CN111667181A

CN111667181A - Task processing method and device, electronic equipment and computer readable storage medium

Info

Publication number: CN111667181A
Application number: CN202010514622.1A
Authority: CN
Inventors: 罗浩
Original assignee: Rajax Network Technology Co Ltd
Current assignee: Rajax Network Technology Co Ltd
Priority date: 2020-06-08
Filing date: 2020-06-08
Publication date: 2020-09-15
Anticipated expiration: 2040-06-08
Also published as: CN111667181B

Abstract

The embodiment of the disclosure discloses a task processing method, a task processing device, an electronic device and a computer readable storage medium, wherein the task processing method comprises the following steps: acquiring task execution historical behavior data of one or more task receivers in a preset historical time period through at least one processor; according to the task execution historical behavior data, calculating the embedding similarity between the task to be distributed and the distributed task distributed to the target task receiver through at least one processor; in response to the task to be assigned being assigned to the target task recipient, adjusting, by at least one processor, a task execution plan of the target task recipient based on an embedded similarity between the task to be assigned and an assigned task that has been assigned to the target task recipient. The technical scheme can effectively improve the service efficiency under the condition of ensuring the service quality.

Description

Task processing method and device, electronic equipment and computer readable storage medium

Technical Field

The present disclosure relates to the field of data processing technologies, and in particular, to a task processing method and apparatus, an electronic device, and a computer-readable storage medium.

Background

With the development of internet technology, more and more service providers provide services for users or other service demand parties through internet platforms, and many internet services need to be delivered by distributors, so that how to improve delivery efficiency and reduce delivery cost is of great importance to guarantee the service quality of the internet platforms and reduce the service cost of the internet platforms under the condition of guaranteeing the delivery quality. In the prior art, the similarity degree between the order to be distributed and one of the orders which are received by the distributor but are not distributed is generally considered, and the order to be distributed is distributed according to the similarity degree, wherein the adjustment of the task execution route of the distributor after distribution is an important factor for improving the distribution efficiency.

Disclosure of Invention

The embodiment of the disclosure provides a task processing method and device, electronic equipment and a computer-readable storage medium.

In a first aspect, a task processing method is provided in an embodiment of the present disclosure.

Specifically, the task processing method includes:

acquiring task execution historical behavior data of one or more task receivers in a preset historical time period through at least one processor;

according to the task execution historical behavior data, calculating the embedding similarity between the task to be distributed and the distributed task distributed to the target task receiver through at least one processor, wherein the embedding similarity is calculated based on the embedding vectors of the task to be distributed and the distributed task;

in response to the task to be assigned being assigned to the target task recipient, adjusting, by at least one processor, a task execution plan of the target task recipient based on an embedded similarity between the task to be assigned and an assigned task that has been assigned to the target task recipient.

With reference to the first aspect, in a first implementation manner of the first aspect, the calculating, by at least one processor, an embedding similarity between the task to be allocated and the allocated task allocated to the target task recipient according to the task execution historical behavior data includes:

calculating, by at least one processor, a first embedded vector of a task execution location from the task execution historical behavior data;

acquiring a second embedded vector corresponding to an execution place of the task to be distributed through at least one processor;

calculating, by at least one processor, a similarity between the second embedding vector and the first embedding vector, determined as an embedding similarity between the task to be allocated and an allocated task that has been allocated to the target task recipient.

With reference to the first aspect and the first implementation manner of the first aspect, in a second implementation manner of the first aspect, the calculating, by the at least one processor, a first embedded vector of a task execution location according to the task execution historical behavior data includes:

generating a task execution weighted undirected graph according to the task execution historical behavior data through at least one processor, wherein the task execution weighted undirected graph comprises two or more nodes used for representing task execution places and edges connecting the nodes, and the nodes are correspondingly provided with corresponding geographic areas;

and calculating node embedded vectors of all nodes in the weighted undirected graph for task execution through at least one processor, and taking the node embedded vectors as first embedded vectors of corresponding task execution places.

With reference to the first aspect, the first implementation manner of the first aspect, and the second implementation manner of the first aspect, in a third implementation manner of the first aspect, the calculating, by the at least one processor, node-embedded vectors of nodes in the task execution weighted undirected graph is implemented as:

calculating, by at least one processor, a sequence of random walk behaviors of the task execution weighted undirected graph;

determining an embedded vector learning model through at least one processor, learning by taking the random walk behavior sequence as the input of the embedded vector learning model, and obtaining node embedded vectors of all nodes in the task execution weighted undirected graph through learning.

With reference to the first aspect, the first implementation manner of the first aspect, the second implementation manner of the first aspect, and the third implementation manner of the first aspect, in a fourth implementation manner of the first aspect, the obtaining, by the at least one processor, the second embedding vector corresponding to the execution location of the task to be allocated is implemented as:

acquiring, by at least one processor, an execution location of the task to be assigned;

matching, by at least one processor, an execution location of the task to be distributed with a geographic area corresponding to a node in the task execution weighted undirected graph;

determining, by at least one processor, a node embedding vector corresponding to the task execution weighted undirected graph node matched with the geographic area of the task execution place to be allocated as a second embedding vector corresponding to the execution place of the task to be allocated.

