CN112699174A - Big data building product drive chain diagram generation method - Google Patents

Big data building product drive chain diagram generation method Download PDF

Info

Publication number
CN112699174A
CN112699174A CN202110306025.4A CN202110306025A CN112699174A CN 112699174 A CN112699174 A CN 112699174A CN 202110306025 A CN202110306025 A CN 202110306025A CN 112699174 A CN112699174 A CN 112699174A
Authority
CN
China
Prior art keywords
model
dbg
driving
sku
order
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110306025.4A
Other languages
Chinese (zh)
Other versions
CN112699174B (en
Inventor
陶锋
张勇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China State Construction eCommerce Co Ltd
Original Assignee
China State Construction eCommerce Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China State Construction eCommerce Co Ltd filed Critical China State Construction eCommerce Co Ltd
Priority to CN202110306025.4A priority Critical patent/CN112699174B/en
Publication of CN112699174A publication Critical patent/CN112699174A/en
Application granted granted Critical
Publication of CN112699174B publication Critical patent/CN112699174B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2465—Query processing support for facilitating data mining operations in structured databases
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/248—Presentation of query results
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/08—Construction

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • General Physics & Mathematics (AREA)
  • Finance (AREA)
  • Development Economics (AREA)
  • Accounting & Taxation (AREA)
  • Economics (AREA)
  • Data Mining & Analysis (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Databases & Information Systems (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Computational Linguistics (AREA)
  • Tourism & Hospitality (AREA)
  • Human Resources & Organizations (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Probability & Statistics with Applications (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Fuzzy Systems (AREA)
  • Primary Health Care (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention provides a method for generating a drive chain diagram of big data building supplies, which comprises the following steps: mining the driving relation of each building article of the order to generate driving relation basic data and establish a structural model; identifying and determining a core object and other objects for the internal structure of the driving relationship, and inputting the core object and other objects into the structure corresponding to the structure model to establish a matching model; calculating driving probability values between the core object and other objects and driving probability values between other objects; and selecting a matching model, comparing other objects in the other matching models with the core object in the matching model, connecting the corresponding class structures of the other objects and the core object when the other objects and the core object are the same, and generating a driving chain diagram by combining the driving probability value. The invention provides a drive chain diagram which can show the influence relation of one main drive commodity and other n groups of commodity sets under different dynamic combination conditions, and can support the sales procurement personnel to more specifically carry out inventory satisfaction and gap analysis.

