Information determination method and device and computer readable storage medium
By obtaining and processing the characteristic information of the company's target products and entrusted items, using feature extraction and similarity calculation technology, the company's main products are automatically determined, and the problems of complex process and low accuracy in the existing technology are solved, and the process is low More efficient and accurate determination of main products.
Patent Information
- Application Number
- CN202311013124.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-11
- Publication Date
- 2025-02-18
AI Technical Summary
In the prior art, there is a problem that the process of determining the authenticity of the main products of the enterprise is complicated and the accuracy is low.
By obtaining the characteristic information of the company's target products and entrusted items, using feature extraction and dimensionality reduction technology, combined with part-of-speech annotation and similarity calculation, we automatically determine whether the target product is the company's main product.
It reduces the complexity of determining the main product, improves efficiency and accuracy, and avoids the dependence of manual questionnaires.
Smart Images
Figure CN119477401A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to information determination technology in the field of logistics, and in particular to an information determination method, device and computer-readable storage medium. Background Art
[0002] Main products refer to the main products in daily life that an enterprise engages in to achieve its business objectives. Verifying the authenticity of the main products published in the enterprise information sheet is of great significance to the study of the development of the enterprise. At present, the authenticity of the main products published in the enterprise information sheet is usually determined manually through questionnaires. However, the above scheme for determining the authenticity of the main products has the problem of complex process and low accuracy. Summary of the invention
[0003] In order to solve the above technical problems, the embodiments of the present application hope to provide an information determination method, device and computer-readable storage medium, which solves the problem that the solution for determining the authenticity of the main product in the related art has a complex process and low accuracy.
[0004] The technical solution of this application is implemented as follows:
[0005] A method for determining information, the method comprising:
[0006] Acquire first characteristic information of a target product of an enterprise; wherein the first characteristic information represents an attribute of the target product;
[0007] Acquire second characteristic information of the consigned items of the enterprise; wherein the second characteristic information represents the attributes of the consigned items;
[0008] For each target product of the enterprise, based on the first characteristic information of the target product and the second characteristic information of the plurality of consigned items, it is determined whether the target product is a main product of the enterprise.
[0009] In the above solution, the step of obtaining the second characteristic information of the consigned items of the enterprise includes:
[0010] Obtaining the consigned items information of the enterprise;
[0011] Determining the part of speech of the words in the consignment information;
[0012] determining target consigned item information from the consigned item information based on the part of speech;
[0013] The target consigned item information is encoded and dimensionally reduced to obtain the second feature information.
[0014] In the above solution, the step of determining target consigned item information from the consigned item information based on the part of speech includes:
[0015] Based on the part of speech, filtering the words in the consigned item information that meet the preset part of speech to obtain the consigned item information to be selected;
[0016] Based on the number of the to-be-selected consigned article information, the to-be-selected consigned article information that meets a preset number is determined from the to-be-selected consigned article information to obtain the target consigned article information.
[0017] In the above solution, the determining the part of speech of the word in the consigned item information includes:
[0018] Encode each character in the consigned item information to obtain encoding information corresponding to each character;
[0019] Based on the coding information, the words in the consigned item information are tagged with parts of speech to obtain the parts of speech of the words.
[0020] In the above solution, the step of determining whether the target product is the main product of the enterprise based on the first characteristic information of the target product and the second characteristic information of the plurality of consigned items includes:
[0021] Determining a similarity between the first feature information and each of the second feature information;
[0022] Based on the sizes of multiple similarities, determine whether the target product is a main product.
[0023] In the above solution, determining whether the target product is the main product based on the magnitudes of multiple similarities includes:
[0024] Based on the magnitudes of the multiple similarities, determining a target similarity satisfying a similarity condition from the multiple similarities;
[0025] When the target similarity is greater than a similarity threshold, determining that the target product is the main product;
[0026] When the target similarity is less than or equal to the similarity threshold, the target product is determined to be a non-main product.
[0027] In the above solution, the step of obtaining the first characteristic information of the target product of the enterprise includes:
[0028] Acquire product information of the target product of the enterprise; wherein the product information at least includes the product name;
[0029] The product information is encoded and dimensionally reduced to obtain the first feature information.
[0030] In the above solution, the step of obtaining the consigned items information of the enterprise includes:
[0031] Obtain the waybill data of the enterprise within the target time period;
[0032] The consigned item information is extracted from the waybill data.
[0033] An information determination device, the device comprising: a processor, a memory, and a communication bus;
[0034] The communication bus is used to realize the communication connection between the processor and the memory;
[0035] The processor is used to execute the information determination program stored in the memory to implement the following steps:
[0036] Acquire first characteristic information of a target product of an enterprise; wherein the first characteristic information represents an attribute of the target product;
[0037] Acquire second characteristic information of the consigned items of the enterprise; wherein the second characteristic information represents the attributes of the consigned items;
[0038] For each target product of the enterprise, based on the first characteristic information of the target product and the second characteristic information of the plurality of consigned items, it is determined whether the target product is a main product of the enterprise.
[0039] A computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the above-mentioned information determination method.