With reference to the first aspect, the first implementation manner of the first aspect, the second implementation manner of the first aspect, the third implementation manner of the first aspect, and the fourth implementation manner of the first aspect, in a fifth implementation manner of the first aspect, before the calculating, by the at least one processor, a similarity between the second embedded vector and the first embedded vector, the method further includes:

and acquiring side information and corresponding weights through at least one processor, and performing weighting adjustment on the first embedded vector and the second embedded vector according to the side information and the weights.

With reference to the first aspect, the first implementation manner of the first aspect, the second implementation manner of the first aspect, the third implementation manner of the first aspect, the fourth implementation manner of the first aspect, and the fifth implementation manner of the first aspect, in a sixth implementation manner of the first aspect, an embodiment of the present disclosure further includes:

and calculating an initial evaluation value of the task to be distributed through at least one processor, and adjusting the initial evaluation value of the task to be distributed according to the embedding similarity.

With reference to the first aspect, the first implementation manner of the first aspect, the second implementation manner of the first aspect, the third implementation manner of the first aspect, the fourth implementation manner of the first aspect, the fifth implementation manner of the first aspect, and the sixth implementation manner of the first aspect, in a seventh implementation manner of the first aspect, the calculating, by the at least one processor, an initial evaluation value of the task to be allocated includes:

determining, by at least one processor, an evaluation value per unit time corresponding to the target task recipient;

acquiring the execution distance of the task to be distributed and the historical execution speed of the target task receiver through at least one processor, and calculating the expected execution time of the task to be distributed according to the execution distance and the historical execution speed;

and multiplying the estimated execution time and the unit time evaluation value through at least one processor to obtain an initial evaluation value of the task to be distributed.

With reference to the first aspect, the first implementation manner of the first aspect, the second implementation manner of the first aspect, the third implementation manner of the first aspect, the fourth implementation manner of the first aspect, the fifth implementation manner of the first aspect, the sixth implementation manner of the first aspect, and the seventh implementation manner of the first aspect, in an eighth implementation manner of the first aspect, the adjusting, according to the embedding similarity, the initial evaluation value of the task to be allocated is implemented as:

and when the embedding similarity is higher than a preset similarity threshold, performing down-regulation processing on the initial evaluation value of the task to be distributed, and enabling the unit time evaluation value corresponding to the evaluation value after the down-regulation processing not to be lower than the unit time evaluation value corresponding to the target task receiver.

In a second aspect, a task processing device is provided in an embodiment of the present disclosure.

Specifically, the task processing device includes:

the acquisition module is configured to acquire task execution historical behavior data of one or more task receivers within a preset historical time period through at least one processor;

a calculation module configured to calculate, by at least one processor, an embedding similarity between the task to be allocated and an allocated task allocated to the target task recipient according to the task execution history behavior data, where the embedding similarity is calculated based on embedding vectors of the task to be allocated and the allocated task;

an adjustment module configured to adjust, by at least one processor, a task execution plan of the target task recipient based on an embedding similarity between the task to be assigned and an assigned task assigned to the target task recipient in response to the task to be assigned being assigned to the target task recipient.

With reference to the second aspect, in a first implementation manner of the second aspect, the computing module includes:

a computation submodule configured to compute, by at least one processor, a first embedded vector of task execution places from the task execution historical behavior data;

the obtaining submodule is configured to obtain, through at least one processor, a second embedded vector corresponding to an execution place of the task to be distributed;

a determining sub-module configured to calculate, by at least one processor, a similarity between the second embedding vector and the first embedding vector, which is determined as an embedding similarity between the task to be allocated and an allocated task that has been allocated to the target task recipient.

With reference to the second aspect and the first implementation manner of the second aspect, in a second implementation manner of the second aspect, the computing submodule is configured to:

acquiring task execution historical behavior data of the one or more task receivers in a preset historical time period through at least one processor;

With reference to the second aspect, the first implementation manner of the second aspect, and the second implementation manner of the second aspect, in a third implementation manner of the second aspect, the calculating, by the at least one processor, a portion of node embedding vectors of each node in the weighted undirected graph of task execution is configured to:

With reference to the second aspect, the first implementation manner of the second aspect, the second implementation manner of the second aspect, and the third implementation manner of the second aspect, in a fourth implementation manner of the second aspect, the obtaining sub-module is configured to:

With reference to the second aspect, the first implementation manner of the second aspect, the second implementation manner of the second aspect, the third implementation manner of the second aspect, and the fourth implementation manner of the second aspect, in a fifth implementation manner of the second aspect, the determining sub-module further includes, before:

and the adjusting submodule is configured to acquire side information and corresponding weights through at least one processor, and perform weighting adjustment on the first embedded vector and the second embedded vector according to the side information and the weights.

With reference to the second aspect, the first implementation manner of the second aspect, the second implementation manner of the second aspect, the third implementation manner of the second aspect, the fourth implementation manner of the second aspect, and the fifth implementation manner of the second aspect, in a sixth implementation manner of the second aspect, the embodiment of the present disclosure further includes:

and the processing module is configured to calculate an initial evaluation value of the task to be distributed through at least one processor and carry out adjustment processing on the initial evaluation value of the task to be distributed according to the embedding similarity.

