CN114022086B

CN114022086B - Purchasing method, device, equipment and storage medium based on BOM identification

Info

Publication number: CN114022086B
Application number: CN202210007440.4A
Authority: CN
Inventors: 肖清华; 李六七; 王安; 杜飞; 刘武
Original assignee: Shenzhen Qianhai Yingzhicheng Information Technology Co ltd
Current assignee: Shenzhen Foresea Allchips Information & Technology Co.,Ltd.
Priority date: 2022-01-06
Filing date: 2022-01-06
Publication date: 2022-04-22
Anticipated expiration: 2042-01-06
Also published as: CN114022086A

Abstract

The invention relates to the field of automatic purchasing and discloses a purchasing method, device, equipment and storage medium based on BOM identification. The method comprises the following steps: acquiring a BOM file of an electronic component to be purchased; according to a preset element device category identification algorithm, category label setting processing is carried out on the BOM file to obtain a label BOM file, and the setting label of the label BOM file comprises the following steps: a component model label and a non-component model label; according to a preset standardization algorithm, carrying out standard standardization processing on data of a non-element device type label in a label BOM file to obtain a label BOM standard file; accessing a preset target commodity database, and performing association analysis processing on the tag BOM standard file and the target commodity database according to a preset comparison analysis algorithm to obtain associated commodity data; and screening the associated commodity data according to the commodity inventory, the commodity unit price and the commodity delivery time length in the target commodity database to obtain the purchased commodity data corresponding to the BOM file.

Description

Purchasing method, device, equipment and storage medium based on BOM identification

Technical Field

The invention relates to the field of automatic purchasing, in particular to a purchasing method, device, equipment and storage medium based on BOM identification.

Background

At present, the quotation of a component bill of material (BOM) mainly comprises two modes of manual quotation and system quotation. The first manual quotation process is that after identifying key information such as product type, model, brand, parameter, package, unit consumption, required quantity, position number, client material number and the like of each material in a bill of materials (BOM) file, searching in commodity data by using the model or the parameter respectively, selecting the optimal commodity according to certain service requirements of the search result, filling the key information such as the model, the brand, the parameter, package, quotation quantity and the like of the commodity in a corresponding quotation table, and finally sending the processed quotation table to a client. The problem of traditional manual quotation: 1. single materials are required to be processed one by one, so that the quotation efficiency is low; 2. the price inquiring file and the quotation file need to be transmitted through communication tools such as WeChat, QQ, mails and the like, and the communication efficiency is low.

Aiming at the problems encountered by the traditional manual quotation, a second system quotation is adopted, and the general scheme is as follows: and enabling a user to upload the BOM file or respectively select which column corresponds to the model, the brand, the package and the required quantity in the file, and then searching by using the model or the parameter. The scheme has the following disadvantages: 1. the problem of abnormal BOM table (such as the condition that a plurality of key information are written in the same column or the same key information is written in different columns) cannot be solved; 2. different writing methods (such as parameters) of key information cannot be processed.

Therefore, a new technology is needed for solving the technical problem that the current system quotation cannot process the technical problems that the BOM table is not standard and the writing method of the key information is different.

Disclosure of Invention

The invention mainly aims to solve the technical problems that the current system quotation cannot process the technical problems of nonstandard BOM tables and different key information writing methods.

The invention provides a purchasing method based on BOM identification in a first aspect, which comprises the following steps:

acquiring a BOM file of an electronic component to be purchased;

according to a preset element device category identification algorithm, category label setting processing is carried out on the BOM file to obtain a label BOM file, and the setting label of the label BOM file comprises the following steps: a component model label and a non-component model label;

according to a preset standardization algorithm, carrying out standard standardization processing on the data of the non-element device type label in the label BOM file to obtain a label BOM standard file;

accessing a preset target commodity database, and performing association analysis processing on the tag BOM standard file and the target commodity database according to a preset comparison analysis algorithm to obtain associated commodity data;

and screening the associated commodity data according to the commodity inventory, the commodity unit price and the commodity delivery time length in the target commodity database to obtain the purchased commodity data corresponding to the BOM file.

Optionally, in a first implementation manner of the first aspect of the present invention, the performing, according to a preset meta-device category identification algorithm, category label setting processing on the BOM file to obtain a labeled BOM file includes:

identifying the BOM files line by line based on a Fastext text classification model trained by a neural network to obtain an identification classification result;

and setting the identification classification result as a label of a data grid corresponding to the BOM file to obtain a labeled BOM file.