[0040] The information determination method, device and computer-readable storage medium provided by the embodiments of the present application obtain first characteristic information of a target product of an enterprise, wherein the first characteristic information represents the attributes of the target product; obtain second characteristic information of a consigned item of the enterprise, wherein the second characteristic information represents the attributes of the consigned item; for each target product of the enterprise, based on the first characteristic information of the target product and the second characteristic information of multiple consigned items, determine whether the target product is the main product of the enterprise; in this way, the mailing behavior of the enterprise is taken into consideration, and based on the obtained first characteristic information representing the attributes of the target product of the enterprise and the second characteristic information representing the attributes of multiple consigned items of the enterprise, it is automatically determined whether the target product is the main product of the enterprise, thereby reducing the complexity of determining whether the target product is the main product of the enterprise, improving the efficiency and accuracy of determining whether the target product is the main product of the enterprise, and eliminating the need to manually determine whether the target product is the main product of the enterprise through questionnaires, thereby solving the problem that the solution for determining whether the target product is the main product in the related art is complicated and has low accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 A flowchart of an information determination method provided in an embodiment of the present application;
[0042] Figure 2 A flowchart of another information determination method provided in an embodiment of the present application;
[0043] Figure 3 A schematic diagram of the structure of a feature extraction model in an information determination method provided in an embodiment of the present application;
[0044] Figure 4 A flowchart of another information determination method provided in an embodiment of the present application;
[0045] Figure 5 A schematic diagram of the structure of a part-of-speech discrimination model in an information determination method provided in an embodiment of the present application;
[0046] Figure 6 A flowchart of an information determination method provided in another embodiment of the present application;
[0047] Figure 7 A schematic diagram of the structure of an information determination device provided in one embodiment of the present application;
[0048] Figure 8 A schematic diagram of the structure of an information determination device provided in one embodiment of the present application. DETAILED DESCRIPTION
[0049] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.
[0050] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0051] The present application provides an information determination method, which can be applied to an information determination device. Figure 1 As shown, the method comprises the following steps:
[0052] Step 101: Obtain first characteristic information of a target product of an enterprise.
[0053] The first characteristic information represents the attributes of the target product.
[0054] In the embodiment of the present application, the first characteristic information may be determined by the information determination device based on the product information of the enterprise; or it may be obtained by other devices and sent to the information determination device; the target product may be part of or all of the products involved in the enterprise information table. The attribute of the target product refers to the inherent characteristics of the target product; the attribute of the target product includes at least one of the attributes of the product name, product material, product specification, product color, etc. of the target product.
[0055] Specifically, the enterprise information table of the enterprise can be obtained, and the product information of the enterprise can be determined according to the enterprise information table, and the feature of the product information of the enterprise can be extracted to obtain the first feature information; wherein the first feature information can be a feature vector representing the attribute of the target product of the enterprise.
[0056] The product information of the enterprise can be the product information corresponding to the main product field in the product field of the enterprise information table; of course, it can also be the product information corresponding to the entire product field in the enterprise information table. The product field includes the main product field and the non-main product field.
[0057] Step 102: Obtain the second characteristic information of the consignment items of the enterprise.
[0058] The second characteristic information represents the attributes of the consigned items.
[0059] In an embodiment of the present application, the second characteristic information may be determined by the information determination device based on the obtained waybill data of the enterprise; of course, it may also be sent to the information determination device after other devices obtain the second characteristic information. Among them, the attributes of the consigned items refer to the inherent characteristics of the product, and the attributes of the consigned items include at least one of the attributes such as the name, material, specification, and color of the consigned items. The second characteristic information may be a feature vector that characterizes the attributes of the consigned items of the enterprise. Among them, the consigned items may be items that the enterprise entrusted the logistics company to send to the consignee at a historical time, that is, historical consigned items; of course, it may also be items that the enterprise entrusted the logistics company to send to the consignee at the current time, that is, current consigned items.
[0060] Specifically, the consigned item information of the enterprise can be obtained, and the features of the consigned item information can be extracted to obtain the second feature information; wherein the consigned item can be all the items that the enterprise has entrusted the logistics company to send to the consignee at a historical time or at a current time. Alternatively, the consigned item information of the enterprise can be filtered to obtain the target consigned item information, and then the features of the target consigned item information can be extracted to obtain the second feature information, wherein the consigned item can be part of the items that the enterprise has entrusted the logistics company to send to the consignee at a historical time or at a current time.
[0061] It should be noted that step 101 and step 102 can be executed simultaneously; step 101 can also be executed before step 102; of course, step 101 can also be executed after step 102; the embodiment of the present application does not limit the execution order between step 101 and step 102.
[0062] Step 103: for each target product of the enterprise, based on the first characteristic information of the target product and the second characteristic information of the plurality of consigned items, determine whether the target product is the main product of the enterprise.
[0063] Each consignment is different.
[0064] In the embodiment of the present application, for each target product of the enterprise, the first characteristic information of the target product and the second characteristic information of each consignment can be calculated to obtain multiple target values, and then based on the target values, it is determined whether the target product is the main product of the enterprise; in this way, considering the mailing behavior of the enterprise during the operation process, based on the first characteristic information of the target product and the second characteristic information of multiple consignments when sending, it is determined whether the target product is the main product of the enterprise, which improves the accuracy and reliability of determining whether the target product is the main product of the enterprise. Among them, each target product is different, and each target product can be understood as each type of target product; each consignment is different, and multiple consignments can also be understood as multiple types of consignments.