With reference to the second aspect, the first implementation manner of the second aspect, the second implementation manner of the second aspect, the third implementation manner of the second aspect, the fourth implementation manner of the second aspect, the fifth implementation manner of the second aspect, and the sixth implementation manner of the second aspect, in a seventh implementation manner of the second aspect, the part of the processing module that calculates the initial evaluation value of the task to be allocated through the at least one processor is configured to:

With reference to the second aspect, the first implementation manner of the second aspect, the second implementation manner of the second aspect, the third implementation manner of the second aspect, the fourth implementation manner of the second aspect, the fifth implementation manner of the second aspect, the sixth implementation manner of the second aspect, and the seventh implementation manner of the second aspect, in an eighth implementation manner of the second aspect, the part of the processing module that performs adjustment processing on the initial evaluation value of the task to be allocated according to the embedding similarity is configured to:

In a third aspect, the disclosed embodiments provide an electronic device, including a memory and at least one processor, where the memory is configured to store one or more computer instructions, where the one or more computer instructions are executed by the at least one processor to implement the method steps of the task processing method in the first aspect.

In a fourth aspect, the disclosed embodiments provide a computer-readable storage medium for storing computer instructions for a task processing device, which contains computer instructions for executing the task processing method in the first aspect described above as a task processing device.

The technical scheme provided by the embodiment of the disclosure can have the following beneficial effects:

according to the technical scheme, the embedding similarity between the task to be distributed and the distributed task distributed to the target task receiver is calculated based on the embedding vectors of the task to be distributed and the distributed task, and finally the task execution plan of the target task receiver is adjusted according to the embedding similarity. According to the technical scheme, the similarity degree between the task to be distributed and all the distributed tasks of a certain target task receiver is considered, and the task execution plan of the target task receiver is adjusted based on the similarity degree obtained through calculation subsequently, so that the service efficiency is effectively improved under the condition of ensuring the service quality.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

Other features, objects, and advantages of the present disclosure will become more apparent from the following detailed description of non-limiting embodiments when taken in conjunction with the accompanying drawings. In the drawings:

FIG. 1 illustrates a flow diagram of a task processing method according to an embodiment of the present disclosure;

FIG. 2 illustrates a delivery route adjustment schematic according to an embodiment of the present disclosure;

fig. 3 illustrates a block diagram of a task processing device according to an embodiment of the present disclosure;

FIG. 4 shows a block diagram of an electronic device according to an embodiment of the present disclosure;

fig. 5 is a schematic structural diagram of a computer system suitable for implementing a task processing method according to an embodiment of the present disclosure.

Detailed Description

Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement them. Also, for the sake of clarity, parts not relevant to the description of the exemplary embodiments are omitted in the drawings.

In the present disclosure, it is to be understood that terms such as "including" or "having," etc., are intended to indicate the presence of the disclosed features, numbers, steps, behaviors, components, parts, or combinations thereof, and are not intended to preclude the possibility that one or more other features, numbers, steps, behaviors, components, parts, or combinations thereof may be present or added.

It should be further noted that the embodiments and features of the embodiments in the present disclosure may be combined with each other without conflict. The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.

The technical scheme provided by the embodiment of the disclosure calculates the embedding similarity between the task to be distributed and the distributed task distributed to the target task receiver based on the embedding vectors of the task to be distributed and the distributed task, and finally adjusts the task execution plan of the target task receiver according to the embedding similarity. According to the technical scheme, the similarity degree between the task to be distributed and all the distributed tasks of a certain target task receiver is considered, and the task execution plan of the target task receiver is adjusted based on the similarity degree obtained through calculation subsequently, so that the service efficiency is effectively improved under the condition of ensuring the service quality.

Fig. 1 shows a flowchart of a task processing method according to an embodiment of the present disclosure, which is applicable to a task processing server, and as shown in fig. 1, the task processing method includes the following steps S101 to S103:

in step S101, acquiring, by at least one processor, task execution history behavior data of one or more task recipients within a preset history time period;

in step S102, according to the task execution history behavior data, calculating, by at least one processor, an embedding similarity between the task to be allocated and an allocated task that has been allocated to the target task recipient, where the embedding similarity is calculated based on embedding vectors of the task to be allocated and the allocated task;

in step S103, in response to the task to be allocated to the target task recipient, a task execution plan of the target task recipient is adjusted by at least one processor based on an embedding similarity between the task to be allocated and an allocated task already allocated to the target task recipient.

As mentioned above, with the development of internet technology, more and more service providers provide services for users or other service demanders through an internet platform, and many internet services require distributors to distribute, so how to improve distribution efficiency and reduce distribution cost is very important for guaranteeing the service quality of the internet platform and reducing the service cost of the internet platform in the case of guaranteeing the distribution quality. In the prior art, the similarity degree between the order to be distributed and one of the orders which are received by the distributor but are not distributed is generally considered, and the order to be distributed is distributed according to the similarity degree, wherein the adjustment of the task execution route of the distributor after distribution is an important factor for improving the distribution efficiency.

In view of the above drawbacks, in this embodiment, a task processing method is provided, which calculates an embedding similarity between a task to be allocated and an allocated task allocated to a target task recipient based on embedding vectors of the task to be allocated and the allocated task, and finally adjusts a task execution plan of the target task recipient according to the embedding similarity. According to the technical scheme, the similarity degree between the task to be distributed and all the distributed tasks of a certain target task receiver is considered, and the task execution plan of the target task receiver is adjusted based on the similarity degree obtained through calculation subsequently, so that the service efficiency is effectively improved under the condition of ensuring the service quality.