Description

Big data building product drive chain diagram generation method
Technical Field
The invention relates to the technical field of big data analysis, in particular to a big data building product drive chain diagram generation method.
Background
With the centralization of the purchasing and selling of the building products, a large number of batch-type purchasing orders are recorded in the selling system, and the orders comprise various types of building products. Behind these purchase orders are the implementation requirements of various construction projects, and in a building construction scene, multiple categories of building supplies are generally required to be purchased for a certain construction.
The popularization of the online sales mode shortens the change of orders and stocks to the minute level or even the second level, and when a new order comes, stock purchasing personnel need to analyze the stock gaps of other building supplies in real time according to data of a system background so as to support a new centralized purchasing strategy. This process requires that the purchasing analyst has more clear concrete understanding to the sales relation between the building supplies in self system, and the instrument that supports the purchasing analyst to carry out the analysis at present is comparatively loaded down with trivial details, still needs a large amount of manual experiences. The problems of long time consumption and easy error exist.
The order requirements of the building products are greatly different from the structures of common consumer goods, the consumption of the building products of all types is closely related, and the deeper relationship is not easy to find by utilizing the traditional big data association clustering method, so that the inventory requirement analysis under the business background is not enough supported.
Disclosure of Invention
The invention aims to provide a big data building product drive chain diagram generation method which runs in a system through calculation programming and updates in real time and can support a buyer to be used for inventory gap analysis of a building product centralized online sales scene. According to the method, a building product driving chain diagram is generated by analyzing a large amount of sales order data in a background system, so that the support relationship among various types of building products in the near term can be reflected clearly, and further high-order data analysis means are enriched.
The embodiment of the invention is realized by the following technical scheme: a big data building product drive chain diagram generation method comprises the following steps:
s1, mining a driving relationship according to an order placing rule among the building products in the obtained order, generating basic data of the driving relationship of the building products and establishing a structural model of the driving relationship;
s2, identifying the internal structure of the driving relationship to determine a core object and other objects of the driving relationship, and inputting the core object and other objects into a class structure corresponding to the structural model of the driving relationship to establish a driving relationship matching model D-model;
s3, calculating driving probability values between the core object and other objects and driving probability values between other objects;
s4, selecting one D-model, comparing other objects in the other D-models with the core object in the D-model, connecting the class structures corresponding to the other objects and the core object when the other objects and the core object are the same, and generating a driving chain diagram by combining the driving probability value calculated in the step S3.
According to a preferred embodiment, step S1 further includes the steps of:
s11, according to the ordering rule of the building supplies, a structural model of the driving relationship is created, wherein the data structure of the structural model of the driving relationship comprises the following steps: main driving commodity structureDG(ii) a Driven commodity collection structureDBG(ii) a Secondary driven commodity class structureS-DBG；DGAndDBGthe main driving relationship among the two dRel-m;DGandS-DBGthe secondary drive relationship between dRel-s;DBGandS-DBGthe auxiliary driving relationship among the two dRel-a;
s12, analyzing the obtained orders to calculate the sum of the comprehensive errors, and generating the driving relationship basic data corresponding to each order, wherein the calculation formula of the sum of the comprehensive errors is as follows:
Figure 419646DEST_PATH_IMAGE001
（1）
in the formula (1), inBWhen the order is smaller than the preset threshold value, the order conforms to the basic data requirement of the driving relationship;ma quantity of SKU for the total number of categories of construction supplies in the order;kordering the current SKU in the order, whereinkThe value range of (a) is [1,m]；Cthe number of purchases for the current SKU;Pis a sheetA price;
Figure 661271DEST_PATH_IMAGE002
the ratio of the time difference of the order is obtained;Afor the material categories to which the SKU belongs, different material categories correspond to different values.
According to a preferred embodiment, step S12
Figure 998712DEST_PATH_IMAGE002
Calculated by the following method:
when in usekWhen the ratio is not less than 1,
Figure 642183DEST_PATH_IMAGE002
is 1.0, whenk=mWhen the temperature of the water is higher than the set temperature,
Figure 977349DEST_PATH_IMAGE002
is 2.0; when in usekIn the case of other values of the value,
Figure 706271DEST_PATH_IMAGE002
calculated by the following formula, the formula is expressed as:
Figure 785085DEST_PATH_IMAGE003
（2）
in the formula (2), the reaction mixture is,T 1the order time for the first SKU is,T m the order time for the last SKU is,T k the order time is the currently calculated SKU order time.
According to a preferred embodiment, step S2 further includes the steps of:
s21, influence value on all SKUs in orderICalculating, using SKU with maximum influence value as driving relation core object, and dividing the driving relation core object into main driving commodityDGWherein the influence value is calculatedI ofThe formula is expressed as:
Figure 96112DEST_PATH_IMAGE005
（3）
S22.excluding driving relationship core objects, remainderm-1 SKU as other object;
s23, repeating the steps until all orders are processed and form the D-model.
According to a preferred embodiment, step S22 further includes:
s221. by the pair of the restm-secondary impact value of 1 SKUDSAnd calculating to determine a driven object in other objects, wherein the formula is as follows:
Figure 867759DEST_PATH_IMAGE007
（4）
s222. for the result of calculationDSSorting by magnitude of value in descending order，Accumulating the remainder one by oneDSWhen the accumulated value is just greater than
Figure 818397DEST_PATH_IMAGE008