Optionally, in a second implementation manner of the first aspect of the present invention, the identifying the BOM file line by the Fasttext classification model based on neural network training, and obtaining an identification classification result includes:

identifying and judging whether the first line character string of the BOM file is the type of the model of the component line by line;

if the type of the component is not identified, performing word segmentation processing on the first-row character string according to a preset jieba function to obtain a word segmentation character string set;

performing recognition layering processing on the word segmentation character string set based on a Fasttext classification model trained by a neural network to obtain hierarchical classification data, wherein the hierarchical classification data comprises: the type of the component;

if the type of the component is identified, checking whether the first-line character string is the type of the component based on a preset ES database and a preset KNN algorithm;

if the type of the component is verified, determining the first-line character string as the type of the component;

and if the verification is not the type of the component, entering the process of identifying the type of the component.

Optionally, in a third implementation manner of the first aspect of the present invention, the performing, according to a preset normalization algorithm, standard normalization processing on the data of the non-component model tag in the tag BOM file to obtain a tag BOM specification file includes:

reading a data character string of the non-element device type label in the label BOM file;

performing equivalence conversion processing on the data character string according to a preset equivalence conversion dictionary and designated character setting to obtain a designated character string;

and replacing the data character string in the tag BOM file with the designated character string to obtain a tag BOM standard file.

Optionally, in a fourth implementation manner of the first aspect of the present invention, the accessing a preset target commodity database, and performing association analysis processing on the tag BOM specification file and the target commodity database according to a preset comparison analysis algorithm to obtain associated commodity data includes:

reading a component model character string corresponding to a component model label in the label BOM specification file;

judging whether the component model character string has matched commodity data in the target commodity database;

if matched commodity data exist, determining the commodity data as associated commodity data;

if no matched commodity data exists, reading a non-component model character string set corresponding to a non-component model label in the label BOM standard file;

calculating the character matching number of the commodity data in the target commodity database based on the non-component type character string set to obtain the character matching number corresponding to each commodity data;

and screening the commodity data with the maximum character matching number as the associated commodity data.

Optionally, in a fifth implementation manner of the first aspect of the present invention, the screening the associated product data according to the product inventory, the product unit price, and the product delivery duration in the target product database, and obtaining the purchased product data corresponding to the BOM file includes:

judging whether the commodity inventory in the commodity database corresponding to the associated commodity data is larger than the purchase quantity corresponding to the BOM file or not;

if the data is larger than the purchasing data, screening the associated commodity data to a pre-selected commodity data set;

and screening out the commodity data with the lowest commodity unit price and the shortest commodity delivery time in the pre-selected commodity data set according to the commodity unit price and the commodity delivery time in the commodity database to obtain the purchased commodity data corresponding to the BOM file.

Optionally, in a sixth implementation manner of the first aspect of the present invention, after the screening the associated product data according to the product inventory, the product unit price, and the product delivery duration in the target product database to obtain the purchased product data corresponding to the BOM file, the method further includes:

and calculating the total purchasing price corresponding to the purchased commodity data.

The invention provides a purchasing device based on BOM identification, which comprises:

the acquisition module is used for acquiring a BOM file of the electronic component to be purchased;

the tag setting module is used for carrying out category tag setting processing on the BOM file according to a preset element device category identification algorithm to obtain a tag BOM file, and the setting tag of the tag BOM file comprises the following steps: a component model label and a non-component model label;

the normalization module is used for carrying out standard normalization processing on the data of the non-element device type labels in the label BOM file according to a preset normalization algorithm to obtain a label BOM standard file;

the comparison module is used for accessing a preset target commodity database, and performing correlation analysis processing on the tag BOM standard file and the target commodity database according to a preset comparison analysis algorithm to obtain correlation commodity data;

and the screening module is used for screening the associated commodity data according to the commodity inventory, the commodity unit price and the commodity delivery time length in the target commodity database to obtain the purchased commodity data corresponding to the BOM file.

The third aspect of the present invention provides a purchasing device based on BOM identification, including: a memory having instructions stored therein and at least one processor, the memory and the at least one processor interconnected by a line; the at least one processor invokes the instructions in the memory to cause the BOM identification based purchasing device to perform the BOM identification based purchasing method described above.

A fourth aspect of the present invention provides a computer-readable storage medium having stored therein instructions, which, when run on a computer, cause the computer to execute the above-mentioned purchasing method based on BOM identification.