In an optional implementation manner of this embodiment, the task refers to a task object that can be allocated, can be executed after being allocated, and can generate a certain execution result after being executed, for example, the task may be a pick task, a delivery task, an order task, or the like. For convenience of explanation, the embodiments of the present disclosure will be explained and illustrated below with the pick-up task as an example.

In an optional implementation manner of this embodiment, the task receiving party refers to a party that receives a task and performs an operation on the received task. When the task is a goods taking task, the task receiver is a goods taking resource for executing the goods taking task, wherein the goods taking resource refers to a resource which can be used for executing the goods taking task, such as a goods taker, a goods taking device, a goods taking robot and the like, and it should be noted that the goods taking resource may include a goods taking resource which provides a dedicated goods taking service for a certain service provider, or may include a crowd-sourced goods taking resource which provides a goods taking service for a plurality of service providers, flexibly undertakes the goods taking task across the service providers according to the demands of different service providers and can be shared by the plurality of service providers. The target task receiver refers to a task receiver for which a task needs to be allocated currently.

In an optional implementation manner of this embodiment, the task execution historical behavior data refers to behavior data that occurs when a certain task recipient executes a task within a preset historical time period, for example, the task execution historical behavior data may include one or more of the following data: task execution location information, task execution start location information, task execution end location information, total duration of task execution, and the like. In order to ensure the accuracy of the subsequent embedded vector, the historical behavior data of the execution of the continuous tasks in the preset historical time period is determined as the historical behavior data of the subsequent embedded vector calculation.

In an optional implementation manner of this embodiment, the embedding similarity is one of similarities, and is calculated based on the embedding vectors corresponding to the task to be allocated and the allocated task, and is used to evaluate a similarity degree between the task to be allocated and the allocated task allocated to the target task recipient.

In view of the fact that after the task to be allocated is allocated to the target task recipient, the task execution route of the target task recipient may change due to the calculation of the embedding similarity, in this embodiment, after the task to be allocated is allocated to the target task recipient, the task execution plan of the target task recipient needs to be adjusted based on the embedding similarity between the task to be allocated and the allocated task allocated to the target task recipient according to all the task information allocated to the target task recipient. When the task is a picking task, the task execution plan may be, for example, picking route adjustment, estimated picking time adjustment, or the like. Taking the adjustment of the pick-up and delivery routes as an example, as shown in fig. 2, it is assumed that before a new task is not assigned, the number of assigned tasks of the target task receiver is 2, the corresponding pick-up task execution locations are location 1 and location 2, the corresponding delivery task execution locations are user 1 and user 2, respectively, and before a new task is not assigned, the task execution route of the target task receiver is shown by a solid line in fig. 2, that is, from the pick-up location 1 to the pick-up location 2 to the delivery user 1 and finally to the delivery user 2, because of the execution location of the new pick-up task: the embedding similarity between the site 3 and the site 2 is greater than the embedding similarity between the site 3 and the site 1, and the execution site of the new delivery task: the embedding similarity between the user 3 and the user 1 is greater than the embedding similarity between the user 3 and the user 2, and therefore, after a new pick-up and delivery task is assigned to the target task recipient, the task execution route of the target task recipient becomes as shown by a dotted line in fig. 2, that is, the task execution route is from the location 1 to the location 2, then to the location 3 closer to the location 2, then to the user 3 at the user 1, closer to the user 1, and finally to the user 2 farther from the user 1.

In an optional implementation manner of this embodiment, in step S102, that is, the step of calculating, by at least one processor, an embedding similarity between the task to be allocated and the allocated task allocated to the target task recipient according to the task execution historical behavior data includes the following steps:

As mentioned above, the embedding similarity is calculated based on the embedding vectors corresponding to the task to be allocated and the allocated task, and is used for evaluating the similarity between the task to be allocated and the allocated task allocated to the target task receiver. In the embodiment, when calculating the embedding similarity between the task to be distributed and the distributed task distributed to the target task receiver, firstly, calculating a first embedding vector of a task execution place according to the task execution historical behavior data through at least one processor; then, acquiring a second embedded vector corresponding to the execution place of the task to be distributed through at least one processor; and finally, calculating the similarity between the second embedding vector and the first embedding vector through at least one processor, and determining the similarity as the embedding similarity between the task to be distributed and the distributed task distributed to the target task receiver.

In an optional implementation manner of this embodiment, the embedded vector refers to a vector data expression for characterizing a certain object, for example, an object, which is a task execution location, may be provided with an embedded vector. More specifically, the execution location of the task to be allocated may correspond to an embedded vector, that is, a second embedded vector, one or more embedded vectors, that is, one or more first embedded vectors, may be calculated according to the historical behavior data of task execution, and the similarity between the first embedded vector and the second embedded vector may be considered as the similarity between the task to be allocated and the allocated task allocated to the target task recipient, and may be used to represent the degree of similarity between the task to be allocated and the allocated task allocated to the target task recipient.

In an optional implementation manner of this embodiment, the step of calculating, by the at least one processor, a first embedded vector of a task execution location according to the task execution historical behavior data includes the steps of:

In order to obtain a vector data expression capable of accurately characterizing the characteristics of a certain object, in the embodiment, a first embedded vector of all task execution places is calculated according to the task execution historical behavior data through a weighted undirected graph. Specifically, firstly, generating a task execution weighted undirected graph according to the task execution historical behavior data through at least one processor, wherein the task execution weighted undirected graph comprises two or more nodes used for representing task execution places and edges connecting the nodes, and the nodes are correspondingly provided with corresponding geographic areas; and then calculating node embedded vectors of all nodes in the weighted undirected graph for task execution through at least one processor, wherein the node embedded vectors can be used as first embedded vectors of corresponding task execution places.