In time, the SKU corresponding to the secondary influence value participating in accumulation is divided into the driven commodity setDBGWhereinD total Is left overm-sum of secondary impact values of 1 SKU;
s223, if remaining SKUs exist, all the SKUs are divided into the secondary driven commodity setS-DBG。
According to a preferred embodiment, step S3 further includes the steps of:
s31, calculatingDGDrive theDBGProbability value of (2)P m The formula is expressed as:
Figure 700902DEST_PATH_IMAGE009
（5）
in the formula (5), the reaction mixture is,k 1 is the same as that in the current D-modelDGAndDBGin (1)SKUA total number of D-models that are identical in sequence;t e is composed ofk 1 In the same D-model, the latest order placing time of the corresponding order is obtained;t s is composed ofk 1 The earliest ordering time corresponding to the order in the same D-model;tthe order placing time of the current D-model corresponding order is obtained;C DG is the same as that in the current D-modelDGThe same order placing times of all the D-models corresponding to the orders are obtained;C DBG is the same as that in the current D-modelDGSKU andDBGall the D-models with the same SKU sequence correspond to the order placing times of the order;
s32, calculatingDGDriving S-DBGProbability value of (2)P S The formula is expressed as:
Figure 53386DEST_PATH_IMAGE010
（6）
in the formula (6), the reaction mixture is,C S-DBG is the same as that in the current D-modelDGSKU and S-DBGAll the D-models with the same SKU sequence correspond to the order placing times of the order;
S33.DBGdriving S-DBGProbability value of (2)P a According toP m AndP s is determined, the formula is expressed as:
Figure 58252DEST_PATH_IMAGE011
（7）。
according to a preferred embodiment, step S4 further includes the steps of:
s41, selecting one D-model, and traversing another D-modelDBGAnd S-DBGThe SKU sequence in (1), and the selected D-modelDGIf the same SKU exists, the SKU will existDBGAnd S-DBGConnecting selected ones of the D-modelsDGAnd establishing a dRel-m, dRel-s or dRel-a association relation;
wherein the result of calculation isP m As a weight of the association dRel-m,P s as a weight of the association dRel-s,P a then the weight value is used as the weight value of the incidence relation dRel-a;
and S42, repeating the steps until all the D-models are compared.
According to a preferred embodiment, in step S41, the association relationship between the D-model and another D-model is established as follows:
modifying the pointing relation of the class structure in the D-model according to the association relation of the dRel-m, the dRel-s and the dRel-a;
calculated in step S41P m As a weight of the association dRel-m,P s as a weight of the association dRel-s,P a then, as the weight of the association dRel-a, the following is specifically mentioned:
according toP m 、P s AndP a and modifying the driving probability member variable in the class structure data class instance in the D-model.
According to a preferred embodiment, step S1 is preceded by the steps of:
combining orders in batches, namely combining scattered orders of the same customer within a preset time range to generate a composite order comprising multiple SKUs;
removing invalid orders, namely removing the combined composite orders with the SKU number of 1, and taking the rest orders as valid orders;
and sequencing the effective orders according to the order placing time of the orders.
According to a preferred embodiment, the material category in step S12 includes a main material category, an auxiliary material category and a tool category, wherein the value a of the main material category is greater than the value a of the auxiliary material category is greater than the value a of the tool category.
The technical scheme of the embodiment of the invention at least has the following advantages and beneficial effects: (1) the driving chain diagram generated by the method provided by the invention can show the influence relationship of one main driving commodity and other n groups of commodity sets under different dynamic combination conditions, and can particularly support sales buyers to carry out inventory satisfaction degree and gap analysis more specifically;
(2) the driving relation matching model D-model is generated according to ordering rules of different types of articles during building construction, so that the driving relation matching model D-model has stronger industrial pertinence, can reflect the influence relation of all types of building articles more visually, and can provide a data basis for visualization of inventory requirements of the building articles;
(3) the drive chain graph uses the SKU as the unique identification of the building product analysis, and can provide a new analysis dimension for the user image of the order initiating customer.
Drawings
Fig. 1 is a flowchart of steps of a method for generating a drive chain diagram according to embodiment 1 of the present invention;
FIG. 2 is a schematic diagram of a D-model logic structure provided in embodiment 1 of the present invention;
FIG. 3 is a logic diagram of a D-model connection result provided in embodiment 1 of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Example 1
Referring to fig. 1, the embodiment provides a method for generating a drive chain diagram of a big data building product, which runs in a system through calculation programming and updates in real time, and can support a buyer to use for stock gap analysis of a centralized online sales scene of the building product. The method generates the building product driving chain diagram by analyzing a large amount of sales order data in the background system. The generated driving chain diagram can clearly reflect the support relationship among various recent building products, and further enriches high-order data analysis means, and the method comprises the following steps:
preprocessing order data, namely preprocessing sales order data obtained by a background system and determining source data which have conditions and are used for generating a driving chain diagram; the pretreatment specifically comprises the following steps:
creating a preprocessing calculation cache pool and a preprocessing result cache pool in a Hive database of Hadoop; loading the sales order data acquired from the background service system into a calculation cache pool, and sequencing the sales order data from far to near according to order placing time; and combining scattered orders of the same customer in a similar time range to form a composite virtual order containing multiple SKUs of the building products.
Using JAVA program to read in the first order of calculation buffer pool to become class object and find the second orderkIndividual scattered orderORD k Within 24 hours of the order, if found to be laternAnd combining the scattered orders to form a composite order containing n +1 SKUs, deleting the original scattered orders, and storing the combined composite order into a preprocessing result cache pool in the Hive. The method for judging the scattered order is to judge whether the number of SKUs contained in the order is 1, and if the number of SKUs in the order is 1, namely only one type of building products is contained, the order is deleted. And repeating the steps until all orders in the calculation preprocessing calculation cache pool are processed.