In the embodiment of the invention, the user only needs to upload the BOM file and does not need to select the column of the key information after uploading. The standardization of different writing methods such as parameters, packaging, brands and the like is realized, the problem that no quotation result or wrong quotation result is caused by different writing methods is solved, and the technical problems that the quotation of the current system cannot process the technical problems that the BOM table is not standard and the key information writing methods are different are solved.

Drawings

FIG. 1 is a schematic diagram of an embodiment of a purchasing method based on BOM identification according to the present invention;

FIG. 2 is a schematic diagram of an embodiment of a purchasing device based on BOM identification in accordance with embodiments of the present invention;

FIG. 3 is a schematic diagram of another embodiment of a purchasing device based on BOM identification in accordance with the present invention;

FIG. 4 is a diagram of an embodiment of a purchasing device based on BOM identification in accordance with embodiments of the present invention.

Detailed Description

The embodiment of the invention provides a purchasing method, device, equipment and storage medium based on BOM identification.

The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

For understanding, a detailed flow of an embodiment of the present invention is described below, and referring to fig. 1, an embodiment of a purchasing method based on BOM identification in an embodiment of the present invention includes:

101. acquiring a BOM file of an electronic component to be purchased;

in this embodiment, the BOM file data may be uploaded to the system by the user, or the BOM data to be purchased may be pulled from the master and slave BOM databases by the system.

102. According to a preset element device category identification algorithm, category label setting processing is carried out on the BOM file to obtain a label BOM file, and the setting label of the label BOM file comprises the following steps: a component model label and a non-component model label;

in the embodiment, the category identification algorithm predicts what category each row in the Excel table represents in a model training + rule mode, the prediction is presented in a probability mode, and the category with the maximum probability value is set as a label. Because the subsequent data processing is to analyze according to the parameter data of the component model and the non-component model, the labels only include a component model label and a non-component model label.

Further, at 102, the following steps may be performed:

1021. identifying BOM files line by line based on a Fasttext classification model trained by a neural network to obtain an identification classification result;

1022. and setting the identification classification result as a label of a data grid corresponding to the BOM file to obtain a labeled BOM file.

In this embodiment, the model judgment is performed by using a Fasttext classification model in combination with a component naming rule. In the early stage, a certain amount of training samples are marked according to a category system, namely, BOM files are labeled, and subsequent models need supervised learning on the linguistic data. The training model adopts a Fasttext classification model, the model training process is a process of continuously adjusting parameters of the neural network, and the cross entropy loss of the model prediction probability output and the real target probability is calculated to carry out back propagation, so that the weight of the neural network is updated. And obtaining a final iteration model under the condition of meeting the discrimination condition of certain precision.

In another embodiment, a naive Bayesian classification model is employed to perform probability statistics based on characteristics of the input information, returning categories and probability values.

The rule checking program details are as follows: acquiring header information and column position information, then performing jieba word segmentation on each row in each column, wherein the word segmentation is based on a dictionary, each word in the dictionary has a corresponding part of speech category, counting the category probability of a word segmentation result, selecting a result with the maximum probability for outputting, and counting a probability formula: total number of lines in word class 1.0/total number of lines in column. And reading the rule configuration file after the class probability value is obtained, searching a corresponding rule under the class to judge the class, and finally returning the class and the probability value. The method adopts a Bayes and professional rule mode, combines Bayesian classification and an industry rule base to identify key columns of the BOM for corresponding processing, ensures the accuracy of column identification under the double verification of a model and the industry rule base, and provides high-purity data for subsequent word segmentation to achieve the purpose of noise reduction. For example, a column contains a combination of "number + unit K", which may represent either the required number (K = thousand) or the resistance of the resistor (K is the unit of the resistance), and if the column is identified as the required number, the word segmentation of the parameter (resistance) is not included.

Further, at 1022, the following steps may be performed:

10221. identifying and judging whether the first line character string of the BOM file is the type of the model of the component line by line;

10222. if the type of the component is not identified, performing word segmentation processing on the first row character string according to a preset jieba function to obtain a word segmentation character string set;

10223. the method comprises the steps of carrying out recognition layering processing on a word character string set based on a Fasttext classification model trained by a neural network to obtain hierarchical classification data, wherein the hierarchical classification data comprises the following steps: the type of the component;

10224. if the type of the component is identified, checking whether the first-line character string is the type of the component based on a preset ES database and a preset KNN algorithm;

10225. if the type of the component is verified, determining the first-line character string as the type of the component;

10226. and if the verification is not the type of the component, entering the process of identifying that the type of the component is not the type of the component.

In step 10221, the initial identification and determination process adopts a first-line character string approximate matching mode, and directly carries out approximate matching with the name in a preset component model dictionary, if hit, the type of the component is determined, and if not hit, the type of the component is not determined.

And if the type of the component is judged for the first time, further logic verification is required. The method enters 10244 to combine the ES database and the k-nearest neighbor method to infer the category to which the model belongs, and further improves the precision by combining with the calculation of the edit distance. And if the type of the component is judged, determining the first-line character string as the type of the component. But if the component model type is not determined, the process returns to step 10222.

And if the type of the component signal is not judged for the first time, a jieba word segmentation frame is adopted, the mode of combining the existing professional dictionary is adopted, the word segmentation is carried out on the data of the whole line, and key information such as the model, the brand, the parameter, the package, the required quantity, the single machine consumption, the position number, the customer material number and the like is respectively extracted.

Line segmentation solves the problem of different key information being written in the same column (e.g., 0402WGF8201TCE Uni-Royal 04028.2 k Ω + -1% 1/16W is written in the same cell, 0402WGF8201TCE is the model, Uni-Royal is the brand, 0402 is the package, 8.2k Ω + -1% 1/16W are parameters), or the same key information being written in different columns (e.g., 8.2k Ω + -1% is written in 1 column, 1/16W is written in another column). The main purpose is to identify meaningful information in the document while typing the correct tag part of speech.

The identification process is carried out by adopting a method of combining a Fasttext classification model with a component naming rule. In the early stage, a certain amount of training samples are marked according to a category system, namely, BOM files are labeled, and subsequent models need supervised learning on the linguistic data. The training model adopts a Fasttext classification model, the model training process is a process of continuously adjusting parameters of the neural network, and the cross entropy loss of the model prediction probability output and the real target probability is calculated to carry out back propagation, so that the weight of the neural network is updated. And obtaining a final iteration model under the condition of meeting the discrimination condition of certain precision.

103. According to a preset standardization algorithm, carrying out standard standardization processing on data of a non-element device type label in a label BOM file to obtain a label BOM standard file;

in this embodiment, different descriptions are unified into a character that can be retrieved by the database, mainly for the case that data retrieval is lost due to different commodity writing methods. For example: one character string "DERF" has three data characters of "5662", "dfaT", and "dafRRT" which are the same, and is converted into "5662" in the data of the non-element model tag.

Further, at 103, the following steps may be performed:

1031. reading a data character string of a non-element device type label in a label BOM file;

1032. performing equivalence conversion processing on the data character string according to a preset equivalence conversion dictionary and designated character setting to obtain a designated character string;

1033. and replacing the data character string in the tag BOM file with the specified character string to obtain a tag BOM standard file.

In steps 1031-. For example, the capacitor package 0402 has different writing methods such as R0402, C0402, 1005, and can be regarded as 0402 when the user writes 1005. Parameter standardization solves the problem that the results cannot be searched or are wrong due to different writing methods of parameters and packaging.

104. Accessing a preset target commodity database, and performing association analysis processing on the tag BOM standard file and the target commodity database according to a preset comparison analysis algorithm to obtain associated commodity data;

in this embodiment, a large number of data commodities exist in the target commodity database, and the names of the tag BOM specification files are approximately matched with the commodity names in the target commodity database, so that associated commodity data can be obtained, and the commodity data can be further screened.

In another embodiment, the component model names of the tag BOM specification file are accurately compared with the trade name names in the target commodity database to obtain associated commodity data.

Further, at 104, the following steps may be performed:

1041. reading a component model character string corresponding to a component model label in a label BOM specification file;

1042. judging whether matched commodity data exist in the target commodity database or not by the component type character strings;

1043. if the matched commodity data exists, determining the commodity data as associated commodity data;

1044. if the matched commodity data does not exist, reading a non-component type character string set corresponding to the non-component type label in the label BOM standard file;

1045. calculating the character matching number of the commodity data in the target commodity database based on the non-component type character string set to obtain the character matching number corresponding to each commodity data;

1046. and screening the commodity data with the maximum character matching number as the associated commodity data.

In 1041-1046 step, in order to find out the product data meeting the necessary parameters maintained in the rules, the commodities associated with all the products are found out. And if necessary parameter product data cannot be found, replacing parameters corresponding to the non-component model labels so as to find approximate parameter products, and searching again by using the replaced parameter values.