In an optional implementation manner of this embodiment, when the task is a pickup task, and when a weighted undirected graph for task execution is generated, identification information of a geographic area where a task execution location in the historical behavior data for task execution is located is used as a node in the weighted undirected graph for task execution, where the node is one or more, then an edge is formed between the nodes according to a task execution path of a task receiver, and the number of times of connection between the two nodes is the weight of the edge. The geographic area where the task execution place is located is obtained by dividing the geographic area in advance according to a preset geographic area division rule, and the identification information of the geographic area where the task execution place is located is also obtained according to the preset geographic area identification rule.

In an optional implementation manner of this embodiment, the step of calculating, by the at least one processor, node embedding vectors of each node in the weighted undirected graph of task execution may be implemented as:

In this embodiment, when calculating the first embedded vector of each node in the task execution weighted undirected graph, first calculating, by at least one processor, a sequence of random walk behaviors of the task execution weighted undirected graph; and then determining an embedded vector learning model through at least one processor, learning by taking the random walk behavior sequence as the input of the embedded vector learning model, and obtaining node embedded vectors of all nodes in the task execution weighted undirected graph through learning.

Specifically, based on the task execution weighted undirected graph, a Deepwalk algorithm can be used for learning to obtain a node embedded vector expression of each node in the task execution weighted undirected graph: firstly, for the task execution weighted undirected graph, randomly walking through a random walk algorithm to obtain a task execution behavior sequence, and then using the obtained task execution behavior sequence as training data of a Skip-Gram algorithm, so that node embedding vectors of each node in the task execution weighted undirected graph can be learned, wherein the deep walk algorithm, the random walk algorithm and the Skip-Gram algorithm are common algorithms in the prior art, and a person skilled in the art can use the algorithms to obtain node embedding vector expressions of each node in the task execution weighted undirected graph, and the details of the disclosure are omitted.

In an optional implementation manner of this embodiment, the step of obtaining, by the at least one processor, the second embedded vector corresponding to the execution location of the task to be allocated may be implemented as:

As mentioned above, the nodes in the task execution weighted undirected graph are correspondingly provided with corresponding geographic areas, which are represented by the identification information of the geographic area where the task execution location is located in the task execution historical behavior data, so that when the second embedded vector corresponding to the execution location of the task to be distributed is obtained, the execution location of the task to be distributed can be determined first; then, matching the execution place of the task to be distributed with a geographical area corresponding to a node in the task execution weighted undirected graph; and determining a node embedding vector corresponding to a node matched with the geographical area where the execution place of the task to be distributed is located in the task execution weighted undirected graph as a second embedding vector corresponding to the execution place of the task to be distributed.

After the similarity between the second embedded vector and each first embedded vector is obtained through calculation by the at least one processor, the similarity can be respectively determined as the embedded similarity between the task to be allocated and all allocated tasks allocated to the target task receiver. In an optional implementation manner of this embodiment, before the step of calculating, by the at least one processor, a similarity between the second embedding vector and the first embedding vector, and determining the similarity as an embedding similarity between the task to be allocated and an allocated task that has been allocated to the target task recipient, the method may further include a step of performing weighted adjustment on the first embedding vector and the second embedding vector based on side information, that is:

In order to further improve the accuracy of the embedded vector, in this embodiment, side information and corresponding weights are also obtained, and the first embedded vector and the second embedded vector are subjected to weighting adjustment according to the side information and the weights. Wherein the side information may include one or more of the following information: the identification information of the geographical area where the new task execution place is located, the task execution distance, the task content type, the task content quantity, the time slice sequence number information to which the task execution belongs, and the like. The task execution distance refers to a distance which a task receiver needs to move when executing a task; the task content type refers to a type to which content included in the task belongs, for example, if the task is a task of fetching food, the task content type refers to the task of fetching food; the task content quantity refers to the quantity of contents contained in the task, for example, if the task is a task of fetching food, the task content quantity refers to the quantity of the food needing to be fetched; the time slice sequence number information to which the task execution belongs refers to a sequence number of the time slice to which the task belongs when the task is executed, for example, if 15 minutes is taken as a time slice and a certain task is executed in the 2 nd 15 minutes, the time slice sequence number information to which the task execution belongs is 2.

After the above-mentioned side information is determined, the first embedding vector and the second embedding vector may be weight-adjusted according to the degree of influence of the side information on the embedding vectors. Specifically, the side information is first converted into a vector form, where the conversion manner of the information vector belongs to a common method in the prior art, and is not described in detail in this disclosure; then respectively setting corresponding weights for the side information and the embedded vector, determining a deep learning network and a loss function, and learning to obtain the weight which enables the loss function to be minimum by taking the side information, the embedded vector and the weight as the input of the deep learning network; and finally, carrying out weighting adjustment on the embedded vector by utilizing each side information vector obtained by learning, the weight corresponding to the side information vector and the weight corresponding to the embedded vector. Assuming that the weights corresponding to the side information vectors obtained by learning are respectively as follows: the weight corresponding to the task execution distance vector is 0.4, the weight corresponding to the task content type vector is 0.2, the weight corresponding to the task content quantity vector is 0.2, the weight corresponding to the time slice sequence number information vector to which the task execution belongs is 0.2, and the weight corresponding to the embedded vector is 1, then the adjusted embedded vector can be represented as: the adjusted embedding vector is (embedding vector before adjustment × 1+ task execution distance vector × 0.4+ task content type vector × 0.2+ task content number vector × 0.2+ time slice index information vector to which task execution belongs × 0.2)/(1+0.4+0.2+0.2+ 0.2).