[0065] The information determination method provided by the embodiment of the present application obtains the first characteristic information of the target product of the enterprise; wherein the first characteristic information represents the attributes of the target product; obtains the second characteristic information of the consigned items of the enterprise; wherein the second characteristic information represents the attributes of the consigned items; for each target product of the enterprise, based on the first characteristic information of the target product and the second characteristic information of multiple consigned items, determines whether the target product is the main product of the enterprise; in this way, the mailing behavior of the enterprise is taken into consideration, and based on the obtained first characteristic information representing the attributes of the target product of the enterprise and the second characteristic information representing the attributes of multiple consigned items of the enterprise, it is automatically determined whether the target product is the main product of the enterprise, thereby reducing the complexity of determining whether the target product is the main product of the enterprise, improving the efficiency and accuracy of determining whether the target product is the main product of the enterprise, and eliminating the need to manually determine whether the target product is the main product of the enterprise through questionnaires, thereby solving the problem that the solution for determining whether the target product is the main product in the related art is complicated and has low accuracy.
[0066] Based on the above embodiments, the embodiments of the present application provide an information determination method, which is applied to an information determination device, referring to Figure 2 As shown, the method comprises the following steps:
[0067] Step 201: Obtain product information of the enterprise's target product.
[0068] The product information at least includes the product name.
[0069] In the embodiment of the present application, the enterprise information table of the enterprise can be obtained from the database according to the enterprise identification, and the product information corresponding to the main product field can be extracted from the enterprise information table to obtain the product information of the target product. Among them, the enterprise identification is used to uniquely refer to the enterprise.
[0070] In a feasible implementation, the enterprise logo may be the name of the enterprise; the product information of the target product may be the product name of the target product, the appearance information of the target product, and the use information of the target product.
[0071] Step 202: Encode and reduce the dimension of the product information to obtain first feature information.
[0072] In an embodiment of the present application, each character in the product information can be encoded to obtain feature information corresponding to each character, and the feature information corresponding to multiple characters can be subjected to dimensionality reduction processing to obtain first feature information. Feature extraction includes encoding and dimensionality reduction; encoding is used to convert the format of product information. The encoding method can include one of One-hot encoding, word2vec encoding, and embedding encoding. The dimensionality reduction method can include one of Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA).
[0073] In a feasible implementation, a pre-trained feature extraction model may be used to encode and perform dimension reduction processing on the product information of the target product to obtain the first feature information.
[0074] The feature extraction model may include a semantic representation model for knowledge enhancement (Enhanced language Representation with Informative Entities, ERNIE) and a maximum pooling module (Max-Poling), referred to as ERNIE+Max-Poling. The semantic representation model may encode each character in the product information and output a feature vector corresponding to each character. The maximum pooling module is used to reduce the dimension of feature vectors corresponding to multiple characters to obtain a first feature vector. The first feature information includes the first feature vector.
[0075] Step 203: Obtain the consignment information of the enterprise.
[0076] In the embodiment of the present application, the consigned item information may be determined by the information determination device based on the enterprise's waybill data; or other electronic devices may obtain the consigned item information and send it to the information determination device.
[0077] Step 204: Determine the part of speech of the words in the consignment information.
[0078] In the embodiment of the present application, the consigned item information may be segmented to obtain a plurality of words, and the parts of speech of the plurality of words may be analyzed to obtain the part of speech of each word.
[0079] Specifically, a word segmenter may be used to analyze and process the consigned item information to obtain a plurality of words, and then a part-of-speech analysis model may be used to perform part-of-speech tagging on the plurality of words to obtain the part-of-speech of each word.
[0080] In a feasible implementation, the part-of-speech analysis model may be a hidden Markov model.
[0081] Step 205: Determine target consigned item information from the consigned item information based on the part of speech.
[0082] In the embodiment of the present application, the words in the consigned item information may be filtered based on the part of speech to obtain the target consigned item information.
[0083] Specifically, a target word with a preset part of speech can be determined from the words in the consigned item information, and the target words in the consigned item information can be filtered to obtain the target consigned item information; of course, the target words in the consigned item information can also be filtered, and then the consigned item information obtained after filtering can be screened to obtain the target consigned item information. In this way, based on the part of speech, the words in the consigned item information are filtered to obtain the target consigned item information, so that the obtained target consigned item information is more accurate, and the non-standard text in the consigned item information is avoided. The impact on the subsequent determination of whether the target product is the main product is avoided, and the accuracy of determining whether the target product is the main product is improved; moreover, by filtering the words in the consigned item information, the number of second feature information determined is reduced, and the complexity of the subsequent determination of whether the target product is the main product is further reduced. Among them, the preset part of speech can be non-noun, such as verbs and prepositions.
[0084] Step 206: Encode and reduce the dimension of the target consigned item information to obtain second feature information.
[0085] In an embodiment of the present application, each character in the target consigned item information may be encoded to obtain feature information corresponding to each character, and then the feature information corresponding to multiple characters may be subjected to dimensionality reduction processing to obtain second feature information.
[0086] Specifically, a pre-trained feature extraction model may be used to encode and perform dimension reduction processing on the target consigned item information to obtain the second feature information.