The preprocessing also comprises invalid order elimination, namely scanning the residual orders in the preprocessing calculation cache pool, and deleting the data of the still scattered orders from the database of the preprocessing calculation cache pool.
Furthermore, the system also comprises a sequencing preprocessing flow, all effective orders are sequenced according to the order placing time of the building supplies, and the sequencing rule is preferably from far to near.
After data acquired from a background service system is subjected to a data preprocessing process, a driving relationship mining step is carried out, wherein the driving relationship mining step specifically comprises the following steps:
creating a driving relation matching model D-model, and creating a structural model of the driving relation according to the ordering rule of the building supplies, wherein the driving relation is a main driving commodityDGDriven commodity collectionDBGAnd secondary driven commodity setS-DBGThree main objects. Furthermore, three relations of a main driving relation dRel-m, a secondary driving relation dRel-s and an auxiliary driving relation dRel-a are included. WhereinDBGAndS-DBGmay include a plurality of construction article SKUs, anDGOnly one SKU, the drive core object, is included as the primary drive item. In this embodiment, the switch is drivenThe system matching model D-model is realized by class oriented programming, and respectively and correspondingly creates SKU,DG、DBG、S-DBGClass I, whereinDG、DBG、S-DBGAnd creating a SKU sequence with a sequence number in the class, wherein the initial state of the SKU sequence is an empty element sequence.
Create dRel-m, dRel-s, dRel-a class structural data class implementation, and willP m 、P s 、P a As member variables are contained therein, respectively.
And (3) creating a large D-model class, containing all the class objects, forming a final data structure, wherein the final formed D-model logical structure is shown in FIG. 2.
Further, generating the driving relationship basic data specifically includes the following steps:
and loading source data from the preprocessing result buffer pool, and simultaneously creating a driving relation basic data pool.
And loading the first order of the preprocessing result buffer pool for analysis and calculation, generating corresponding driving relationship basic data, and calculating the matching degree of the driving relationship.
Calculating the first step, calculating the ratio of the next single time difference
Figure 496186DEST_PATH_IMAGE002
In particular, the amount of the surfactant is,
Figure 916803DEST_PATH_IMAGE002
obtained by the following methodkWhen the ratio is not less than 1,
Figure 389373DEST_PATH_IMAGE002
is 1.0; when in usek=mWhen the temperature of the water is higher than the set temperature,
Figure 502822DEST_PATH_IMAGE002
is 2.0;
kin the case of other values of the value,
Figure 742567DEST_PATH_IMAGE002
is between [1.0, 2.0 ]]According to the ratio betweenkIndividual constructionOrdering time of building articlesT k 1 st time for placing order of building articlesT 1The time of ordering with the last building articleT m The formula is as follows:
Figure 966875DEST_PATH_IMAGE003
the sum of the composite errors is then calculatedBThe formula is as follows:
Figure 28372DEST_PATH_IMAGE001
in the formula (I), the compound is shown in the specification,Bi.e. the sum of the composite errors in the process of calculating the degree of matching of the driving relationshipBWhen the order is smaller than the preset threshold value, the order conforms to the basic data requirement of the driving relationship;ma quantity of SKU for the total number of categories of construction supplies in the order;kordering the current SKU in the order, whereinkThe value range of (a) is [1,m]；Cthe number of purchases for the current SKU;Pis monovalent;
Figure 312723DEST_PATH_IMAGE002
the ratio of the time difference of the order is obtained;Afor the material categories to which the SKU belongs, different material categories correspond to different values;
in this embodiment, the material category includes main material category, auxiliary material category and tool category, and usually, the main material categoryAOf value greater than that of the auxiliary materialAOf value greater than that of the tool typeAA value; preferably, the main material corresponds toAValue 1.0, corresponding to the auxiliary materialAValue 0.75, corresponding to tool classAThe value was 0.5. In this exampleBIs set to 0.2, i.e., whenBWhen the value of (2) is less than 0.2, the order is considered to be in accordance with the driving relationship basic data, and the order data is stored in a driving relationship basic data pool. And repeating the steps to know that all orders in the preprocessing result buffer pool are processed. And then emptying or deleting the preprocessing result buffer pool.
Further, after mining the driving relationship of the order, identifying the core object in the driving relationship, specifically including the following steps:
loading a first order in a driving relationship basic data pool, and firstly identifying a main driving building product, wherein the main driving building product refers to the SKU with the maximum influence or action in the order, namely the SKU with the maximum influence value, and the calculation influence value formula is as follows:
Figure 725250DEST_PATH_IMAGE012
calculating an impact value for a SKUIAfter that, the SKU with the largest influence value is classified into the D-model class structureDGSKU sequence of class structure.
By aligning the restm-secondary impact value of 1 SKUDSAnd calculating to determine a driven object in other objects, wherein the formula is as follows:
Figure 815565DEST_PATH_IMAGE013
further, for the calculatedDSSorting by magnitude of value in descending order，Accumulating the remainder one by oneDSWhen the accumulated value is just greater than
Figure 997148DEST_PATH_IMAGE008
In time, the SKU corresponding to the secondary influence value participating in accumulation is divided into the driven commodity setDBGWhereinD total Is left overm-sum of secondary impact values of 1 SKU; if there are remaining SKUs, all the secondary driven commodity set is sorted intoS-DBG。
The above steps are repeated until all orders are processed and a D-model is formed.
Further, after all the metadata are generated into a D-model, calculating a driving probability value between the core object and another object and a driving probability value between the other objects specifically includes:
and copying the basic data pool of the driving relationship in the memory to generate a pool to be processed for calculating the driving probability, thereby avoiding the confusion of the reference object.