In an optional implementation manner of this embodiment, the method further includes the following steps:

In the prior art, the similarity degree between the order to be distributed and one of the orders which are received by the distributor but are not distributed is generally considered, and the distribution price of the order to be distributed is deducted according to the similarity degree, so that the receiving efficiency of the distributor for the order is improved, and the distribution cost is reduced. However, the processing method does not consider the similarity between the order to be distributed and all orders which are received but not distributed by the distributor, and the deduction of the distribution price of the order to be distributed mainly depends on subjective experience, so that the accuracy of the distribution price adjustment of the order to be distributed is greatly influenced. Therefore, in this embodiment, the initial evaluation value of the task to be allocated is first calculated, and then the adjustment processing is performed on the initial evaluation value of the task to be allocated according to the embedding similarity calculated above.

In an optional implementation manner of the present embodiment, the evaluation value refers to a cost required to execute the task, that is, an execution value of the task. When the task is a goods taking task, the evaluation value of the task is the goods taking price of the goods taking task.

In an optional implementation manner of this embodiment, the initial evaluation value refers to an initial evaluation value to be adjusted of a certain task to be allocated, which is calculated according to a preset rule.

In an optional implementation manner of this embodiment, the step of calculating, by at least one processor, an initial evaluation value of the task to be allocated includes the following steps:

In this embodiment, in order to determine the initial and subsequent evaluation values to be adjusted of the task to be allocated, at least one processor first determines or obtains an evaluation value per unit time corresponding to the target task receiver, where when the task is a pickup task, the target task receiver is a pickup resource, and then the evaluation value per unit time corresponding to the target task receiver refers to a pickup value per unit time of the pickup resource, such as a salary of the pickup resource; then, obtaining an execution distance of the task to be distributed and a historical execution speed of the target task receiver through at least one processor, and calculating to obtain an expected execution time of the task to be distributed according to the execution distance and the historical execution speed, wherein the execution distance of the task to be distributed refers to a distance between a current position of the target task receiver and an execution position of the task to be distributed when the task to be distributed is executed, and the historical execution speed of the target task receiver can be obtained by calculating an average speed of the target task receiver when the target task receiver executes the task within a preset historical time period; and finally, multiplying the estimated execution time by the unit time evaluation value through at least one processor to obtain an initial evaluation value of the task to be distributed, for example, if the estimated execution time of a certain task to be distributed is 0.5 hour, and the unit time evaluation value corresponding to the target task receiver is 20 yuan/hour, the initial evaluation value of the task to be distributed is 0.5 hour × 20 yuan/hour — 10 yuan.

In an optional implementation manner of this embodiment, the step of performing adjustment processing on the initial evaluation value of the task to be allocated according to the embedding similarity may be implemented as:

when the embedding similarity is higher than a preset similarity threshold, performing down-regulation processing on the initial evaluation value of the task to be distributed through at least one processor, and enabling the unit time evaluation value corresponding to the evaluation value after the down-regulation processing not to be lower than the unit time evaluation value corresponding to the target task receiver.

When the embedding similarity is higher than a preset similarity threshold, the task to be allocated is considered to be relatively similar to all allocated tasks allocated to the target task receiver, if the target task receiver receives the task to be allocated, certain task execution time can be saved during execution, in order to improve the task receiving efficiency of the target task receiver and reduce the task execution cost, the initial evaluation value of the task to be allocated can be adjusted downwards according to the similarity between the task to be allocated and all allocated tasks allocated to the target task receiver, but the adjustment treatment needs to meet a certain principle, namely the unit time evaluation value corresponding to the adjusted evaluation value is not lower than the unit time evaluation value corresponding to the target task receiver, namely the task execution pay of the target task receiver cannot be reduced, that is, the down-regulation processing of the initial evaluation value of the task to be allocated needs to satisfy the following formula:

wherein, P₁And P₂An evaluation value, t, representing the assigned task of the target task recipient₁₂Indicating that the target task recipient has completed execution of the assigned task P₁And P₂Required total task execution time, P₃Representing the adjusted evaluation value t of the task to be distributed₁₂₃Representing the target task recipient executionCompleted assigned task P₁、P₂And the total task execution time required by the task to be distributed.

The technical solution of the present disclosure is explained and illustrated below by taking two application scenarios as examples.