[0087] In one possible implementation, Figure 3 As shown, the target consigned item information is "digital product" and the feature extraction model includes ERNIE and Max-poling; ERNIE can encode the characters "number", "code", "product" and "product" in the target consigned item information respectively to obtain the feature vector corresponding to each character, and then the maximum pooling module can be used to reduce the dimension of the feature vectors corresponding to multiple characters to obtain the second feature vector. Among them, the second feature information includes the second feature vector.
[0088] Step 207: Determine the similarity between the first feature information and each piece of second feature information.
[0089] In the embodiment of the present application, the similarity between the first feature information and each second feature information may be determined based on the types of the first feature information and the second feature information.
[0090] Specifically, when both the first feature information and the second feature information are vectors, that is, the first feature information is a first feature vector and the second feature information is a second feature vector, the cosine similarity between the first feature vector and each second feature vector may be calculated, wherein the similarity includes the cosine similarity.
[0091] In a feasible implementation, the cosine similarity can be calculated using formula (1):
[0092]
[0093] in," s ” represents the cosine similarity between the first feature vector of the i-th target product of the enterprise and the second feature vector of the j-th consignment item; Represents the first eigenvector of the i-th target product of the enterprise; The second eigenvector representing the j-th type of consigned items of the enterprise.
[0094] Step 208: Determine whether the target product is a main product based on the magnitudes of the multiple similarities.
[0095] In the embodiment of the present application, the magnitudes of multiple similarities can be screened to obtain a target similarity, and then based on the target similarity and the similarity threshold, it is determined whether the target product is the main product; of course, it is also possible to determine whether the target product is the main product based on the distribution of the magnitudes of multiple similarities. Among them, the target value includes the similarity.
[0096] It should be noted that the embodiments of the present application take into account the mailing behavior of the enterprise during its operations, and calculate the similarity between the first feature information and each second feature information, and then determine whether the target product is the main product based on the size of multiple similarities, thereby improving the accuracy of determining whether the target product is the main product.
[0097] The information determination method provided in the embodiment of the present application takes into account the mailing behavior of the enterprise, and automatically determines whether the target product is the main product of the enterprise based on the first feature information characterizing the attributes of the target product of the enterprise and the second feature information characterizing the attributes of multiple consigned items of the enterprise. This reduces the complexity of determining whether the target product is the main product of the enterprise, improves the efficiency and accuracy of determining whether the target product is the main product of the enterprise, and eliminates the need for manual determination of whether the target product is the main product of the enterprise through questionnaires. This solves the problem that the solution for determining whether the target product is the main product in the related art is complicated and has low accuracy.
[0098] Based on the above embodiments, the embodiments of the present application provide an information determination method, which is applied to an information determination device, referring to Figure 4 As shown, the method comprises the following steps:
[0099] Step 301: Obtain product information of the enterprise's target product.
[0100] Step 302: Encode and reduce the dimension of the product information to obtain first feature information.
[0101] Step 303: Obtain the enterprise's waybill data within the target time period.
[0102] Among them, the waybill data is a written contract between the carrier (logistics company) and the shipper (enterprise) regarding the transportation of goods.
[0103] In an embodiment of the present application, the target time period can be determined based on the user's input operation, or it can be determined based on the current time and a preset duration; in this way, the waybill data is obtained according to the determined target time period to ensure the timeliness of the obtained waybill data, so as to improve the reliability of determining whether the target product is the main product.
[0104] Step 304: extract consignment information from the waybill data.
[0105] In an embodiment of the present application, semantic analysis may be performed on the waybill data to extract consigned item information from the waybill data.
[0106] In one feasible implementation, the consigned item information may be “digital products, graphics card”; in another feasible implementation, the consigned item information may be “digital products, graphics card help me”.
[0107] Step 305: Encode each character in the consigned item information to obtain encoding information corresponding to each character, and perform part-of-speech tagging on the words in the consigned item information based on the encoding information to obtain the part-of-speech of the words.
[0108] In an embodiment of the present application, an encoder can be used to encode each character in the consigned item information to obtain the encoding information corresponding to each character, and then based on the position information and encoding information between multiple characters, each word in the consigned item information is tagged with a part of speech to obtain the part of speech of each word.
[0109] Specifically, the encoder in the part-of-speech discrimination model can be used to encode each character in the consigned item information to obtain a feature vector corresponding to each character, and the part-of-speech tagging model in the part-of-speech discrimination model can be used to process the feature vectors corresponding to multiple characters to tag each word in the consigned item information to obtain the part of speech of each word. The part-of-speech discrimination model includes an encoder and a part-of-speech tagging model. The encoding information corresponding to each character includes a feature vector corresponding to each character.
[0110] In one possible implementation, Figure 5 As shown, the encoder in the part-of-speech discrimination model can be ERNIE, the part-of-speech tagging model in the part-of-speech discrimination model can be a conditional random field (Conditional Random Field, CRF), and the part-of-speech discrimination model is referred to as ERNIE+CRF.
[0111] like Figure 5 As shown in the figure, the consignment information is "digital products, graphics card help me". ERNIE can be used to encode the characters "number", "code", "product", "product", "," "display", "card", "help" and "me" respectively to obtain the feature vector corresponding to each character, and then CRF is used as the decoder to perform part-of-speech tagging to obtain the part of speech corresponding to each word; Figure 5 As shown, the first 4 characters from left to right, namely "digital products", are marked as nouns, the 5th character "," is marked as invalid, the 6th and 7th characters, namely "graphics card", are marked as nouns, and the 7th and 8th characters, namely "help me", are marked as verbs.
