 5203.

The memory unit 520 may also include a program/utility 5204 having a set (at least one) of program modules 5205, such program modules 5205 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.

Bus 530 may be one or more of any of several types of bus structures including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

The electronic device 500 may also communicate with one or more external devices 600 (e.g., keyboard, pointing device, bluetooth device, etc.), with one or more devices that enable a tenant to interact with the electronic device 500, and/or with any devices (e.g., router, modem, etc.) that enable the electronic device 500 to communicate with one or more other computing devices. Such communication may be through input/output (I/O) interfaces 550. Also, the electronic device 500 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN) and/or a public network, such as the Internet) via the network adapter 560. The network adapter 560 may communicate with other modules of the electronic device 500 via the bus 530. It should be appreciated that although not shown in the figures, other hardware and/or software modules may be used in conjunction with the electronic device 500, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.

Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein may be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiment of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a usb disk, a removable hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, or a network device, etc.) to execute the above processing method for canceling an order according to the embodiment of the present invention.

Compared with the prior art, the invention has the advantages that:

Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.

Claims

1. A method for establishing a prediction model of express delivery time, comprising:

acquiring historical logistics track data;

training a classification model by using the track sample; and

2. The method of creating a prediction model of express delivery time of claim 1, wherein after obtaining historical logistics trajectory data and before extracting characteristic data of a logistics trajectory from the historical logistics trajectory data, further comprising:

3. The method of claim 2, wherein the key links include at least a package pick-up link, a branch origination link, a trunk transportation link, a branch destination link, and a package delivery link.

4. The method of claim 2, wherein the data cleansing of the historical logistics track data based on at least a portion of the key links of the logistics track comprises:

5. The method of claim 4, wherein the package collecting step according to the logistics track comprises the following steps:

6. The method of claim 4, wherein the package delivery by logistics track link, the data cleansing of the historical logistics track data comprises:

And cleaning the abnormal signing record aiming at the same logistics track.

7. The method of modeling express delivery time of claim 2, wherein the abnormal package types include at least returned packages, rejected packages, false packages, and pickup packages.

8. The method of claim 1, wherein the track segment characteristics include at least a time interval between parcels from one location to another.

9. The method of claim 8, wherein the time interval between parcels in the logistics track from one station to another is calculated based on historical logistics track data comprising the two stations.

10. The method of creating a express delivery time prediction model of claim 8, wherein the track segment characteristics are encoded at a plurality of time granularities.

11. The method of claim 8, wherein the sites include a pickup site, an origination distribution site, a destination distribution site, and a dispatch site.

12. The method of claim 1, wherein the operational time period characteristics include at least one or more of a time period during which the package is picked up, a time period during which the package is signed in, and a time period during which the logistics operation is performed on the package at a site.

13. The method of claim 12, wherein the period of time that the logistics operation is performed on the package at a site is calculated based on historical logistics trajectory data including the site.

14. The method of creating a express delivery time prediction model of claim 12, wherein the operational period characteristics are encoded at a time granularity.

15. The method of creating a express delivery time prediction model of claim 12, wherein the logistics operation includes arrival of a parcel at the site and/or departure of a parcel from the site.

16. The method of claim 1, wherein the characteristic data further comprises package pickup location characteristics, the package pickup location characteristics including package pickup location, package delivery location, and a time interval from package pickup location to package delivery location.

17. The method of creating a express delivery time prediction model of claim 1, wherein the characteristic data further comprises express company characteristics.

18. The method of creating a express delivery time prediction model of claim 1, wherein the training a classification model using the trajectory samples comprises:

19. The method of building a predictive model of express delivery time of claim 1, further comprising:

20. An express delivery time prediction method is characterized by comprising the following steps:

acquiring current logistics track data;

extracting characteristic data of the acquired logistics track data;

inputting the extracted feature data into a courier delivery time prediction model, the courier delivery time prediction model being built via a method of building a courier delivery time prediction model according to any one of claims 1 to 19; and

21. An apparatus for creating a prediction model of delivery time for an item of courier, comprising:

22. An express delivery time prediction apparatus, comprising:

an input module configured to input the extracted feature data into a time of delivery prediction model established via a method of establishing a time of delivery prediction model according to any one of claims 1 to 19; and

23. An electronic device, characterized in that the electronic device comprises:

a processor;

a memory having stored thereon a computer program that, when executed by the processor, performs:

a method of building a prediction model of express delivery time according to any one of claims 1 to 19; and/or

The express delivery time prediction method of claim 20.

24. A computer-readable storage medium, having a computer program stored thereon, which when executed by a processor performs:

The express delivery time prediction method of claim 20.