Application scenario one

The task is a goods taking task, the target task receiver is a goods taking resource, the unit time evaluation value corresponding to the target task receiver is the unit time goods taking value of the goods taking resource, namely the salary of the goods taking resource, and the task to be allocated is the goods taking task to be allocated. Firstly, calculating an initial price of a goods taking task to be distributed, specifically, firstly, determining an salary corresponding to the goods taking resource, then obtaining an execution distance of the goods taking task to be distributed and a historical execution speed of the goods taking resource, calculating an expected execution time of the goods taking task to be distributed according to the execution distance and the historical execution speed, and multiplying the expected execution time and the salary to obtain the initial price of the goods taking task to be distributed. Then, according to task execution historical behavior data of the goods taking resources, calculating the embedding similarity between the goods taking task to be distributed and the distributed goods taking task distributed to the goods taking resources, specifically, firstly, calculating a first embedding vector of a goods taking task execution place according to the task execution historical behavior data of the goods taking resources; then acquiring a second embedded vector corresponding to the execution place of the goods taking task to be distributed; and finally, calculating the similarity between the second embedded vector and the first embedded vector as the embedded similarity between the goods taking task to be allocated and the allocated goods taking task allocated to the goods taking resources. Finally, the action route of the pickup resource and the initial price of the pickup task to be allocated are adjusted according to the calculated embedding similarity, for example, when the embedding similarity is higher than a preset similarity threshold, the initial price of the pickup task to be allocated is adjusted downward, but it should be noted that the salary corresponding to the price after the downward adjustment process should not be lower than the previously determined salary of the pickup resource.

Application scenario two

The task is a guest sending task, namely, a guest is sent to an appointed terminal point from a starting point where the guest is located, the target task receiver is a guest sending resource, the unit time evaluation value corresponding to the target task receiver is the unit time guest sending value of the guest sending resource, namely the salary of the guest sending resource, and the task to be allocated is the guest sending task to be allocated. Firstly, calculating an initial price of a to-be-allocated guest-sending task, specifically, firstly, determining an salary corresponding to the guest-sending resource, then obtaining an execution distance of the to-be-allocated guest-sending task and a historical execution speed of the guest-sending resource, calculating an expected execution time of the to-be-allocated guest-sending task according to the execution distance and the historical execution speed, and multiplying the expected execution time and the salary to obtain the initial price of the to-be-allocated guest-sending task. Then, according to task execution historical behavior data of the delivery resources, calculating the embedding similarity between the delivery task to be distributed and the distributed delivery task, specifically, firstly, calculating a first embedding vector of a starting/ending execution place of the delivery task according to the task execution historical behavior data of the delivery resources; then acquiring a second embedded vector corresponding to the initial execution place of the to-be-distributed passenger sending task; and finally, calculating the similarity between the second embedded vector and the first embedded vector as the embedded similarity between the to-be-distributed guest sending task and the distributed guest sending task distributed to the guest sending resource. Finally, the action route of the guest-sending resource and the initial price of the guest-sending task to be allocated are adjusted according to the calculated embedding similarity, for example, when the embedding similarity is higher than a preset similarity threshold, the initial price of the guest-sending task to be allocated is adjusted downward, but it should be noted that the salary corresponding to the price after the downward adjustment process should not be lower than the previously determined salary of the guest-sending resource.

The following are embodiments of the disclosed apparatus that may be used to perform embodiments of the disclosed methods.

Fig. 3 shows a block diagram of a task processing device according to an embodiment of the present disclosure, which may be implemented as part or all of an electronic device, and may be implemented as a task processing server, through software, hardware, or a combination of both. As shown in fig. 3, the task processing device includes:

an obtaining module 301 configured to obtain, by at least one processor, task execution history behavior data of one or more task recipients within a preset history time period;

a calculating module 302 configured to calculate, by at least one processor, an embedding similarity between the task to be allocated and an allocated task allocated to the target task recipient according to the task execution history behavior data, where the embedding similarity is calculated based on embedding vectors of the task to be allocated and the allocated task;

an adjustment module 303 configured to adjust, by at least one processor, a task execution plan of the target task recipient based on an embedding similarity between the task to be allocated and an allocated task already allocated to the target task recipient in response to the task to be allocated being allocated to the target task recipient.

In view of the above drawbacks, in this embodiment, a task processing device is provided, which calculates an embedding similarity between a task to be allocated and an allocated task allocated to a target task recipient based on embedding vectors of the task to be allocated and the allocated task, and finally adjusts a task execution plan of the target task recipient according to the embedding similarity. According to the technical scheme, the similarity degree between the task to be distributed and all the distributed tasks of a certain target task receiver is considered, and the task execution plan of the target task receiver is adjusted based on the similarity degree obtained through calculation subsequently, so that the service efficiency is effectively improved under the condition of ensuring the service quality.

In an optional implementation manner of this embodiment, the calculating module 302 includes:

the computing sub-module is configured to acquire task execution historical behavior data of the one or more task receivers in a preset historical time period through at least one processor, and compute a first embedded vector of a task execution place according to the task execution historical behavior data;

As mentioned above, the embedding similarity is calculated based on the embedding vectors corresponding to the task to be allocated and the allocated task, and is used for evaluating the similarity between the task to be allocated and the allocated task allocated to the target task receiver. In this embodiment, when calculating the embedding similarity between the task to be distributed and the distributed task distributed to the target task receiver, the calculating sub-module calculates, by at least one processor, a first embedding vector of a task execution location according to the task execution historical behavior data; the obtaining submodule obtains a second embedded vector corresponding to the execution place of the task to be distributed through at least one processor; the determining submodule calculates the similarity between the second embedding vector and the first embedding vector through at least one processor, and determines the similarity as the embedding similarity between the task to be distributed and the distributed task distributed to the target task receiver.