Loading a first D-model into the pool to be processed, and respectively acquiring the D-modelsDG、DBG、S-DBGThe SKU sequence in (1). Searching and calculating in the whole driving relation basic data pool according to the three SKU sequencesC DG 、C DBG 、C S-DBG The value of (c).C DG Is the same as that in the current D-modelDGThe same order placing times of all the D-models corresponding to the orders are obtained;C DBG is the same as that in the current D-modelDGSKU andDBGall the D-models with the same SKU sequence correspond to the order placing times of the order;C S-DBG is the same as that in the current D-modelDGSKU and S-DBGAll D-models with the same SKU sequence correspond to the order placement times.
Next, calculateDGDrive theDBGProbability value of (2)P m The formula is expressed as:
Figure 186821DEST_PATH_IMAGE009
in the formula (I), the compound is shown in the specification,k 1 is the same as that in the current D-modelDGAndDBGin (1)SKUA total number of D-models that are identical in sequence;t e is composed ofk 1 In the same D-model, the latest order placing time of the corresponding order is obtained;t s is composed ofk 1 The earliest ordering time corresponding to the order in the same D-model;tthe order placing time of the current D-model corresponding order is the time unit of day.
Further, calculatingDGDriving S-DBGProbability value of (2)P S The formula is expressed as:
Figure 86644DEST_PATH_IMAGE010
in particular, the method comprises the following steps of,DBGdriving S-DBGProbability value of (2)P a According toP m AndP s is determined, the formula is expressed as:
Figure 652754DEST_PATH_IMAGE011
P s 、P m 、P a and after the calculation is finished, directly operating the D-model instance object to realize numerical value updating.
Repeating the steps until all the driving probabilities calculate the D-model of the pool to be processedP s 、P m 、P a And finishing the updating. Furthermore, objects from the driving probability calculation completion pool to the driving probability calculation standby pool are re-referenced, and confusion is avoided.
Then, a driving chain diagram generation step is performed, specifically as follows:
and (3) creating a new graph data structure, wherein the initial state is 0 node and 0 relation. And randomly selecting a D-model as a node to be added into the graph data structure.
Traverse another D-modelDBGAnd S-DBGThe SKU sequence in (1), and the selected D-modelDGIf the same SKU exists, the SKU will existDBGAnd S-DBGConnecting selected ones of the D-modelsDGAnd establishing a dRel-m, dRel-s or dRel-a incidence relation, namely modifying the pointing relation in the D-model seed linked list data class instance.
At the same time will be calculatedP m As a weight of the association dRel-m,P s as a weight of the association dRel-s,P a the drive probability P member variable in the chained list dataclass instance in the D-model is modified as a weight for the association dRel-a. Preferably, the D-models are selected from the remaining D-models, except for the first selected D-modelDGCan only be connected.
Selecting the next D-model, repeating the operation,up to all D-modelsDBGAndS-DBGthe alignment was completed.
Referring to fig. 3, a minimum scale chain diagram result logic is shown: when in useDBGAnd S-DBGWith some SKU appearing in other D-modelsDGThe same SKU of the connection result logic diagram of (a), is described in detail below:
randomly selecting one D-model as the current D-model; while scanning other D-modelsDGWhen a SKU sequence in the S-DBG is found with a D-modelDGAre the same (e.g., both 20200911), the SKU (e.g., example SKU)b) With the new D-modelDGConnecting;
similarly, when a SKU sequence in a DBG is found in a D-modelDGAre the same (e.g., both 20200323), the SKU (e.g., example SKU)3) With the new D-modelDGThe connection is made.
Through the steps, a minimum-scale driving chain diagram consisting of three D-model connections is obtained.
In summary, the drive chain diagram generated by the method provided by the invention can show the influence relationship between one main drive commodity and other n groups of commodity sets under different dynamic combination conditions, and can specifically support the sales procurement personnel to more specifically perform inventory satisfaction and gap analysis; the driving relation matching model D-model is generated according to ordering rules of different types of articles during building construction, so that the driving relation matching model D-model has stronger industrial pertinence, can reflect the influence relation of all types of building articles more visually, and can provide a data basis for visualization of inventory requirements of the building articles; the drive chain graph uses the SKU as the unique identification of the building product analysis, and can provide a new analysis dimension for the user image of the order initiating customer.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes will occur to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. A big data building product drive chain diagram generation method is characterized by comprising the following steps:
s1, mining a driving relationship according to an order placing rule among the building products in the obtained order, generating basic data of the driving relationship of the building products and establishing a structural model of the driving relationship;
s2, identifying the internal structure of the driving relationship to determine a core object and other objects of the driving relationship, and inputting the core object and other objects into a class structure corresponding to the structural model of the driving relationship to establish a driving relationship matching model D-model;
s3, calculating driving probability values between the core object and other objects and driving probability values between other objects;
s4, selecting one D-model, comparing other objects in the other D-models with the core object in the D-model, connecting the class structures corresponding to the other objects and the core object when the other objects and the core object are the same, and generating a driving chain diagram by combining the driving probability value calculated in the step S3.
2. The big data construction product drive chain map generation method according to claim 1, wherein the step S1 further comprises the steps of:
s11, according to the ordering rule of the building supplies, a structural model of the driving relationship is created, wherein the data structure of the structural model of the driving relationship comprises the following steps: main driving commodity structureDG(ii) a Driven commodity collection structureDBG(ii) a Secondary driven commodity class structureS-DBG；DGAndDBGthe main driving relationship among the two dRel-m;DGandS-DBGthe secondary drive relationship between dRel-s;DBGandS- DBGthe auxiliary driving relationship among the two dRel-a;
s12, analyzing the obtained orders to calculate the sum of the comprehensive errors, and generating the driving relationship basic data corresponding to each order, wherein the calculation formula of the sum of the comprehensive errors is as follows:
Figure 858410DEST_PATH_IMAGE001
（1）
in the formula (1), inBWhen the order is smaller than the preset threshold value, the order conforms to the basic data requirement of the driving relationship;ma quantity of SKU for the total number of categories of construction supplies in the order;kordering the current SKU in the order, whereinkThe value range of (a) is [1,m]；Cthe number of purchases for the current SKU;Pis monovalent;
Figure 775550DEST_PATH_IMAGE002
the ratio of the time difference of the order is obtained;Afor the material categories to which the SKU belongs, different material categories correspond to different values.
3. The big data construction product drive chain map generation method according to claim 2, wherein in step S12
Figure 555288DEST_PATH_IMAGE002
Calculated by the following method:
when in usekWhen the ratio is not less than 1,
Figure 12814DEST_PATH_IMAGE002
is 1.0, whenk=mWhen the temperature of the water is higher than the set temperature,
Figure 827186DEST_PATH_IMAGE002
is 2.0; when in usekIn the case of other values of the value,
Figure 649649DEST_PATH_IMAGE002
calculated by the following formula, the formula is expressed as:
Figure 916682DEST_PATH_IMAGE003
（2）
in the formula (2), the reaction mixture is,T 1the order time for the first SKU is,T m the order time for the last SKU is,T k for the currently calculated SKAnd U ordering time.
4. The big data construction product drive chain map generation method according to claim 2 or 3, wherein the step S2 further comprises the steps of:
s21, influence value on all SKUs in orderICalculating, using SKU with maximum influence value as driving relation core object, and dividing the driving relation core object into main driving commodityDGWherein the influence value is calculatedI ofThe formula is expressed as:
Figure 115582DEST_PATH_IMAGE004
（3）
s22, excluding the driving relation core object and remainingm-1 SKU as other object;
s23, repeating the steps until all orders are processed and form the D-model.
5. The big data construction product drive chain map generation method according to claim 4, wherein the step S22 further comprises:
s221. by the pair of the restm-secondary impact value of 1 SKUDSAnd calculating to determine a driven object in other objects, wherein the formula is as follows:
Figure 597510DEST_PATH_IMAGE006
（4）
s222. for the result of calculationDSSorting by magnitude of value in descending order，Accumulating the remainder one by oneDSWhen the accumulated value is just greater than
Figure 590874DEST_PATH_IMAGE007
In time, the SKU corresponding to the secondary influence value participating in accumulation is divided into the driven commodity setDBGWhereinD total Is left overm-sum of secondary impact values of 1 SKU;
s223, if remaining SKUs exist, dividing all SKUsSecondary driven commodity collectionS-DBG。
6. The big data construction product drive chain map generation method according to claim 2, wherein the step S3 further comprises the steps of:
s31, calculatingDGDrive theDBGProbability value of (2)P m The formula is expressed as:
Figure 345203DEST_PATH_IMAGE008
（5）
in the formula (5), the reaction mixture is,k 1 is the same as that in the current D-modelDGAndDBGin (1)SKUA total number of D-models that are identical in sequence;t e is composed ofk 1 In the same D-model, the latest order placing time of the corresponding order is obtained;t s is composed ofk 1 The earliest ordering time corresponding to the order in the same D-model;tthe order placing time of the current D-model corresponding order is obtained;C DG is the same as that in the current D-modelDGThe same order placing times of all the D-models corresponding to the orders are obtained;C DBG is the same as that in the current D-modelDGSKU andDBGall the D-models with the same SKU sequence correspond to the order placing times of the order;
s32, calculatingDGDriving S-DBGProbability value of (2)P S The formula is expressed as:
Figure 82215DEST_PATH_IMAGE009
（6）
in the formula (6), the reaction mixture is,C S-DBG is the same as that in the current D-modelDGSKU and S-DBGAll the D-models with the same SKU sequence correspond to the order placing times of the order;
S33.DBGdriving S-DBGProbability value of (2)P a According toP m AndP s size determination of (D), formularyShown as follows:
Figure 605600DEST_PATH_IMAGE010
（7）。
7. the big data construction product drive chain map generation method according to claim 6, wherein the step S4 further comprises the steps of:
s41, selecting one D-model, and traversing another D-modelDBGAnd S-DBGThe SKU sequence in (1), and the selected D-modelDGIf the same SKU exists, the SKU will existDBGAnd S-DBGConnecting selected ones of the D-modelsDGAnd establishing a dRel-m, dRel-s or dRel-a association relation;
wherein the result of calculation isP m As a weight of the association dRel-m,P s as a weight of the association dRel-s,P a then the weight value is used as the weight value of the incidence relation dRel-a;
and S42, repeating the steps until all the D-models are compared.
8. The method for generating the drive chain map of the big data construction product according to claim 7, wherein in step S41, an association relationship between the D-model and another D-model is established as follows:
modifying the pointing relation of the class structure in the D-model according to the association relation of the dRel-m, the dRel-s and the dRel-a;
calculated in step S41P m As a weight of the association dRel-m,P s as a weight of the association dRel-s,P a then, as the weight of the association dRel-a, the following is specifically mentioned:
according toP m 、P s AndP a and modifying the driving probability member variable in the class structure data class instance in the D-model.
9. The big data construction product drive chain map generation method according to claim 1, wherein step S1 is preceded by the steps of:
combining orders in batches, namely combining scattered orders of the same customer within a preset time range to generate a composite order comprising multiple SKUs;
removing invalid orders, namely removing the combined composite orders with the SKU number of 1, and taking the rest orders as valid orders;
and sequencing the effective orders according to the order placing time of the orders.
10. The big data construction product drive chain map generation method of claim 2, wherein the big material class in step S12 includes a main material class, an auxiliary material class and a tool class, wherein the value a of the main material class is greater than the value a of the auxiliary material class.
CN202110306025.4A 2021-03-23 2021-03-23 Big data building product drive chain diagram generation method Active CN112699174B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110306025.4A CN112699174B (en) 2021-03-23 2021-03-23 Big data building product drive chain diagram generation method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110306025.4A CN112699174B (en) 2021-03-23 2021-03-23 Big data building product drive chain diagram generation method