And sorting the matching result data according to the matching number of the parameters according to the size. For example, the user gives 5 parameters, 5 parameters in the commodity result set are 2 in accordance, 4 parameters are 5 in accordance, 3 parameters are 10 in accordance, the 5 commodities with the consistent parameters are ranked from 1 to 2, the 4 commodities with the consistent parameters are ranked from 3 to 7, the 3 commodities with the consistent parameters are ranked from 8 to 17, and so on. And determining the commodity data with the largest quantity in the parameter matching as the associated commodity data.

105. And screening the associated commodity data according to the commodity inventory, the commodity unit price and the commodity delivery time length in the target commodity database to obtain the purchased commodity data corresponding to the BOM file.

In this embodiment, the commodity data includes purchasing key information such as commodity inventory, commodity unit price, and commodity delivery time of the associated commodity data, and if a commodity meeting the inventory requirement, the lowest price, and the shortest delivery time is to be found from the associated commodity data, the priority of the three parameters may be manually set according to a policy.

Further, 105 may perform the following steps:

1051. judging whether the commodity inventory in the commodity database corresponding to the associated commodity data is larger than the purchase quantity corresponding to the BOM file;

1052. if the data is larger than the purchasing data, screening the associated commodity data to a pre-selected commodity data set;

1053. and screening out the commodity data with the lowest commodity unit price and the shortest commodity delivery time in the pre-selected commodity data set according to the commodity unit price and the commodity delivery time in the commodity database to obtain the purchased commodity data corresponding to the BOM file.

In the 1051-1053 step, the price, inventory and delivery information of all the products in the result set are obtained. And preferentially judging whether the commodity inventory meets the quantity required by the user or not and preferentially meeting the inventory. Then, the price of the commodity is judged, and the lowest price is prioritized. And finally, judging the delivery period of the commodity, wherein the shortest delivery period has priority. The above recommendation rules are performed on the premise of sorting according to parameters. I.e. the ordering of stock, price, delivery, must be acted upon in the ordering of parameters.

Optionally, the following steps may also follow step 105:

106. and calculating the total purchase price corresponding to the purchased commodity data.

In this embodiment, the total purchase price is a purchase quantity, i.e., a unit price of the purchased commodity, and this step is mainly to better connect with other data computer systems for data analysis, which is not limited herein.

With reference to fig. 2, the purchasing apparatus based on BOM identification in the embodiment of the present invention is described above, and referring to fig. 2, the purchasing apparatus based on BOM identification in the embodiment of the present invention includes:

an obtaining module 201, configured to obtain a BOM file of an electronic component to be purchased;

the tag setting module 202 is configured to perform category tag setting processing on the BOM file according to a preset meta-device category identification algorithm, so as to obtain a tag BOM file, where the setting tag of the tag BOM file includes: a component model label and a non-component model label;

the normalization module 203 is configured to perform standard normalization processing on the data of the non-component type tag in the tag BOM file according to a preset normalization algorithm to obtain a tag BOM specification file;

the comparison module 204 is configured to access a preset target commodity database, perform association analysis processing on the tag BOM specification file and the target commodity database according to a preset comparison analysis algorithm, and obtain associated commodity data;

and the screening module 205 is configured to perform screening processing on the associated commodity data according to the commodity inventory, the commodity unit price, and the commodity delivery duration in the target commodity database, so as to obtain the purchased commodity data corresponding to the BOM file.

Referring to fig. 3, another embodiment of the purchasing apparatus based on BOM identification according to the embodiment of the present invention includes:

Wherein the label setting module 202 comprises:

the recognition unit 2021 is configured to recognize the BOM files line by line based on a Fasttext classification model trained by a neural network, so as to obtain recognition and classification results;

the setting unit 2022 is configured to set the identification and classification result as a tag of a data lattice corresponding to the BOM file, so as to obtain a tagged BOM file.

The identification unit 2021 is specifically configured to:

identifying and judging whether the first-line character string of the BOM file is of the type of the component model line by line according to a preset Bayesian classification algorithm;

Wherein the normalization module 203 is specifically configured to:

Wherein the comparison module 204 is specifically configured to:

judging whether the component model character string has matched commodity data in the purchased commodity database;

calculating the character matching number of the commodity data in the purchased commodity database based on the non-component type character string set to obtain the character matching number corresponding to each commodity data;

Wherein the screening module 205 is specifically configured to:

Wherein, the purchasing apparatus based on BOM identification further includes an aggregation module 206, and the aggregation module 206 is specifically configured to:

Fig. 2 and fig. 3 above describe in detail the purchasing apparatus based on BOM identification in the embodiment of the present invention from the perspective of the modular functional entity, and in the following, describe in detail the purchasing device based on BOM identification in the embodiment of the present invention from the perspective of hardware processing.

Fig. 4 is a schematic structural diagram of a purchasing device based on BOM identification according to an embodiment of the present invention, where the purchasing device 400 based on BOM identification may generate relatively large differences due to different configurations or performances, and may include one or more processors (CPUs) 410 (e.g., one or more processors) and a memory 420, one or more storage media 430 (e.g., one or more mass storage devices) storing applications 433 or data 432. Memory 420 and storage medium 430 may be, among other things, transient or persistent storage. The program stored on storage medium 430 may include one or more modules (not shown), each of which may include a sequence of instructions operating on purchasing device 400 based on BOM identification. Still further, processor 410 may be configured to communicate with storage medium 430 to execute a series of instruction operations in storage medium 430 on purchasing device 400 based on BOM identification.

BOM identification based purchasing device 400 may also include one or more power supplies 440, one or more wired or wireless network interfaces 450, one or more input-output interfaces 460, and/or one or more operating systems 431, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will appreciate that the BOM identification based purchasing device configuration shown in fig. 4 does not constitute a limitation of a BOM identification based purchasing device, and may include more or fewer components than shown, or some components in combination, or a different arrangement of components.

The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, and which may also be a volatile computer-readable storage medium, having stored therein instructions, which, when executed on a computer, cause the computer to perform the steps of the method for purchasing based on BOM identification.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses, and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A purchasing method based on BOM identification is characterized by comprising the following steps:

acquiring a BOM file of an electronic component to be purchased;

screening the associated commodity data according to commodity inventory, commodity unit price and commodity delivery time length in the target commodity database to obtain purchased commodity data corresponding to the BOM file;

the method for setting category labels of the BOM file according to the preset element device category identification algorithm to obtain the labeled BOM file comprises the following steps:

setting the identification classification result as a label of a data grid corresponding to the BOM file to obtain a labeled BOM file;

the identifying the BOM files line by the Fasttext classification model based on the neural network training to obtain the identifying and classifying result comprises the following steps:

2. The purchasing method based on BOM identification of claim 1, wherein the step of performing standard normalization processing on the data of the non-component model tag in the tag BOM file according to a preset normalization algorithm to obtain a tag BOM specification file comprises:

3. The purchasing method based on BOM identification of claim 1, wherein the accessing a preset target goods database, and performing the association analysis processing on the tag BOM specification file and the target goods database according to a preset comparison analysis algorithm to obtain the associated goods data comprises:

4. The purchasing method based on BOM identification as claimed in claim 1, wherein said screening the associated merchandise data according to the merchandise inventory, the merchandise unit price, and the merchandise delivery duration in the target merchandise database, and obtaining the purchased merchandise data corresponding to the BOM file comprises:

5. The purchasing method based on BOM identification as claimed in claim 1, wherein said screening the associated merchandise data according to the merchandise inventory, merchandise unit price and merchandise delivery time length in the target merchandise database, and after obtaining the purchased merchandise data corresponding to the BOM file, further comprises:

6. A purchasing device based on BOM recognition is characterized in that the purchasing device based on BOM recognition comprises:

the screening module is used for screening the associated commodity data according to commodity inventory, commodity unit price and commodity delivery time length in the target commodity database to obtain purchased commodity data corresponding to the BOM file;

wherein, the label setting module includes:

the identification unit is used for identifying the BOM files line by line based on a Fasttext classification model trained by a neural network to obtain an identification classification result;

the setting unit is used for setting the identification classification result as a label of a data grid corresponding to the BOM file to obtain a labeled BOM file;

wherein the identification unit is specifically configured to:

7. A purchasing device based on BOM identification is characterized in that the purchasing device based on BOM identification comprises: a memory having instructions stored therein and at least one processor, the memory and the at least one processor interconnected by a line;

the at least one processor invoking the instructions in the memory to cause the BOM identification based procurement device to perform the BOM identification based procurement method of any of claims 1-5.

8. A computer-readable storage medium, having a computer program stored thereon, wherein the computer program, when executed by a processor, implements a BOM identification based procurement method according to any of claims 1-5.