In an optional implementation manner of this embodiment, the calculation sub-module may be configured to:

In an optional implementation manner of this embodiment, the calculating, by the at least one processor, a portion of the node embedding vector of each node in the weighted undirected graph of task execution may be configured to:

In an optional implementation manner of this embodiment, the obtaining sub-module may be configured to:

In the determining submodule, after the similarity between the second embedding vector and each first embedding vector is obtained through calculation by at least one processor, the similarity can be respectively determined as the embedding similarity between the task to be allocated and all allocated tasks allocated to the target task receiver.

In an optional implementation manner of this embodiment, before the determining the sub-module, a part that performs weighting adjustment on the first embedded vector and the second embedded vector based on side information may further be included, that is, the calculating module 302 includes:

the adjusting submodule is configured to obtain side information and corresponding weights through at least one processor, and carry out weighting adjustment on the first embedded vector and the second embedded vector according to the side information and the weights;

In an optional implementation manner of this embodiment, the apparatus further includes:

In an optional implementation manner of this embodiment, the part of the processing module that calculates, by at least one processor, the initial evaluation value of the task to be assigned may be configured to:

In an optional implementation manner of this embodiment, a portion of the processing module, which performs adjustment processing on the initial evaluation value of the task to be allocated according to the embedding similarity, may be configured to:

wherein, P₁And P₂An evaluation value, t, representing the assigned task of the target task recipient₁₂Indicating that the target task recipient has completed execution of the assigned task P₁And P₂Total task requiredExecution time, P₃Representing the adjusted evaluation value t of the task to be distributed₁₂₃Indicating that the target task recipient has completed execution of the assigned task P₁、P₂And the total task execution time required by the task to be distributed.

The present disclosure also discloses an electronic device, fig. 4 shows a block diagram of an electronic device according to an embodiment of the present disclosure, and as shown in fig. 4, the electronic device 400 includes a memory 401 and a processor 402; wherein the content of the first and second substances,

the memory 401 is used to store one or more computer instructions that are executed by the processor 402 to implement the above-described method steps.

As shown in fig. 5, the computer system 500 includes a processing unit 501 that can execute various processes in the above-described embodiments according to a program stored in a Read Only Memory (ROM)502 or a program loaded from a storage section 508 into a Random Access Memory (RAM) 503. In the RAM503, various programs and data necessary for the operation of the system 500 are also stored. The processing unit 501, the ROM502, and the RAM503 are connected to each other by a bus 504. An input/output (I/O) interface 505 is also connected to bus 504.

The following components are connected to the I/O interface 505: an input portion 506 including a keyboard, a mouse, and the like; an output portion 507 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage portion 508 including a hard disk and the like; and a communication section 509 including a network interface card such as a LAN card, a modem, or the like. The communication section 509 performs communication processing via a network such as the internet. The driver 510 is also connected to the I/O interface 505 as necessary. A removable medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 510 as necessary, so that a computer program read out therefrom is mounted into the storage section 508 as necessary. The processing unit 501 may be implemented as a CPU, a GPU, a TPU, an FPGA, an NPU, or other processing units.

In particular, the above described methods may be implemented as computer software programs, according to embodiments of the present disclosure. For example, embodiments of the present disclosure include a computer program product comprising a computer program tangibly embodied on a medium readable thereby, the computer program comprising program code for performing the object recognition method. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 509, and/or installed from the removable medium 511.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowcharts or block diagrams may represent a module, a program segment, or a portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). 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 blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The units or modules described in the embodiments of the present disclosure may be implemented by software or hardware. The units or modules described may also be provided in a processor, and the names of the units or modules do not in some cases constitute a limitation of the units or modules themselves.

As another aspect, the present disclosure also provides a computer-readable storage medium, which may be the computer-readable storage medium included in the apparatus in the above-described embodiment; or it may be a separate computer readable storage medium not incorporated into the device. The computer readable storage medium stores one or more programs for use by one or more processors in performing the methods described in the present disclosure.

The foregoing description is only exemplary of the preferred embodiments of the disclosure and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the invention in the present disclosure is not limited to the specific combination of the above-mentioned features, but also encompasses other embodiments in which any combination of the above-mentioned features or their equivalents is possible without departing from the inventive concept. For example, the above features and (but not limited to) the features disclosed in this disclosure having similar functions are replaced with each other to form the technical solution.

Claims

1. A method of task processing, comprising:

2. The method of claim 1, the calculating, by at least one processor, an embedding similarity between the task to be assigned and an assigned task assigned to the target task recipient according to the task execution historical behavior data, comprising:

3. The method of claim 2, the calculating, by at least one processor, a first embedded vector of task execution locations from the task execution historical behavior data, comprising:

4. The method of claim 3, the computing, by at least one processor, node-embedded vectors for respective nodes in the task execution weighted undirected graph implemented as:

5. A task processing device comprising:

6. The apparatus of claim 5, the computing module comprising:

7. The apparatus of claim 6, the computation submodule configured to:

8. The apparatus of claim 7, the computing, by at least one processor, the portion of the node-embedded vectors for the respective nodes in the weighted undirected graph of task execution configured to:

9. An electronic device comprising a memory and at least one processor; wherein the memory is to store one or more computer instructions, wherein the one or more computer instructions are to be executed by the at least one processor to implement the method steps of any one of claims 1-4.

10. A computer readable storage medium having stored thereon computer instructions which, when executed by a processor, carry out the method steps of any of claims 1-4.