Publications (2)

Publication Number Publication Date
CN112699174A true CN112699174A (en) 2021-04-23
CN112699174B CN112699174B (en) 2021-08-17

Family

ID=75515423

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110306025.4A Active CN112699174B (en) 2021-03-23 2021-03-23 Big data building product drive chain diagram generation method

Country Status (1)

Country Link
CN (1) CN112699174B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113077390A (en) * 2021-06-04 2021-07-06 中建电子商务有限责任公司 Image rectification algorithm based on deep learning

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100138277A1 (en) * 2008-12-03 2010-06-03 At&T Intellectual Property I, L.P. Product migration analysis using data mining
CN102663594A (en) * 2012-03-31 2012-09-12 浙江工商大学 Electronic commerce catalog distribution system
CN103700005A (en) * 2013-12-17 2014-04-02 南京信息工程大学 Association-rule recommending method based on self-adaptive multiple minimum supports
CN105701681A (en) * 2016-01-06 2016-06-22 北京京东尚科信息技术有限公司 Method and device used for predicting order amount
CN106446943A (en) * 2016-09-19 2017-02-22 大连海事大学 Commodity correlation big data sparse network quick clustering method
CN106649714A (en) * 2016-12-21 2017-05-10 重庆邮电大学 topN recommendation system and method for data non-uniformity and data sparsity
CN108564244A (en) * 2018-02-26 2018-09-21 新智数字科技有限公司 Order processing method, apparatus, equipment and semiconductor readable storage medium storing program for executing
CN111047009A (en) * 2019-11-21 2020-04-21 腾讯科技（深圳）有限公司 Event trigger probability pre-estimation model training method and event trigger probability pre-estimation method
CN112365295A (en) * 2020-12-01 2021-02-12 杭州拼便宜网络科技有限公司 Method, device and equipment for predicting order taking probability and readable storage medium

