Consignor label generation method, and logistics operation method and system based on consignor label

Abstract

The present invention proposes a method for generating a consignor label, a logistics operation method and a system based on the consignor label, and relates to the technical field of logistics services. First, the process stages of the logistics service are divided, and the logistics service labels of different process stages are defined. The historical order logistics data is segmented by natural language processing, and the logistics needs of the historical orders are analyzed. The logistics service labels are matched to the consignors based on the result of cosine similarity calculation to form a basic label library for the consignors. Based on the basic label library for the consignors, the consignors' orders are managed for logistics operations, and the corresponding logistics operation personnel are notified to mine and analyze the data in the logistics service process. According to the logistics service operation status of the order, the consignor logistics service label of the order is updated, thereby updating the basic label library for the consignor, standardizing the labeling management method for logistics service needs, realizing the refined label definition and generation of the consignors' logistics service needs at different logistics stages, and improving the logistics service experience and quality.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics services, and more specifically, to a method for generating a consignor label, and a logistics operation method and system based on the consignor label. Background Art

[0002] Currently, there are a large number of small and medium-sized enterprise shippers in the field of large-ticket logistics services. With the improvement of the production and operation level of shipper companies, shippers have more and more demands for the services provided by logistics companies. The logistics service demands of these small and medium-sized enterprise shippers are scattered, different companies have different logistics service requirements, and the demand for logistics service process is difficult to manage.

[0003] Generally, the logistics needs of cargo owners are described through text notes when placing an order. During the logistics operation process, the operator manages the needs of the logistics process through the text information of the cargo owner's notes. For example, a label-based logistics service provision method and device is disclosed in the prior art. The method obtains the stages of the target cargo owner and his cargo, including the collection stage or the delivery stage. The cargo owner has label notes in the collection stage. Through each label, the preferred service of each cargo owner can be understood, and targeted services can be provided to the cargo owner to manage the logistics process of the cargo corresponding to the express delivery. However, in fact, the logistics process is a multi-stage operation process, including core stages such as pickup, trunk transportation, and distribution. The logistics needs of different logistics stages cannot be matched through simple label text notes. Personalized text notes can only be interpreted and executed in the logistics process through the personal understanding of the operator, and cannot be verified and improved in the final logistics execution process.

[0004] In summary, in the actual multi-stage logistics operation process, it is difficult for logistics operators to manage the logistics service needs of different shippers, resulting in inconsistent logistics service experience for small and medium-sized enterprise shippers each time, and inability to guarantee customer satisfaction. Summary of the invention

[0005] In order to solve the problem that the current logistics operation management method based on the simple note text label of the consignor cannot adapt to the multi-stage logistics service needs, the present invention proposes a consignor label generation method, a logistics operation method and system based on the consignor label, and standardizes the labeling management method of the consignor's logistics service needs, so as to realize the refined operation management of different logistics stages, match the consignor's logistics service needs, and thus improve the logistics service experience and quality.

[0006] In order to achieve the above technical effects, the technical solution of the present invention is as follows:

[0007] A method for generating a consignor label, comprising:

[0008] S1. Divide the process stages of logistics services, define the logistics service labels for different process stages of logistics services, and form a standard logistics service label library;

[0009] S2. Generate a logistics corpus based on the standard logistics service label library, together with the general corpus to form a text corpus, obtain historical logistics order data, where the historical logistics order data includes shipper information, the logistics requirement text in the shipper's order remarks, and the logistics requirement text in the logistics operator's remarks, and perform preprocessing operations on the historical order logistics data;

[0010] S3. Perform word segmentation operations on both the text corpus and the preprocessed historical order logistics data, and output the word segmentation results;

[0011] S4. Input the standard logistics service label library and the word segmentation results output in S3 into the Word2Vec model, and output the word vector information of the standard logistics service label library and the word vector information of the logistics requirement text in the shipper's order remarks;

[0012] S5. Calculate the cosine similarity between the word vector matrix in the word vector information of the logistics requirement text in the shipper's order remarks and the word vector matrix in the word vector information of the standard logistics service label library. Denote the cosine similarity calculation result as r, set the automatic matching parameter of similarity as w1, and set the manual matching parameter of similarity as w2, where w1 > w2;

[0013] S6. When r >= w1, automatically match logistics service labels for the shipper based on the word vector information of the standard logistics service label library to form a shipper basic label library; when w2 <= r < w1, there are logistics requirements in the logistics requirement text in the shipper's order remarks that are outside the word vector information of the standard logistics service label library, then the operator manually reviews the logistics requirement text in the shipper's order remarks. If the review passes, add the logistics requirement as a newly added logistics service label to the shipper basic label library.

[0014] In this technical solution, the logistics operation process is divided into different process stages, the logistics service labels for different process stages of logistics services are defined and split into different logistics stages. Word segmentation operations are performed on the historical order logistics data using natural language processing to analyze the logistics requirements of historical orders. Based on the result of cosine similarity calculation, logistics service labels are matched for the shipper, and finally a shipper basic label library is formed.

[0015] Preferably, the process stages of the logistics service described in step S1 include: a pick-up stage, a trunk transportation stage and a delivery stage; wherein, the core labels of the pick-up stage are: pick-up by tail truck, appointment before pick-up, handle with care, and do not invert; the core labels of the trunk transportation stage are: do not dismantle the pallet, handle with care, do not invert, and do not press; the core labels of the delivery stage are: unloading by forklift, handle with care, appointment before delivery, and signing the return note; there is a sequence between the process stages of the logistics service, and one core label can correspond to different process stages; the formed standard logistics service label library includes: logistics service labels, logistics service process stages, and the correspondence between logistics service labels and logistics service process stages.

[0016] Preferably, when generating a logistics corpus based on a standard logistics service label library, the logistics service labels of different process stages of the logistics service and their corresponding process stages are annotated, and the general corpus is generated based on open source data; the preprocessing operation of the historical order logistics data includes: firstly, data cleaning of the logistics demand text of the cargo owner's order notes and the logistics demand text of the logistics operator's notes, and secondly, regular matching of the logistics demand text of the cargo owner's order notes and the logistics demand text of the logistics operator's notes;

[0017] The word segmentation operation performed on the text corpus and the pre-processed historical order logistics data in step S3 includes: generating a DAG directed acyclic graph through the text corpus, combining word frequency scanning, and outputting the first word segmentation result of the logistics vocabulary with the maximum probability; inputting the pre-processed historical order logistics data into the HMM model, and using the Viterbi algorithm to solve the HMM model, and outputting the second word segmentation result;

[0018] The first word segmentation result is merged with the second word segmentation result, and the duplicate data and stop words after the merger are removed, and the word segmentation result is output, including: owner information, order number, cargo information, cargo packaging information, consignee and word segmentation result.

[0019] Preferably, the word vector information of the standard logistics service label library includes: process stages of logistics services, logistics service labels and word vector matrix; the word vector information of the cargo owner's order notes logistics demand text includes: cargo owner information, order number, cargo information, cargo packaging information, word segmentation results and word vector matrix.

[0020] Preferably, in step S6, when matching the logistics service label for the consignor, the logistics service label in the word vector information of the standard logistics service label library is matched according to the dimensions of the consignor, the industry to which the consignor belongs, the shipping area, the consignee, the product category, and the product packaging, and the corresponding relationship between the logistics service label and the logistics service process stage of the consignor is determined to form a basic label library of the consignor; the basic label library of the consignor includes: the logistics service label of the consignor, the logistics service process stage of the consignor, and the corresponding relationship between the logistics service label of the consignor and the logistics service process stage of the consignor;

[0021] When the operation staff manually reviews the logistics demand text in the consignor's order notes, if the logistics demand in the logistics demand text in the consignor's order notes meets the preset specifications, the review is passed.

[0022] The present invention also proposes a logistics operation method based on the consignor tag, comprising:

[0023] S101. The consignor places an order, fills in basic information of the order, and determines the dimensions of the order based on the basic information of the order;

[0024] S102. According to the order dimension, the consignor's logistics service tag is matched to the order from the consignor's basic tag library. The consignor checks and confirms the consignor's logistics service tag and confirms to submit the order;

[0025] S103. According to different process stages in the logistics service, the order is pushed to the operator of the corresponding process stage. The operator checks the logistics service tag of the consignor at this process stage and performs logistics service operations;

[0026] S104. After the process stage of S103 is completed, the order is triggered in the order of the process stages to notify the operator of the next process stage until the logistics service is completed;

[0027] S105. After the logistics service is completed, the consignor's logistics service label of the order is updated according to the logistics service operation status of the order, thereby updating the consignor's basic label library.

[0028] Preferably, the basic information of the order includes: weight, quantity, consignee, select the category of goods, packaging, and quantity; the dimensions of the order include: the owner, the industry to which the owner belongs, the shipping area, the consignee, the category of goods, and the packaging of goods;

[0029] If the consignor is a new consignor, when matching the consignor's logistics service label for the order from the consignor's basic label library according to the order dimensions, the order dimensions at this time include: the consignor's industry, shipping area, cargo category, cargo packaging, and the top N consignor's logistics service labels with the highest frequency of use at different process stages are selected according to the dimensions as the consignor's logistics service labels matched for the new consignor's order;

[0030] When the consignor is viewing the consignor's logistics service label, if there are additional logistics requirements, the consignor can enter the logistics requirement text remarks or select the corresponding logistics service label from the standard logistics service label library to meet the additional logistics requirements.

[0031] Preferably, at different stages of the logistics service process, the operator can also input logistics demand notes text, and the logistics demand notes text input by the operator automatically matches the standard logistics service label and serves as the newly added consignor logistics service label of the order;

[0032] The operation status of the logistics service for an order includes the complaint handling of the logistics service and the claim settlement result of the shipper. In the operation of the logistics service for the shipper's order, if there is a complaint handling and / or a shipper's claim settlement, a new shipper logistics service label is matched for the order; after the logistics service is completed, the shipper logistics service label of the order is updated using the new shipper logistics service label, thereby updating the shipper basic label library.

[0033] Preferably, after step S105, the following steps are further included:

[0034] S106. Set an acquisition cycle, and acquire the logistics requirement note text related to the order after the logistics service is completed according to the acquisition cycle. The logistics requirement note text includes the logistics requirement text note input by the shipper and the logistics requirement text note input by the operator.

[0035] S107. Perform word segmentation on the logistics requirement note text, and then input it into the Word2Vec model to output a word matrix vector.

[0036] S108. Calculate the cosine similarity of the word matrix vector, set the automatic matching parameter of the similarity as w1 and the manual matching parameter of the similarity as w2, where w1 > w2. When w2 <= r < w1, retain the word matrix vector at this time.

[0037] S109. Cluster the retained word matrix vectors to obtain all clustering regions, process all word vectors within the clustering regions, and determine whether the word vectors within the clustering regions need to be submitted for manual review.

[0038] S110. After the word vectors within the clustering regions that need to be submitted for manual review pass the manual review, the logistics requirement note text corresponding to the word vectors is used as a new logistics service label and added to the shipper basic label library to form a new shipper basic label library.

[0039] S111. Cluster the new shipper basic label library to generate refined process stages of the logistics service; when clustering, cluster according to the process stages of the logistics service, use the Kmeans++ algorithm for the clustering method, set the number of clusters as m, and after clustering is completed, calculate for the labels in each clustering region according to the following formula,

[0040]

[0041]

[0042] where j is the order of the jth clustering region, m is the total number of clustering regions, that is, the number of clusters, n is the total number of labels included in all clustering regions, d jIndicates the label of the jth cluster area, nj is the total number of labels of the jth cluster area, ci is the number of times the corresponding i-th label is used; fi is the weight ratio of the corresponding cluster area. The generation of the process stage is automatically judged based on the ratio of the area, which is specifically judged according to the following formula and system parameter configuration:

[0043] If fi>=g, where g is the judgment parameter, the service stage splitting and refinement are prompted. When an area that meets the judgment criteria is generated, it is submitted to the operator for manual review. The operator determines whether the process stage of the logistics service needs to be refined based on the logistics tags gathered in the new area and the changing trend of the logistics service.

[0044] If the operator passes the manual review, a new logistics service subdivision process stage will be created, the corresponding label will be automatically bound to the new process stage, and it will be pushed according to the operation position sequence bound by the operator.

[0045] The present invention also proposes a logistics operation system based on a consignor tag, the system comprising:

[0046] Shipowner client, shipper basic label library module, operation client, update and maintenance client;

[0047] The consignor client includes:

[0048] The order unit supports the shipper to place an order and is used to fill in the basic information of the order and determine the dimensions of the order based on the basic information of the order;

[0049] The label matching unit matches the consignor's logistics service label for the order from the consignor's basic label library according to the order dimensions, so that the consignor can view and confirm the consignor's logistics service label and confirm the order submission;

[0050] The basic label library module for shippers is used to generate the basic label library for shippers;

[0051] The operation client pushes the order to the operation personnel at the corresponding process stage according to the different process stages of the logistics service. The operation personnel check the logistics service tag of the consignor at this process stage and perform logistics service operations. After the process stage is completed, the order is triggered in the order of the process stages to notify the operation personnel at the next process stage until the logistics service is completed.

[0052] Update and maintain the client. After the logistics service is completed, update the consignor's logistics service label of the order according to the logistics service operation status of the order, thereby updating the consignor's basic label library.

[0053] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0054] The present invention proposes a method for generating consignor labels, a logistics operation method and system based on consignor labels. First, the process stages of logistics services are divided, and logistics service labels of different process stages are defined. Natural language processing is used to perform word segmentation operations on historical order logistics data, and the logistics needs of historical orders are analyzed. The logistics service labels are matched to consignors based on the results of cosine similarity calculation to form a consignor basic label library. Then, based on the consignor basic label library, logistics operation management is performed on the consignor's orders. During the logistics operation process, the corresponding logistics operation personnel are notified, and data mining and analysis are performed on the logistics service process. According to the logistics service operation status of the order, the consignor logistics service label of the order is updated, thereby updating the consignor basic label library, standardizing the labeling management method of logistics service needs, realizing the refined label definition and generation of personalized logistics service needs of consignors in different logistics stages, improving the logistics service experience and quality, simplifying the logistics service operation management process, and improving the logistics operation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A schematic diagram showing a flow chart of a method for generating a consignor label proposed in Embodiment 1 of the present invention;

[0056] Figure 2 A schematic diagram showing a flow chart of a logistics operation method based on a consignor tag proposed in Embodiment 2 of the present invention;

[0057] Figure 3 A schematic diagram showing the structure of a logistics operation system based on consignor tags proposed in Example 3 of the present invention. DETAILED DESCRIPTION

[0058] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;

[0059] In order to better illustrate the present embodiment, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the actual size;

[0060] It is understandable to those skilled in the art that descriptions of certain well-known contents in the drawings may be omitted.

[0061] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0062] The positional relationships described in the drawings are only for illustrative purposes and should not be construed as limiting the present patent;

[0063] Example 1

[0064] This embodiment proposes a method for generating a consignor label. For a flowchart of the method, see Figure 1 ,like Figure 1 As shown, including:

[0065] S1. Divide the process stages of logistics services, define logistics service labels for different process stages of logistics services, and form a standard logistics service label library;

[0066] In this embodiment, the process stages of logistics services include: the pickup stage, the trunk transportation stage and the delivery stage; these stages are relatively core stages in logistics services, and in practice, different process stage definitions can be made according to different logistics subdivision fields; taking the division of this embodiment as an example, when defining logistics service labels, the core labels of the pickup stage are: tail truck pickup, appointment before pickup, handle with care, and do not invert, the core labels of the trunk transportation stage are: do not remove the pallet, handle with care, do not invert, and do not press; the core labels of the delivery stage are: forklift unloading, handle with care, appointment before delivery, and sign the return slip; there is a sequence between the process stages of logistics services to ensure the normal operation of the service process, and a core label can correspond to different process stages, such as handle with care, which can correspond to the stages of pickup, trunk transportation, and delivery. The formed standard logistics service label library includes: logistics service labels, logistics service process stages, and the corresponding relationship between logistics service labels and logistics service process stages.

[0067] S2. Generate a logistics corpus based on the standard logistics service tag library, and form a text corpus together with the general corpus to obtain historical logistics order data. The historical logistics order data includes cargo owner information, cargo owner order notes logistics demand text, and logistics operator notes logistics demand text. Perform preprocessing operations on the historical order logistics data;

[0068] In this embodiment, when the logistics corpus is generated based on the standard logistics service label library, the logistics service labels of different process stages of the logistics service and their corresponding process stages are annotated, and the general corpus is generated based on open source data; the preprocessing operation of the historical order logistics data includes: first, data cleaning of the cargo owner's order note logistics demand text and the logistics operator's note logistics demand text, such as special characters, full-width to half-width conversion and other basic processing. Secondly, regular matching is performed on the cargo owner's order note logistics demand text and the logistics operator's note logistics demand text, and operations such as removing duplicate text and extracting keywords are performed, and then step S3 is executed.

[0069] S3. Perform word segmentation operations on the text corpus and the pre-processed historical order logistics data, and output the word segmentation results;

[0070] The word segmentation operation performed on the text corpus and the pre-processed historical order logistics data includes: generating a DAG directed acyclic graph through the text corpus, combining word frequency scanning, and outputting the first word segmentation result of the logistics vocabulary with the maximum probability; inputting the pre-processed historical order logistics data into the HMM model, and using the Viterbi algorithm to solve the HMM model, and outputting the second word segmentation result;

[0071] The first word segmentation result is merged with the second word segmentation result, and the duplicate data and stop words after the merger are removed, and the word segmentation result is output, including: owner information, order number, cargo information, cargo packaging information, consignee and word segmentation result.

[0072] Here, Hidden Markov Model (HMM) is a statistical model that is used to describe a Markov process with hidden unknown parameters, and to determine the hidden parameters of the process from observable parameters. These parameters are then used for further analysis.

[0073] The Viterbi algorithm is a dynamic programming algorithm used to find the most likely sequence of observed events - the Viterbi path - the hidden state sequence, especially in the context of Markov information sources and hidden Markov models. It is used here to facilitate the solution of the HMM model.

[0074] The results of the above two steps of word segmentation make comprehensive use of the word segmentation capabilities of the corpus of the vocabulary and the corpus not logged into the vocabulary, merge the word segmentation results of the two steps, and remove duplicate data.

[0075] In addition, the processed data is used to remove stop words, output the word segmentation information and save it to the database.

[0076] S4. Input the standard logistics service tag library and the word segmentation results output by S3 into the Word2Vec model, and output the word vector information of the standard logistics service tag library and the word vector information of the logistics demand text of the consignor's order notes;

[0077] Word2vec is a process that uses a one-layer neural network (CBOW) to map a one-hot sparse word vector to an n-dimensional (n is generally several hundred) dense vector. In natural language processing, it can be used for word embedding. According to the actual short notes in the logistics industry, generally within 15 words, the text size is set to 15 and the feature vector dimension is 250. Input the standard logistics service tag library and the word segmentation results output by S3 into the Word2Vec model, output the word vector matrix and save it to the database. The word vector information of the standard logistics service tag library includes: the process stage of logistics service, logistics service tags and word vector matrix; the word vector information of the cargo owner's order note logistics demand text includes: cargo owner information, order number, cargo information, cargo packaging information, word segmentation results and word vector matrix.

[0078] S5. Calculate the cosine similarity between the word vector matrix in the shipper's order remarks logistics demand text word vector information and the word vector matrix in the standard logistics service label library word vector information. Denote the cosine similarity calculation result as r, set the similarity automatic matching parameter as w1, and set the similarity manual matching parameter as w2, where w1 > w2;

[0079] In the cosine similarity calculation, for the cosine similarity of two matrices, that is: assume a and b are two n-dimensional vectors, a is the word vector matrix in the shipper's order remarks logistics demand text word vector information, and b is the word vector matrix in the standard logistics service label library word vector information, then the cosine of the angle between a and b is equal to:

[0080]

[0081] The closer the cosine value is to 1, the closer the angle is to 0 degrees, that is, the more similar the two vectors are. When the angle is equal to 0, it means the two vectors exactly match;

[0082] S6. When r >= w1, automatically match logistics service labels for the shipper based on the word vector information in the standard logistics service label library to form the shipper's basic label library; when w2 <= r < w1, there are logistics demands in the shipper's order remarks logistics demand text that are outside the word vector information in the standard logistics service label library. Then, the operator manually reviews the shipper's order remarks logistics demand text. If the review passes, the logistics demand is added to the shipper's basic label library as a newly added logistics service label.

[0083] When matching logistics service labels for the shipper, match the logistics service labels in the word vector information of the standard logistics service label library according to the dimensions of the shipper, the shipper's industry, the shipping area, the consignee, the goods category, and the goods packaging, determine the corresponding relationship between the logistics service label and the shipper's logistics service process stage, and form the shipper's basic label library; the shipper's basic label library includes: the shipper's logistics service label, the shipper's logistics service process stage, and the corresponding relationship between the shipper's logistics service label and the shipper's logistics service process stage;

[0084] When the operator manually reviews the shipper's order remarks logistics demand text, if the logistics demand in the shipper's order remarks logistics demand text meets the preset norms, that is, it is reasonable and legitimate, and the logistics service can meet it and is not an unreasonable demand, then the review passes. In addition, in order to reasonably control the cost of manual review, the manual review is actually carried out periodically, and the values of w1 and w2 need to be configured according to the actual operation situation.

[0085] The above process as a whole divides the logistics operation process into different process stages, defines the logistics service labels of different process stages of logistics services, and splits them into different logistics stages. It uses natural language processing to perform word segmentation operations on historical order logistics data, parses the logistics needs of historical orders, and matches logistics service labels for shippers based on the results of cosine similarity calculation, ultimately forming a basic label library for shippers.

[0086] Example 2

[0087] See also Figure 2 Based on the basic tag library of the consignor in Example 1, the present invention also proposes a logistics operation method based on the consignor tag. The flowchart of the method is as follows: Figure 2 As shown, see Figure 2 , the steps include:

[0088] S101. The consignor places an order, fills in the basic information of the order, and determines the dimensions of the order based on the basic information of the order; the basic information of the order includes: weight, quantity, consignee, select the category of goods, packaging, and quantity; the dimensions of the order include: consignor, industry to which the consignor belongs, shipping area, consignee, category of goods, and packaging of goods;

[0089] If the consignor is a new consignor, when matching the consignor's logistics service label for the order from the consignor's basic label library according to the dimensions of the order, the dimensions of the order at this time include: the industry to which the consignor belongs, the shipping area, the category of goods, and the packaging of goods. The top N consignor logistics service labels with the highest frequency of use at different process stages are taken according to the dimensions as the consignor logistics service labels matched for the new consignor's order, which is set to 3 in this embodiment.

[0090] S102. According to the order dimension, the consignor's logistics service tag is matched to the order from the consignor's basic tag library. The consignor checks and confirms the consignor's logistics service tag and confirms to submit the order;

[0091] When the consignor is viewing the consignor's logistics service label, if there are additional logistics requirements, the consignor can enter the logistics requirement text notes or select the corresponding logistics service label from the standard logistics service label library to meet the additional logistics requirements. In addition, consistent with ordinary logistics services, the operation of the logistics service label is still kept simple, and the consignor client does not display different service stages.

[0092] S103. According to different process stages in the logistics service, the order is pushed to the operator of the corresponding process stage. The operator checks the logistics service tag of the consignor at this process stage and performs logistics service operations;

[0093] In the actual process, after the order is automatically reviewed and approved, it will be automatically pushed to the operator at the corresponding process stage according to the process stage defined by the logistics service. The operator will check the logistics service label of this stage and perform logistics service operations in accordance with the logistics service standards and the requirements of the shipper.

[0094] S104. After the process stage of step S103 is completed, the order is triggered in the order of the process stages to notify the operator of the next process stage until the logistics service is completed;

[0095] At different stages of the logistics service process, operators can also enter logistics demand notes text. The logistics demand notes text entered by operators will automatically match the standard logistics service label and serve as the new consignor logistics service label for the order.

[0096] S105. After the logistics service is completed, the consignor's logistics service label of the order is updated according to the logistics service operation status of the order, thereby updating the consignor's basic label library.

[0097] During the logistics service process, the result data of the service process will be traced back. The logistics service operation status of the order includes the complaint handling of the logistics service and the claim settlement results of the consignor. Among them, the claim settlement results of the consignor are set according to the claim type. Different claim types correspond to different labels. Complaints are also set according to the complaint type and complaint handling result type. Rules are set. In the logistics service operation of the consignor's order, if complaint handling and / or consignor claims occur, a new consignor logistics service label will be added to match the order; after the logistics service is completed, the new consignor logistics service label will be used to update the consignor logistics service label of the order, thereby updating the consignor basic label library.

[0098] In this embodiment, as the logistics order is generated, the new logistics service label and the user service requirements are continuously updated, and the new logistics service label and the logistics service process stage need to be continuously updated according to the specific situation of the business. Therefore, after step S105, the following steps are also included:

[0099] S106. Set an acquisition cycle, and acquire the logistics demand notes text related to the order after the logistics service is completed according to the acquisition cycle, the logistics demand notes text includes the logistics demand text notes entered by the cargo owner and the logistics demand text notes entered by the operator;

[0100] S107. Perform a word segmentation operation on the logistics demand note text, and then input it into the Word2Vec model to output a word matrix vector;

[0101] The process of word segmentation and inputting into the Word2Vec model and outputting the word matrix vector is the same as the corresponding process in Example 1

[0102] S108. Calculate the cosine similarity of the word matrix vectors, set the automatic similarity matching parameter as w1 and the manual similarity matching parameter as w2, where w1 > w2. When w2 <= r < w1, retain the word matrix vector at this time;

[0103] The calculation process of the cosine similarity processing here is the same as that in Embodiment 1.

[0104] S109. Cluster the retained word matrix vectors to obtain all the clustering regions, process all the word vectors within the clustering regions, and determine whether the word vectors within the clustering regions need to be submitted for manual review;

[0105] The specific process of clustering is as follows:

[0106] 1) Randomly select 1 word matrix vector, set it as the region center point, mark the corresponding weight as 1, and create region p.

[0107] 2) Traverse all the remaining word matrix vectors, and calculate the cosine similarity with the region center points of the existing regions;

[0108] 3) When the similarity >= w1, add the word matrix vector to the corresponding region and mark the weight as 1;

[0109] When w2 <= similarity < w1, add the word matrix vector to the corresponding region and mark the weight as 0.5;

[0110] When the similarity < w2, set the word matrix vector as the region center point, mark the weight as 1, and create region p + 1;

[0111] 4) Repeat step 2 until all the word matrix vectors are traversed.

[0112] Obtain all the clustering regions, and calculate all the word matrix vectors within the regions according to the following formula:

[0113]

[0114]

[0115] When r > w3, the default value of w3 is 0.3, submit all the vectors in the corresponding region for manual review. The reviewer reviews whether the corresponding information can be added to the shipper's basic label library through the management background.

[0116] S110. After the word vectors within the clustering regions that need to be submitted for manual review pass the manual review, the logistics demand note text corresponding to the word vectors is used as a new logistics service label and added to the shipper's basic label library to form a new shipper's basic label library;

[0117] S111. Cluster the new cargo owner basic label library to generate detailed logistics service process stages.

[0118] Specifically: When clustering, clustering is performed according to the process stage of logistics services. The clustering method uses the Kmeans++ algorithm and the number of clusters is set to m. After clustering is completed, the label of each cluster area is calculated according to the following formula:

[0119]

[0120]

[0121] Among them, j is the order of the jth cluster region, m is the total number of cluster regions, that is, the number of clusters, n is the total number of labels contained in all cluster regions, and d j Indicates the label of the jth cluster area, nj is the total number of labels of the jth cluster area, ci is the number of times the corresponding i-th label is used; fi is the weight ratio of the corresponding cluster area. The generation of the process stage is automatically judged based on the ratio of the area, which is specifically judged according to the following formula and system parameter configuration:

[0122] If fi>=g, g is the judgment parameter, and the default value of g is 50%;

[0123] The system will prompt the user to split the service phases and refine the service phases. When an area that meets the criteria is generated, it will be submitted to the operator for manual review. The operator will determine whether the process phase of the logistics service needs to be refined based on the logistics tags gathered in the new area and the changing trend of the logistics service.

[0124] If the operator passes the manual review, a new logistics service subdivision process stage is created, the corresponding label is automatically bound to the new process stage, and it is pushed according to the operation position sequence bound by the operator;

[0125] In addition, it is necessary to save the audit records of the operators. The audit records should save the relevant information of the corresponding area for the operators to make judgments, including: audit date, area f i Value and the number of labels for the corresponding area.

[0126] Example 3

[0127] See also Figure 3 The present invention also proposes a logistics operation system based on a consignor tag, the system comprising:

[0128] Shipowner client, shipper basic label library module, operation client, update and maintenance client;

[0129] The consignor client includes:

[0130] The order unit supports the shipper to place an order and is used to fill in the basic information of the order and determine the dimensions of the order based on the basic information of the order;

[0131] The label matching unit matches the consignor's logistics service label for the order from the consignor's basic label library according to the order dimensions, so that the consignor can view and confirm the consignor's logistics service label and confirm the order submission;

[0132] The basic label library module for shippers is used to generate the basic label library for shippers;

[0133] The operation client pushes the order to the operation personnel at the corresponding process stage according to the different process stages of the logistics service. The operation personnel check the logistics service tag of the consignor at this process stage and perform logistics service operations. After the process stage is completed, the order is triggered in the order of the process stages to notify the operation personnel at the next process stage until the logistics service is completed.

[0134] Update and maintain the client. After the logistics service is completed, update the consignor's logistics service label of the order according to the logistics service operation status of the order, thereby updating the consignor's basic label library.

[0135] Obviously, the above embodiments of the present invention are only examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A method for generating a consignor label, characterized in that: Including: S1. Divide the process stages of logistics services, define the logistics service labels for different process stages of logistics services, and form a standard logistics service label library. S2. Generate a logistics corpus based on the standard logistics service label library, which together with the general corpus constitutes a text corpus. Obtain historical logistics order data, which includes shipper information, the logistics demand text in the shipper's order remarks, and the logistics demand text in the remarks of logistics operators, and perform preprocessing operations on the historical order logistics data. S3. Perform word segmentation operations on both the text corpus and the preprocessed historical order logistics data, and output the word segmentation results. S4. Input the standard logistics service label library and the word segmentation results output in S3 into the Word2Vec model, and output the word vector information of the standard logistics service label library and the word vector information of the logistics demand text in the shipper's order remarks. S5. Calculate the cosine similarity between the word vector matrix in the word vector information of the logistics demand text in the shipper's order remarks and the word vector matrix in the word vector information of the standard logistics service label library. The cosine similarity calculation result is denoted as r. Set the automatic matching parameter of similarity as w1 and the manual matching parameter of similarity as w2, where w1 > w2. S6. When r >= w1, automatically match logistics service labels for the shipper based on the word vector information of the standard logistics service label library to form a shipper's basic label library. When w2 <= r < w1, there are logistics demands in the logistics demand text in the shipper's order remarks that are outside the word vector information of the standard logistics service label library. Then, the operator manually reviews the logistics demand text in the shipper's order remarks. If the review passes, the logistics demand is added as a new logistics service label to the shipper's basic label library.

2. The method for generating a consignor label according to claim 1, characterized in that: The process stages of the logistics services described in step S1 include: the pick-up stage, the trunk transportation stage, and the delivery stage. Among them, the core labels for the pick-up stage are: tailgate truck pick-up, pre-pick-up appointment, handle with care, do not invert. The core labels for the trunk transportation stage are: do not disassemble the pallet, handle with care, do not invert, do not press. The core labels for the delivery stage are: forklift unloading, handle with care, pre-delivery appointment, sign the receipt. There is a sequential order among the process stages of the logistics services, and one core label can correspond to different process stages. The formed standard logistics service label library includes: logistics service labels, logistics service process stages, and the corresponding relationship between logistics service labels and logistics service process stages.

3. The method for generating a consignor label according to claim 2, characterized in that: When generating a logistics corpus based on the standard logistics service label library, label the logistics service labels for different process stages of logistics services and their corresponding process stages. The general corpus is generated based on open-source data. The preprocessing operations performed on the historical order logistics data include: first, perform data cleaning on the logistics demand text in the shipper's order remarks and the logistics demand text in the remarks of logistics operators. Second, perform regular matching on the logistics demand text in the shipper's order remarks and the logistics demand text in the remarks of logistics operators. The word segmentation operation performed on the text corpus and the pre-processed historical order logistics data in step S3 includes: generating a DAG directed acyclic graph through the text corpus, combining word frequency scanning, and outputting the first word segmentation result of the logistics vocabulary with the maximum probability; inputting the pre-processed historical order logistics data into the HMM model, and using the Viterbi algorithm to solve the HMM model, and outputting the second word segmentation result; The first word segmentation result is merged with the second word segmentation result, and the duplicate data and stop words after the merger are removed, and the word segmentation result is output, including: owner information, order number, cargo information, cargo packaging information, consignee and word segmentation result.

4. The method for generating a consignor label according to claim 3, characterized in that: The word vector information of the standard logistics service label library includes: the process stages of logistics services, logistics service labels and word vector matrix; the word vector information of the cargo owner's order note logistics demand text includes: cargo owner information, order number, cargo information, cargo packaging information, word segmentation results and word vector matrix.

5. The method for generating a consignor label according to claim 4, characterized in that: In step S6, when matching the logistics service label for the consignor, the logistics service label in the word vector information of the standard logistics service label library is matched according to the dimensions of the consignor, the industry to which the consignor belongs, the shipping area, the consignee, the product category, and the product packaging, and the corresponding relationship between the logistics service label and the logistics service process stage of the consignor is determined to form the basic label library of the consignor; The basic label library of the consignor includes: consignor logistics service labels, consignor logistics service process stages, and the corresponding relationship between consignor logistics service labels and consignor logistics service process stages; When the operation staff manually reviews the logistics demand text in the consignor's order notes, if the logistics demand in the logistics demand text in the consignor's order notes meets the preset specifications, the review is passed.

6. A logistics operation method based on consignor tag, characterized in that: include: S101. The consignor places an order, fills in basic information of the order, and determines the dimensions of the order based on the basic information of the order; S102. According to the order dimension, the consignor's logistics service tag is matched to the order from the consignor's basic tag library. The consignor checks and confirms the consignor's logistics service tag and confirms to submit the order; S103. According to different process stages in the logistics service, the order is pushed to the operator of the corresponding process stage. The operator checks the logistics service tag of the consignor at this process stage and performs logistics service operations; S104. After the process stage of step S103 is completed, the order is triggered in the order of the process stages to notify the operator of the next process stage until the logistics service is completed; S105. After the logistics service is completed, the logistics service label of the consignor of the order is updated according to the logistics service operation of the order, thereby updating the basic label library of the consignor; After step S105, the following steps are also included: S106. Set an acquisition cycle, and acquire the logistics demand notes text related to the order after the logistics service is completed according to the acquisition cycle, the logistics demand notes text includes the logistics demand text notes entered by the cargo owner and the logistics demand text notes entered by the operator; S107. Perform a word segmentation operation on the logistics demand note text, and then input it into the Word2Vec model to output a word matrix vector; S108. Calculate the cosine similarity of the word matrix vectors, set the automatic matching parameter of similarity as w1 and the manual matching parameter of similarity as w2, where w1 > w2. When w2 <= r < w1, retain the word matrix vectors at this time; S109. Cluster the retained word matrix vectors to obtain all clustering regions, process all word vectors within the clustering regions, and determine whether the word vectors within the clustering regions need to be submitted for manual review; S110. After the word vectors within the clustering regions that need to be submitted for manual review pass the manual review, the logistics demand note text corresponding to the word vectors is used as a new logistics service label and added to the shipper's basic label library to form a new shipper's basic label library; S111. Cluster the new shipper's basic label library to generate refined process stages of the logistics service; When clustering, cluster according to the process stages of the logistics service. The clustering method uses the Kmeans++ algorithm, set the number of clusters as m. After clustering, calculate the labels of each clustering region according to the following formula. Among them, j is the order of the jth cluster area, m is the total number of cluster areas, that is, the number of clusters, and n is the total number of labels contained in all cluster areas. represents the label of the jth clustering area, n j is the total number of labels in the jth clustering area, c i is the number of times the i-th tag is used; f i To correspond to the weight proportion of the clustering area, the generation of the process stage is automatically judged based on the proportion of the area, and is specifically judged according to the following formula and system parameter configuration: If f i >=g, where g is the judgment parameter, the service stage is split and refined. When an area that meets the judgment criteria is generated, it is submitted to the operator for manual review. The operator determines whether the process stage of the logistics service needs to be refined based on the logistics tags gathered in the new area and the changing trend of the logistics service; If the operator passes the manual review, a new sub-process stage of the logistics service is created, the corresponding label is automatically bound to the new process stage, and it is pushed according to the operation position sequence bound by the operator.

7. The logistics operation method based on the consignor tag according to claim 6 is characterized in that: The basic information of the order includes: cubic weight, quantity, consignee, select the category of goods, packaging, quantity; the dimensions of the order include: shipper, shipper's industry, shipping area, consignee, category of goods and goods packaging; If the shipper is a new shipper, when matching the shipper's logistics service label for the order from the shipper's basic label library according to the dimensions of the order, at this time, the dimensions of the order include: shipper's industry, shipping area, category of goods, goods packaging. Take the top N shipper's logistics service labels with the most frequent usage frequencies in different process stages according to the dimensions as the shipper's logistics service labels matched for the new shipper's order; When the shipper views the shipper's logistics service label, if there is an additional logistics demand, enter the logistics demand text note or select the corresponding logistics service label from the standard logistics service label library to meet the additional logistics demand.

8. The logistics operation method based on the consignor tag according to claim 7 is characterized in that: In different process stages of the logistics service, the operator can also enter the logistics demand note text. The logistics demand note text entered by the operator is automatically matched with the standard logistics service label and used as the new shipper's logistics service label for the order; The logistics service operation situation of the order includes the complaint handling of the logistics service and the shipper's claim settlement result. In the logistics service operation of the shipper's order, if there is a complaint handling and / or shipper's claim settlement, match the new shipper's logistics service label for the order; After the logistics service is completed, update the shipper's logistics service label of the order using the new shipper's logistics service label, thereby updating the shipper's basic label library.

9. A logistics operation system based on consignor tags, characterized in that: The system includes: Shipper client, shipper's basic label library module, operator client, update and maintenance client; The shipper client includes: Order unit, which supports the shipper to place an order, is used to fill in the basic information of the order, and determine the dimensions of the order based on the basic information of the order; Label matching unit, which matches the shipper's logistics service label for the order from the shipper's basic label library according to the dimensions of the order, provides the shipper to view and confirm the shipper's logistics service label, and confirm and submit the order; The shipper basic label library module is used to generate the shipper basic label library; The operation client pushes the order to the operation personnel at the corresponding process stage according to different process stages in the logistics service. The operation personnel view the shipper logistics service labels at this process stage and carry out the logistics service operation; After the process stage is completed, the order is triggered to notify the operation personnel at the next process stage in the order of the process stage until the logistics service is completed; The update and maintenance client updates the shipper logistics service labels of the order according to the logistics service operation situation of the order after the logistics service is completed, so as to update the shipper basic label library; Set the acquisition period, and acquire the logistics demand note text involved in the order after the logistics service is completed according to the acquisition period. The logistics demand note text includes the logistics demand text note input by the shipper and the logistics demand text note input by the operation personnel; Perform word segmentation on the logistics demand note text, and then input it into the Word2Vec model to output the word matrix vector; Perform cosine similarity calculation on the word matrix vector, set the similarity automatic matching parameter as w1 and the similarity manual matching parameter as w2, w1>w2. When w2<=r<w1, retain the word matrix vector at this time; Cluster the retained word matrix vectors to obtain all the clustering regions, process all the word vectors in the clustering regions, and determine whether the word vectors in the clustering regions need to be submitted for manual review; After the word vectors in the clustering regions that need to be submitted for manual review pass the manual review, the logistics demand note text corresponding to the word vectors is used as a new logistics service label and added to the shipper basic label library to form a new shipper basic label library; Cluster the new shipper basic label library to generate the process stages of the refined logistics service; When clustering, cluster according to the process stages of the logistics service. The clustering method uses the Kmeans++ algorithm, set the number of clusters as m. After clustering, calculate the labels of each clustering region according to the following formula, Among them, j is the order of the jth cluster area, m is the total number of cluster areas, that is, the number of clusters, and n is the total number of labels contained in all cluster areas. represents the label of the jth clustering area, n j is the total number of labels in the jth clustering area, c i is the number of times the i-th tag is used; f i To correspond to the weight proportion of the clustering area, the generation of the process stage is automatically judged based on the proportion of the area, and is specifically judged according to the following formula and system parameter configuration: If f i >=g, where g is the judgment parameter, the service stage is split and refined. When an area that meets the judgment criteria is generated, it is submitted to the operator for manual review. The operator determines whether the process stage of the logistics service needs to be refined based on the logistics tags gathered in the new area and the changing trend of the logistics service; If the operation personnel pass the manual review, a new logistics service sub-process stage is created, the corresponding label is automatically bound to the new process stage, and it is pushed according to the operation position sequence bound by the operation personnel.

Images