[0112] Step 306: Based on the part of speech, filter the words in the consigned item information that meet the preset part of speech to obtain the consigned item information to be selected.
[0113] In the embodiment of the present application, target words with a preset part of speech can be determined from the words in the consigned item information, and the target words in the consigned item information are filtered to obtain the consigned item information to be selected. In this way, the consigned item information can be cleaned.
[0114] In a feasible implementation, the consigned item information is "digital products, graphics card help me", and the preset part of speech may be a non-noun. Then, after filtering the consigned item information based on the part of speech, the selected consigned item information obtained is "digital products, graphics card"; wherein the selected consigned item information only contains nouns.
[0115] Step 307: Based on the number of the to-be-selected consigned item information, determine the to-be-selected consigned item information that meets the preset number from the to-be-selected consigned item information to obtain the target consigned item information.
[0116] The information of the items to be selected for consignment may be the attribute information of the items to be selected for consignment, and the information of the target consignment may be the attribute information of the target consignment.
[0117] In a feasible implementation, the attribute information of the to-be-selected consigned item may be the name of the to-be-selected consigned item; and the target consigned item information may be the name of the target consigned item.
[0118] In the embodiment of the present application, the number of each category of the to-be-selected consigned item information can be counted, and the to-be-selected consigned item information of the category whose number is greater than the preset number from the multiple categories of the to-be-selected consigned item information is determined as the target consigned item information. In this way, by screening the multiple types of to-be-selected consigned item information based on the number of the to-be-selected consigned item information, to filter out the to-be-selected consigned items with a small number of shipments, the validity and reliability of the determined target consigned item information is improved.
[0119] Step 308: Encode and reduce the dimension of the target consigned item information to obtain second feature information.
[0120] Step 309: Determine the similarity between the first feature information and each piece of second feature information.
[0121] Step 310: Based on the sizes of the multiple similarities, determine a target similarity that meets the similarity condition from the multiple similarities.
[0122] In an embodiment of the present application, multiple similarities can be sorted in order from large to small, and the similarities of the top-ranked target numbers can be determined from the sorted similarities to obtain the target similarity; wherein the target number is greater than or equal to 1; that is, the similarity condition is that after sorting multiple similarities in order from large to small, the top N similarities are used as the target similarity; N is an integer greater than 0.
[0123] In a feasible implementation, the number of targets is 1, and the target similarity is the maximum similarity among multiple similarities.
[0124] Step 311: When the target similarity is greater than the similarity threshold, determine that the target product is the main product.
[0125] In the embodiment of the present application, the target similarity may be compared with a similarity threshold, and when the target similarity is greater than the similarity threshold, the target product is determined to be the main product.
[0126] In a feasible implementation, the target product is the product corresponding to the main product field in the enterprise information table, that is, the target product is the main product announced by the enterprise. When the target similarity is greater than the similarity threshold, it can be determined that the main product announced by the enterprise is the main product in the actual operation of the enterprise.
[0127] Step 312: When the target similarity is less than or equal to the similarity threshold, determine that the target product is a non-main product.
[0128] In the embodiment of the present application, the target similarity may be compared with a similarity threshold, and when the target similarity is less than or equal to the similarity threshold, the target product is determined to be a non-main product.
[0129] In a feasible implementation, the target product is the product corresponding to the main product field in the enterprise information table, that is, the target product is the main product announced by the enterprise. When the target similarity is less than or equal to the similarity threshold, it can be determined that the main product announced by the enterprise is not the real main product in the actual operation of the enterprise.
[0130] The following is a detailed explanation of the information determination method provided in the embodiment of the present application in combination with the application scenario.
[0131] like Figure 6As shown, the waybill data and enterprise information tables of multiple enterprises in the target time period can be obtained. For each enterprise, the information corresponding to the main product field is extracted from the enterprise information table of the enterprise to obtain the product information of the target product of the enterprise. Then, according to the enterprise name, the waybill data of multiple enterprises and the enterprise information tables of multiple enterprises are matched to obtain the multiple consignment items information corresponding to each enterprise; the part-of-speech discrimination model (ERNIE+CRF) is used to perform part-of-speech tagging on the words in each consignment item information to obtain the part-of-speech of each word, and based on the part-of-speech, the multiple consignment items information is screened to obtain the target consignment item information corresponding to each enterprise; then, since the product information of the target products of multiple enterprises may overlap, and the target consignment item information of multiple enterprises may overlap, in order to reduce the calculation The product information of target products of multiple enterprises can be aggregated and deduplicated, and the target consigned items information of multiple enterprises can be aggregated and deduplicated, and each product information after aggregation and deduplication can be input into the feature extraction model (ERNIE+Max-Polling) to obtain the first feature information representing the attribute of each target product of each enterprise, and the second feature information representing the attribute of each consigned item of each enterprise, and then for the mth enterprise, for the i-th target product of the mth enterprise, the similarity between the first feature information of the i-th target product and the second feature information of each type of consigned item of the mth enterprise can be calculated, and then the maximum similarity is determined from multiple similarities. When the maximum similarity is greater than the similarity threshold, it is determined that the i-th target product is the main product of the mth enterprise. Wherein, when the first feature information and the second feature information are vectors, the cosine similarity between the first feature information and the second feature information can be calculated to obtain the similarity; wherein, m is a positive integer.
[0132] It should be noted that since ERNIE is slow in acquiring text features, aggregation and deduplication can avoid more repeated calculations to improve verification speed and save resources.
[0133] Among them, the consigned item information in the waybill data is the name of a certain product under standard circumstances, and the part of speech is a noun. However, there are non-standard texts in the actual waybill data, such as the non-standard word "help me" that appears in "["digital products, graphics card help me"]". Such non-standard words are usually not nouns. Therefore, the part of speech of each word in the consigned item information can be determined by the part-of-speech tagging method, and the words in the consigned item information are cleaned (screened) based on the part of speech to obtain the target consigned item information.
[0134] Among them, for the consignment information T = w1, w2…w n, it is necessary to convert T into an array that can be input into ERNIE+CRF, and then use ERNIE in ERNIE+CRF to convert the array into a feature vector corresponding to each character in the consignment information. The feature vector can be expressed as E = (e1, e2…e n ); where e9 is the i-th character w i The obtained 768-dimensional feature vector with contextual semantics is then input into CRF. CRF can perform part-of-speech tagging on each word in the consigned item information based on E, and obtain multiple tagging results corresponding to the consigned item information. The Viterbi dynamic programming algorithm is then used to determine the target tagging result from the multiple tagging results, and the part-of-speech of each word corresponding to the target tagging result is output.
[0135] In addition, when determining the first characteristic information and the second characteristic information, it is necessary to convert the product information of the target product and the target consigned item information into an array that can be input into ERNIE+Max-Polling, and then obtain the feature vector corresponding to each character in the product information through ERNIE, and obtain the maximum eigenvalue of each character in the same dimension through Max-Pooling to obtain the first feature vector; and obtain the feature vector corresponding to each character in the target consigned item information through ERNIE, and obtain the maximum eigenvalue of each character in the same dimension through Max-Pooling to obtain the second feature vector.
[0136] Among them, ERNIE is a model trained in the field of a large amount of unlabeled text, and the entity information in the text is integrated into the training process. Therefore, it can better obtain the feature vector of each character in the product information of the target product and the feature vector of each character in the target consignment information.
[0137] This application uses the text similarity calculation method in natural language processing to achieve a more accurate match between the consigned goods and the main products provided in the enterprise information table, thereby improving the accuracy and reliability of verification. At the same time, through the analysis of waybill data, more data analysis on the business conditions of the enterprise can be provided, which is of great significance for enterprise management and business decision-making. By using the huge logistics waybill data, it can accurately verify whether the main products provided in the enterprise information table are consistent with its actual business conditions, help enterprises to understand their own business conditions in a timely manner, improve management efficiency and decision-making accuracy, and has high practicality and economic benefits.
[0138] It should be noted that, for the description of the same steps and the same contents in this embodiment as those in other embodiments, reference can be made to the description in other embodiments and will not be repeated here.
[0139] The information determination method provided in the embodiments of the present application takes into account the mailing behavior of the enterprise, and automatically determines whether the target product is the main product of the enterprise based on the first feature information characterizing the attributes of the target product of the enterprise and the second feature information characterizing the attributes of multiple consigned items of the enterprise. This reduces the complexity of determining whether the target product is the main product of the enterprise, improves the efficiency and accuracy of determining whether the target product is the main product of the enterprise, and eliminates the need for manual determination of whether the target product is the main product of the enterprise through questionnaires. This solves the problem that the solutions for determining whether the target product is the main product in the related art are complex and have low accuracy.
[0140] Based on the above embodiments, the embodiments of the present application provide an information determination device, which can be applied to Figure 1 , Figure 2 and Figure 4 In the information determination method provided in the corresponding embodiment, refer to Figure 7 As shown, the information determination device 4 includes: a processor 41, a memory 42 and a communication bus 43, wherein:
[0141] The communication bus 43 is used to realize the communication connection between the processor 41 and the memory 42;
[0142] The processor 41 is used to execute the information determination program stored in the memory 42 to implement the following steps:
[0143] Acquire first characteristic information of a target product of the enterprise; wherein the first characteristic information represents an attribute of the target product;
[0144] Acquire second characteristic information of the consigned items of the enterprise; wherein the second characteristic information represents the attributes of the consigned items;
[0145] For each target product of the enterprise, based on the first characteristic information of the target product and the second characteristic information of the plurality of consigned items, it is determined whether the target product is the main product of the enterprise.
[0146] In other embodiments of the present application, the processor 41 is used to execute the information determination program stored in the memory 42 to obtain the second characteristic information of the consigned items of the enterprise, so as to implement the following steps:
[0147] Obtain the consignment information of the enterprise;
[0148] Determine the part of speech of words in the consignment information;
[0149] Based on the part of speech, determining target consignment item information from the consignment item information;
[0150] The target consigned item information is encoded and dimensionally reduced to obtain the second feature information.
[0151] In other embodiments of the present application, the processor 41 is used to execute the information determination program stored in the memory 42 to determine the target consigned item information from the consigned item information based on the part of speech, so as to implement the following steps:
[0152] Based on the part of speech, the words in the consigned item information that meet the preset part of speech are filtered to obtain the consigned item information to be selected;
[0153] Based on the quantity of the to-be-selected consigned item information, the to-be-selected consigned item information that meets a preset quantity is determined from the to-be-selected consigned item information to obtain the target consigned item information.
[0154] In other embodiments of the present application, the processor 41 is used to execute the information determination program stored in the memory 42 to determine the part of speech of the word in the consigned item information, so as to implement the following steps:
[0155] Encode each character in the consignment information to obtain the encoding information corresponding to each character;
[0156] Based on the coding information, the words in the consignment information are tagged with parts of speech to obtain the parts of speech of the words.
[0157] In other embodiments of the present application, the processor 41 is used to execute the information determination program stored in the memory 42 to determine whether the target product is the main product of the enterprise based on the first feature information of the target product and the second feature information of the plurality of consigned items, so as to implement the following steps:
[0158] Determining a similarity between the first feature information and each second feature information;
[0159] Based on the size of multiple similarities, determine whether the target product is the main product.
[0160] In other embodiments of the present application, the processor 41 is used to execute the information determination program stored in the memory 42 to determine whether the target product is a main product based on the magnitudes of multiple similarities, so as to implement the following steps:
[0161] Based on the magnitudes of the multiple similarities, determining a target similarity satisfying a similarity condition from the multiple similarities;
[0162] When the target similarity is greater than the similarity threshold, the target product is determined to be the main product;
[0163] When the target similarity is less than or equal to the similarity threshold, the target product is determined to be a non-main product.
[0164] In other embodiments of the present application, the processor 41 is used to execute the information determination program stored in the memory 42 to obtain the first feature information of the target product of the enterprise, so as to implement the following steps:
[0165] Obtain product information of the target product of the enterprise; wherein the product information at least includes the product name;
[0166] The product information is encoded and dimensionally reduced to obtain the first feature information.
[0167] In other embodiments of the present application, the processor 41 is used to execute the information determination program stored in the memory 42 to obtain the consigned item information of the enterprise, so as to implement the following steps:
[0168] Obtain the company's waybill data within the target time period;
[0169] Extract consignment information from waybill data.
[0170] It should be noted that the specific implementation process of the steps executed by the processor in this embodiment can refer to Figure 1 , Figure 2 and Figure 4 The implementation process of the information determination method provided in the corresponding embodiment will not be repeated here.
[0171] The information determination device provided in the embodiment of the present application takes into account the mailing behavior of the enterprise, and automatically determines whether the target product is the main product of the enterprise based on the first feature information characterizing the attributes of the target product of the enterprise and the second feature information characterizing the attributes of multiple consigned items of the enterprise. This reduces the complexity of determining whether the target product is the main product of the enterprise, improves the efficiency and accuracy of determining whether the target product is the main product of the enterprise, and eliminates the need for manual determination of whether the target product is the main product of the enterprise through questionnaires. This solves the problem that the solution for determining whether the target product is the main product in the related art is complicated and has low accuracy.
[0172] Based on the above embodiments, the embodiments of the present application provide an information determination device, which can be applied to Figure 1 , Figure 2 and Figure 4 In the information determination method provided in the corresponding embodiment, refer to Figure 8 As shown, the information determination device 5 includes: an acquisition unit 51 and a processing unit 52, wherein:
[0173] The acquisition unit 51 is used to acquire first characteristic information of a target product of an enterprise; wherein the first characteristic information represents an attribute of the target product;
[0174] The acquisition unit 51 is further used to acquire second characteristic information of the consigned items of the enterprise; wherein the second characteristic information represents the attributes of the consigned items;
[0175] The processing unit 52 is used to determine, for each target product of the enterprise, whether the target product is a main product of the enterprise based on the first characteristic information of the target product and the second characteristic information of the plurality of consigned items.
[0176] In other embodiments of the present application, the acquisition unit 51 is further configured to perform the following steps:
[0177] Obtain the consignment information of the enterprise;
[0178] Determine the part of speech of words in the consignment information;
[0179] Based on the part of speech, determining target consignment item information from the consignment item information;
[0180] The target consigned item information is encoded and dimensionally reduced to obtain the second feature information.
[0181] In other embodiments of the present application, the processing unit 52 is further configured to perform the following steps:
[0182] Based on the part of speech, the words in the consigned item information that meet the preset part of speech are filtered to obtain the consigned item information to be selected;
[0183] Based on the quantity of the to-be-selected consigned item information, the to-be-selected consigned item information that meets a preset quantity is determined from the to-be-selected consigned item information to obtain the target consigned item information.
[0184] In other embodiments of the present application, the processing unit 52 is further configured to perform the following steps:
[0185] Encode each character in the consignment information to obtain the encoding information corresponding to each character;
[0186] Based on the coding information, the words in the consignment information are tagged with parts of speech to obtain the parts of speech of the words.
[0187] In other embodiments of the present application, the processing unit 52 is further configured to perform the following steps:
[0188] Determining a similarity between the first feature information and each second feature information;
[0189] Based on the size of multiple similarities, determine whether the target product is the main product.
[0190] In other embodiments of the present application, the processing unit 52 is further configured to perform the following steps:
[0191] Based on the magnitudes of the multiple similarities, determining a target similarity satisfying a similarity condition from the multiple similarities;
[0192] When the target similarity is greater than the similarity threshold, the target product is determined to be the main product;
[0193] When the target similarity is less than or equal to the similarity threshold, the target product is determined to be a non-main product.
[0194] In other embodiments of the present application, the acquisition unit 51 is further configured to perform the following steps:
[0195] Obtain product information of the target product of the enterprise; wherein the product information at least includes the product name;
[0196] The product information is encoded and dimensionally reduced to obtain the first feature information.
[0197] In other embodiments of the present application, the acquisition unit 51 is further configured to perform the following steps:
[0198] Obtain the company's waybill data within the target time period;
[0199] Extract consignment information from waybill data.
[0200] It should be noted that the interaction process between the various units in the embodiment of the present application can refer to Figure 1 , Figure 2 and Figure 4 The implementation process of the information determination method provided in the corresponding embodiment will not be repeated here.
[0201] The information determination device provided in the embodiment of the present application takes into account the mailing behavior of the enterprise, and automatically determines whether the target product is the main product of the enterprise based on the first feature information characterizing the attributes of the target product of the enterprise and the second feature information characterizing the attributes of multiple consigned items of the enterprise. This reduces the complexity of determining whether the target product is the main product of the enterprise, improves the efficiency and accuracy of determining whether the target product is the main product of the enterprise, and eliminates the need for manual determination of whether the target product is the main product of the enterprise through questionnaires. This solves the problem that the solution for determining whether the target product is the main product in the related art is complicated and has low accuracy.
[0202] Based on the foregoing embodiments, the embodiments of the present application provide a computer-readable storage medium, which stores one or more programs, and the one or more programs can be executed by one or more processors to implement Figure 1 , Figure 2 as well as Figure 4 The corresponding embodiments provide information for determining the steps.
[0203] It should be noted that the above-mentioned computer-readable storage medium can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory (Flash Memory), a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM) and other memories; it can also be various electronic devices including one or any combination of the above-mentioned memories, such as mobile phones, computers, tablet devices, personal digital assistants, etc.
[0204] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.
[0205] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.
[0206] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course, by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present application.
[0207] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0208] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0209] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0210] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.
Claims
1. A method for determining information, characterized in that: The method comprises: Acquire first characteristic information of a target product of an enterprise; wherein the first characteristic information represents an attribute of the target product; Acquire second characteristic information of the consigned items of the enterprise; wherein the second characteristic information represents the attributes of the consigned items; For each target product of the enterprise, based on the first characteristic information of the target product and the second characteristic information of the plurality of consigned items, it is determined whether the target product is a main product of the enterprise.
2. The method according to claim 1, characterized in that The obtaining of the second characteristic information of the consigned articles of the enterprise includes: Obtaining the consignment information of the enterprise; Determining the part of speech of the words in the consignment information; determining target consigned item information from the consigned item information based on the part of speech; The target consigned item information is encoded and dimensionally reduced to obtain the second feature information.
3. The method according to claim 2, characterized in that The determining target consigned item information from the consigned item information based on the part of speech includes: Based on the part of speech, filtering the words in the consigned item information that meet the preset part of speech to obtain the consigned item information to be selected; Based on the number of the to-be-selected consigned article information, the to-be-selected consigned article information that meets a preset number is determined from the to-be-selected consigned article information to obtain the target consigned article information.
4. The method according to claim 2, characterized in that: The determining the part of speech of the word in the consigned item information includes: Encode each character in the consigned item information to obtain encoding information corresponding to each character; Based on the coding information, the words in the consigned item information are tagged with parts of speech to obtain the parts of speech of the words.
5. The method according to claim 1, characterized in that The determining, based on the first characteristic information of the target product and the second characteristic information of the plurality of consigned items, whether the target product is the main product of the enterprise includes: Determining a similarity between the first feature information and each second feature information; Based on the sizes of multiple similarities, determine whether the target product is a main product.
6. The method according to claim 5, characterized in that The determining whether the target product is a main product based on the magnitudes of the multiple similarities includes: Based on the magnitudes of the multiple similarities, determining a target similarity satisfying a similarity condition from the multiple similarities; When the target similarity is greater than a similarity threshold, determining that the target product is the main product; When the target similarity is less than or equal to the similarity threshold, the target product is determined to be a non-main product.
7. The method according to claim 1, characterized in that The obtaining of the first characteristic information of the target product of the enterprise includes: Acquire product information of the target product of the enterprise; wherein the product information at least includes the product name; The product information is encoded and dimensionally reduced to obtain the first feature information.
8. The method according to claim 2, characterized in that: The obtaining of the consigned items information of the enterprise includes: Obtain the waybill data of the enterprise within the target time period; The consigned item information is extracted from the waybill data.
9. An information determination device, characterized in that The device comprises: a processor, a memory and a communication bus; The communication bus is used to realize the communication connection between the processor and the memory; The processor is used to execute the information determination program stored in the memory to implement the following steps: Acquire first characteristic information of a target product of an enterprise; wherein the first characteristic information represents an attribute of the target product; Acquire second characteristic information of the consigned items of the enterprise; wherein the second characteristic information represents the attributes of the consigned items; For each target product of the enterprise, based on the first characteristic information of the target product and the second characteristic information of the plurality of consigned items, it is determined whether the target product is a main product of the enterprise.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the information determination method according to any one of claims 1 to 8.