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100138277A1 (en) * 2008-12-03 2010-06-03 At&T Intellectual Property I, L.P. Product migration analysis using data mining
CN102663594A (en) * 2012-03-31 2012-09-12 浙江工商大学 Electronic commerce catalog distribution system
CN103700005A (en) * 2013-12-17 2014-04-02 南京信息工程大学 Association-rule recommending method based on self-adaptive multiple minimum supports
CN105701681A (en) * 2016-01-06 2016-06-22 北京京东尚科信息技术有限公司 Method and device used for predicting order amount
CN106446943A (en) * 2016-09-19 2017-02-22 大连海事大学 Commodity correlation big data sparse network quick clustering method
CN106649714A (en) * 2016-12-21 2017-05-10 重庆邮电大学 topN recommendation system and method for data non-uniformity and data sparsity
CN108564244A (en) * 2018-02-26 2018-09-21 新智数字科技有限公司 Order processing method, apparatus, equipment and semiconductor readable storage medium storing program for executing
CN111047009A (en) * 2019-11-21 2020-04-21 腾讯科技（深圳）有限公司 Event trigger probability pre-estimation model training method and event trigger probability pre-estimation method
CN112365295A (en) * 2020-12-01 2021-02-12 杭州拼便宜网络科技有限公司 Method, device and equipment for predicting order taking probability and readable storage medium

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
P KAZIENKO: "Product Recommendation in e-Commerce Using Direct and Indirect Confidence for Historical User Sessions", 《DISCOVERY SCIENCE, 7TH INTERNATIONAL CONFERENCE, DS 2004, PADOVA, ITALY, OCTOBER 2-5, 2004, PROCEEDINGS》 *
李敏波 等: "工业大数据分析技术与轮胎销售数据预测", 《计算机工程与应用》 *
赵爱民: "黄金首饰进销存及客户偏好大数据分析系统设计与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *
黄嘉发: "基于数据挖掘的销售预警模型分析与研究", 《中国优秀硕士学位论文全文数据库 经济与管理科学辑》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113077390A (en) * 2021-06-04 2021-07-06 中建电子商务有限责任公司 Image rectification algorithm based on deep learning

Also Published As

Publication number Publication date
CN112699174B (en) 2021-08-17

Similar Documents

Publication Publication Date Title
US10387779B2 (en) Systems and methods for multi-objective evolutionary algorithms with soft constraints
CN103678447B (en) Multivariate transaction classification
Pervin et al. An integrated inventory model with variable holding cost under two levels of trade-credit policy
Shirole et al. Customer segmentation using rfm model and k-means clustering
CN103903148A (en) C2C system, B2C system, and application methods thereof
Jones et al. US multinationals in British manufacturing before 1962
CN114611959A (en) O2O big data technology-based product selection strategy system
CN112699174B (en) Big data building product drive chain diagram generation method
CN111080396A (en) Commodity pricing method and device based on competitive commodities and storage medium
CN111275294B (en) Commodity information batch processing analysis method, device and equipment
Patil et al. Fuzzy TOPSIS and grey relation analysis integration for supplier selection in fiber industry
CN102254240A (en) Data object substitution device and data object substitution method
TW202141386A (en) Method and system for processing consumption data
JP5018313B2 (en) Information structuring apparatus, information structuring method, and program
Winarso et al. Supplier selection of 40th container in PT Tribudhi Pelita Indonesia using analytical hierarchy process (AHP) method
CN102567826A (en) Method and system for processing material bills
Widyadana et al. Production inventory models for deteriorating items with stochastic machine unavailability time, lost sales and price-dependent demand
CN114663022A (en) Decision tree-based warehousing model decision optimization method
Sadeghi et al. An EPQ model for deteriorating products with delayed payments and shortage
Shen et al. Data Mining the Data Processing Technologies for Inventory Management.
Alborzi et al. A multi-objective model for supply chain network design under stochastic demand
Selvaraju et al. EOQ models for deteriorative items with constant, linear and quadratic holding cost and shortages-a comparative study
CN112116178A (en) Ordering method and device for offline stores
Nikitin et al. Shopping Basket Analisys for Mining Equipment: Comparison and Evaluation of Modern Methods
Ahmad et al. Apparel supply chain optimization focusing on right supplier selection

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant