CN117557346A - Full-link intelligent business decision analysis method based on dynamic consumption data - Google Patents
Full-link intelligent business decision analysis method based on dynamic consumption data Download PDFInfo
- Publication number
- CN117557346A CN117557346A CN202410039603.6A CN202410039603A CN117557346A CN 117557346 A CN117557346 A CN 117557346A CN 202410039603 A CN202410039603 A CN 202410039603A CN 117557346 A CN117557346 A CN 117557346A
- Authority
- CN
- China
- Prior art keywords
- page
- product
- individual
- individual consumer
- browsing
- 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
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 30
- 238000000034 method Methods 0.000 claims abstract description 25
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 17
- 230000006399 behavior Effects 0.000 claims description 56
- 230000008569 process Effects 0.000 claims description 15
- 238000012217 deletion Methods 0.000 claims description 12
- 230000037430 deletion Effects 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 12
- 230000003993 interaction Effects 0.000 claims description 9
- 229940127554 medical product Drugs 0.000 claims description 9
- 230000001186 cumulative effect Effects 0.000 claims description 5
- 238000010586 diagram Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000000295 complement effect Effects 0.000 claims description 3
- 238000003064 k means clustering Methods 0.000 claims description 3
- 238000011002 quantification Methods 0.000 claims description 3
- 238000011897 real-time detection Methods 0.000 claims description 3
- 101150113144 CARTPT gene Proteins 0.000 abstract 1
- 230000000875 corresponding effect Effects 0.000 description 14
- 238000005516 engineering process Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000007418 data mining Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- 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
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0641—Shopping interfaces
- G06Q30/0643—Graphical representation of items or shoppers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/958—Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0222—During e-commerce, i.e. online transactions
-
- 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
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0226—Incentive systems for frequent usage, e.g. frequent flyer miles programs or point systems
-
- 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
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0611—Request for offers or quotes
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Development Economics (AREA)
- Data Mining & Analysis (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the technical field of business decision analysis and discloses a full-link intelligent business decision analysis method based on dynamic consumption data. The method comprises the steps of collecting the browsing page condition and consumption data of individual consumers in real time, collecting the dynamic interests of the individual consumers, and quantifying the interests of the individual consumers; furthermore, the invention clusters past consumption data of individual consumers through a clustering algorithm, and realizes the presumption evolution of the consumption data of group consumers by constructing a group consumer co-purchasing network; finally, the shopping situation of the consumer is further analyzed through analysis of the shopping cart sequences of the individual consumers; and meanwhile, a proper intelligent business decision is made according to the real-time collected interests of the individual consumers and the shopping situation analysis of the consumers.
Description
Technical Field
The invention relates to the technical field of business decision analysis, in particular to a full-link intelligent business decision analysis method based on dynamic consumption data.
Background
With the continuous development of data mining technology, analysis and research on consumer historical data have become important reference bases for business decision-making by relevant management departments of merchants. However, with the continuous development of data mining technology and the deep mining of consumer consumption data, the problem of information overload is also caused, the information quantity in mass media is far higher than the information quantity which can be consumed by the audience, and a large amount of irrelevant redundant information seriously interferes with the accurate resolution and correct selection of relatively useful information by the audience.
With the upgrade of consumption demands, whether traditional physical retail or conventional internet online retail cannot meet the shopping demands of people, so with the transition of information technology, the full-link intelligent business decision analysis method based on dynamic consumption data has the characteristics of high efficiency, accuracy and the like and is considered as an effective tool for promoting the purchase conversion rate.
In the system for fully integrating capturing and analyzing business information to generate predictive decisions and simulations of patent application CN109478296a, business related data that has a large impact on business decisions is collected by a crawler, and deep mining of business data is not achieved, which has a great limitation.
Disclosure of Invention
(one) solving the technical problems
Aiming at the defects of the prior art, the invention provides a full-link intelligent business decision analysis method based on dynamic consumption data, which has the advantages of high efficiency, accuracy and the like and solves the problem of information overload.
(II) technical scheme
In order to solve the technical problem of information overload, the invention provides the following technical scheme:
the embodiment discloses a full-link intelligent business decision analysis method based on dynamic consumption data, which specifically comprises the following steps:
s1, collecting dynamic interests of individual consumers based on real-time browsing conditions and consumption data of the individual consumers to obtain dynamic interest data of the individual consumers;
further, the collecting the dynamic interests of the individual consumers based on the real-time browsing conditions and the consumption data of the individual consumers comprises:
s11, carrying out structural representation on browsing behaviors of individual consumers, and carrying out quantitative indexes on the browsing behaviors of the individual consumers;
further, quantitative indexes of browsing behaviors of individual consumers from three aspects of selecting interestingness, accessing time preference and page interestingness are set;
further, setting U as a set of all page categories in the system, and H as a set of sub-paths of all browsing page categories;
further, setting an individual consumer browsing set H as belonging to the set H, and setting a limited sequence group s, s epsilon H, wherein s represents a sequence formed by the individual consumer browsing page set; setting the ith browsing page in the limited sequence group s as the ith position in the limited sequence group s, setting n choices of individual consumers when selecting the next page category, wherein n represents that n different page categories exist;
further, the specific step of individual consumer selection of interest level index quantification includes:
after the individual consumer leaves the page category q, browsing the page with higher relative selection times in the page category in the next step, and setting the page as the page with high interest level selected by the individual consumer;
further, setting the selection interest degree of the ith page as the probability of the individual consumer transferring the browsing page based on the collected individual consumer data;
wherein N is i Frequency of individual consumers leaving page q to next step i-th page is represented, N k The frequency of the kth page is indicated,representing the probability that an individual consumer leaves page q to the next step, page i;
further, the specific step of quantifying the access time preference index includes:
quantifying the individual consumer's access time preference for pages based on the individual consumer's duration in the q-th page:
time (q) represents the duration of the individual consumer on the q-th page, T i Represents the time of the start of the ith page, T q A time representing the start of the q-th page;
further, setting page relative browsing time rate to accurately quantify access time preference degree;
wherein m represents a total of m pages visited by an individual consumer in one use, timeratio (i) represents a relative browsing Time rate of the individual consumer in the m pages visited by the individual consumer, and Time (i) represents a duration of the individual consumer in the ith page;
further, the specific step of quantifying the page interestingness index includes:
quantifying page interestingness based on individual consumer page access rates;
wherein, visit (i) represents the number of accesses of individual consumers to the ith page type in one use,representing page interest of individual consumers in the ith page type;
s12, constructing an individual consumer dynamic interest model based on the individual consumer access data and the set individual consumer browsing behavior representation index;
further, based on the three quantized interest variables in the S11, clustering modeling is carried out on the web page through a K-Means clustering algorithm;
data set x= { X of N browsing pages of one V dimension is set 1 ,x 2 ,x 3 ,...,x N }∈ϴ N×V Setting each page to contain V interest variables, and dividing N browsing pages into K surrounding cluster centers C= { C through a clustering algorithm 1 ,c 2 ,...,c k Non-empty disjoint clusters s= { S 1 ,s 2 ,...,s k The specific steps include:
where W is the sum of the errors of all pages within K clusters from the respective cluster center, each cluster center c uniquely represents one cluster S, and d (x, c) represents the distance of page x from cluster center c.
Further, verifying the clustered clusters by setting profile coefficients;
for a certain page x, the contour coefficients are:
wherein a (x i ) Representing the calculation x i Average distance from other pages in the cluster, b (x i ) Represents x i The minimum of the average distance to all pages in other clusters,representing the contour coefficients of page xi;
s is obtained from the contour coefficient formula i At [ -1,1]When s approaches-1, it represents page x i The more toward other clusters than the own cluster, x i Assigned to the wrong cluster;
if s i When the trend is 0, the page data set X does not have any cluster, and the pages in the cluster are randomly allocated;
if s i Trend towards 1; indicate page x i Away from other clusters, towards the cluster in which it is located;
further, for the set K cluster clusters of the page data set X, the average profile coefficient thereof is set as:
where N represents the number of pages in the dataset.
S2, clustering the dynamic interest data of the individual consumers based on a clustering algorithm;
further, a sample set of past consumption data of individual consumers is defined as an object set Q, wherein each consumption data is set as an object, and any one cluster represents a group PQ (i=1, 2,., k) a set of similar objects, k representing the total number of clusters;
s21, randomly selecting an initial clustering center;
selecting K objects from the data set as initial cluster centers based on the following formula;
wherein,representing an initial cluster center, f representing the feature number of the parameter cluster, max representing a set composed of the maximum value of each feature, and min representing a set composed of the minimum value of each feature;
s22, associating each object with the cluster center point closest to the object based on a distance algorithm, and gathering all points associated with the same cluster center into a cluster;
distance algorithm:
wherein d P Representing the distance of the object in the P cluster from the center of the original cluster,Q i which represents the i-th cluster and,representing an initial cluster center;
s23, calculating attribute characteristic average values in the clusters to form a new average value object, and taking the object as a next clustering center;
s24, repeating the steps until the obtained clustering center is not changed, and stopping iteration;
s3, presuming evolution group consumers based on a random graph model;
s31, constructing a group consumer co-purchase network;
setting a common purchase network g= { N, E }, where N is a set of product nodes and E is a set of common purchase links interconnecting them;
further, by an n×n-dimensional adjacency matrix y= { y ij I, j=1, 2,..n represents a co-purchase network g= { N, E };
when y is ij =1, indicating that a link exists from product i to product j, when y ij =0, indicating that no link exists from product i to product j;
setting the weight of each product in the adjacency matrix by means of a binary group based on the price of the product, wherein 1 in the binary group (1, z) represents the existence of a link from product i to product j, and z is the price of product i;
s32, setting the topology attribute of the network G according to the link sequence of the co-purchase network G= { N, E }, and presuming the consumption concept of the evolution consumer based on the topology attribute in the network G;
product yield: representing a co-purchase network y= { y ij The number of the link edges is purchased jointly from the outside of the starting product i;
wherein,representing the product output of the product i;
product input degree: representing a co-purchase network y= { y ij The number of inward co-purchase link edges in the product i from other product streams;
wherein,representing the product input degree of the product i;
outward strength of the product: representing a weighted adjacency matrix w= { W ij A cumulative purchase amount of the outward co-purchase link from the origin product i in the co-purchase network;
wherein,representing the outward strength of product i;
product internal guiding strength: representing a weighted adjacency matrix w= { W ij Common purchase network represented by }, a plurality of common purchase networksCumulative purchase amount of inward co-purchase links from other products to product i;
wherein,product internal lead strength of product i;
further, based on a continuous power law modelFitting->，/>，/>，；
Further, based on the fitting result, a product with high product income represents that the product has high popularity, and the product is a universal product, so that more flow can be brought to a product page of the product, and the product can be sold together with a large number of other commodities;
further, products with high product yields indicate that individual consumers can directly jump to more other products from the detail web pages of the products, so that complementary relations exist between the products and the other products;
s4, carrying out decision modeling according to the sequence of the dynamic electronic shopping carts of the individual consumers;
setting interaction options of individual consumers in one shopping browse, including clicking, collecting, purchasing, deleting and purchasing five interaction behaviors of Shop epsilon { click, tag, add, remove, pubose };
further, setting a shopping browsing threshold value accessed by the individual consumer once, when the individual consumer starts browsing entering an e-commerce website in one shopping browsing process, if the time after one interaction occurs exceeds 30 minutes, the individual consumer does not send any webpage request, and the completion of the shopping browsing process is marked; the next web page request will be marked as the initial point of the new shopping browsing process;
further, in the shopping browsing process of one-time access of the individual consumer, constructing a time sequence product path diagram based on the browsing behavior of the individual consumer;
further, setting the weight of the individual consumer behavior, setting deletion < click < collection < purchase < payment, setting the scoring values of five behaviors to be [ -1, 2,3,4] based on the weight of the individual consumer behavior, and adding corresponding points to the corresponding products when the individual consumer has the five behaviors in the shopping browsing process of one visit;
further, set upScoring individual consumer clicks, num is the number of corresponding actions of the product,for the number of clicks, +.>For the reward factor of the corresponding click behavior +.>Is a click-to-active factor;
=/>×/>×/>
setting upScore for individual consumer collection,/->For the reward factor of the corresponding collection behavior +.>Add purchase score for individual consumer, +.>For the reward factor corresponding to the purchasing behavior +.>Pay a score for individual consumers->For a reward factor corresponding to a payment behaviour +.>Deletion of scores for individual consumers, < >>To correspond to the bonus factor of the deletion behavior,for the number of times of product collection->For the number of times of product purchase->Pay for product number>Product deletion times;
=/>×/>
=/>×/>
=/>×/>
=/>×/>
collecting and quantifying the behaviors of individual consumers on the medical product, setting different weights for different behaviors based on different influence degrees of the five behaviors on the final score, and finally quantifying the score of the medical product based on the weights;
Ru,i=×/>+/>×/>+/>×/>+/>×/>+/>×/>
+/>+/>+/>+/>=1
wherein R is u,i Scoring individual consumer u after quantifying medical product i,as a weight factor for the click-through,weighting factors for collections，/>For weighting factors purchased additionally, +.>Weight factor for payment->Is a deleted weight factor.
S5, realizing decision analysis of the intelligent business based on real-time detection of the intention of the consumer;
and selecting L commodities before scoring and L commodities before the interestingness according to the scoring of each product and the interestingness of the consumer for popularization.
(III) beneficial effects
Compared with the prior art, the invention provides a full-link intelligent business decision analysis method based on dynamic consumption data, which has the following beneficial effects:
according to the method, the dynamic interests of the individual consumers are dynamically analyzed in real time in a mode of collecting the page browsing conditions and consumption data of the individual consumers in real time and quantifying the collected data, so that the accuracy, the effectiveness and the reliability of the intelligent business decision analysis method are improved.
According to the invention, the analysis of the consumption data of the individual consumers is realized by clustering the past consumption data of the individual consumers, and the consumption data of the individual consumers is quantitatively classified by continuously iterating the clustering method, so that the accuracy of the intelligent business decision analysis method is improved.
The invention improves the relation between consumption data of individual consumers by establishing a random graph model to infer evolution group consumers, so that the intelligent business decision analysis method is more accurate.
The invention reasonably judges the consumption trend of the individual consumer through the analysis of the sequence of the dynamic electronic shopping carts of the individual consumer, and improves the accuracy of the intelligent business decision analysis method.
Drawings
FIG. 1 is a schematic diagram of the flow configuration of the intelligent business decision making method of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The embodiment discloses a full-link intelligent business decision analysis method based on dynamic consumption data, which specifically comprises the following steps:
s1, collecting dynamic interests of individual consumers based on real-time browsing conditions and consumption data of the individual consumers to obtain dynamic interest data of the individual consumers;
further, the collecting the dynamic interests of the individual consumers based on the real-time browsing conditions and the consumption data of the individual consumers comprises:
s11, carrying out structural representation on browsing behaviors of individual consumers, and carrying out quantitative indexes on the browsing behaviors of the individual consumers;
further, quantitative indexes of browsing behaviors of individual consumers from three aspects of selecting interestingness, accessing time preference and page interestingness are set;
further, setting U as a set of all page categories in the system, and H as a set of sub-paths of all browsing page categories;
further, setting an individual consumer browsing set H as belonging to the set H, and setting a limited sequence group s, s epsilon H, wherein s represents a sequence formed by the individual consumer browsing page set; setting the ith browsing page in the limited sequence group s as the ith position in the limited sequence group s, setting n choices of individual consumers when selecting the next page category, wherein n represents that n different page categories exist;
further, the specific step of individual consumer selection of interest level index quantification includes:
after the individual consumer leaves the page category q, browsing the page with higher relative selection times in the page category in the next step, and setting the page as the page with high interest level selected by the individual consumer;
further, setting the selection interest degree of the ith page as the probability of the individual consumer transferring the browsing page based on the collected individual consumer data;
wherein N is i Frequency of individual consumers leaving page q to next step i-th page is represented, N k The frequency of the kth page is indicated,representing the probability that an individual consumer leaves page q to the next step, page i;
further, the specific step of quantifying the access time preference index includes:
quantifying the individual consumer's access time preference for pages based on the individual consumer's duration in the q-th page:
time (q) represents the duration of the individual consumer on the q-th page, T i Represents the time of the start of the ith page, T q A time representing the start of the q-th page;
further, setting page relative browsing time rate to accurately quantify access time preference degree;
wherein m represents a total of m pages visited by an individual consumer in one use, timeratio (i) represents a relative browsing Time rate of the individual consumer for the ith page among the m pages visited by the individual consumer, and Time (i) represents a duration of the individual consumer in the ith page;
further, the specific step of quantifying the page interestingness index includes:
quantifying page interestingness based on individual consumer page access rates;
wherein, visit (i) represents the number of accesses of individual consumers to the ith page type in one use,representing page interest of individual consumers in the ith page type;
s12, constructing an individual consumer dynamic interest model based on the individual consumer access data and the set individual consumer browsing behavior representation index;
further, based on the three quantized interest variables in the S11, clustering modeling is carried out on the web page through a K-Means clustering algorithm;
data set x= { X of N browsing pages of one V dimension is set 1 ,x 2 ,x 3 ,...,x N }∈ϴ N×V Setting each page to contain V interest variables, and dividing N browsing pages into K surrounding cluster centers C= { C through a clustering algorithm 1 ,c 2 ,...,c k Non-empty disjoint clusters s= { S 1 ,s 2 ,...,s k The specific steps include:
where W is the sum of the errors of all pages within K clusters from the respective cluster center, each cluster center c uniquely represents one cluster S, and d (x, c) represents the distance of page x from cluster center c.
Further, verifying the clustered clusters by setting profile coefficients;
for a certain page x, the contour coefficients are:
wherein a (x i ) Representing the calculation x i Average distance from other pages in the cluster, b (x i ) Represents x i The minimum of the average distance to all pages in other clusters,representing page x i Is a contour coefficient of (2);
s is obtained from the contour coefficient formula i At [ -1,1]When s approaches-1, it represents page x i The more toward other clusters than the own cluster, x i Assigned to the wrong cluster;
if s i When the trend is 0, the page data set X does not have any cluster, and the pages in the cluster are randomly allocated;
if s i Trend towards 1; indicate page x i Away from other clusters, towards the cluster in which it is located;
further, for the set K cluster clusters of the page data set X, the average profile coefficient thereof is set as:
where N represents the number of pages in the dataset.
S2, clustering the dynamic interest data of the individual consumers based on a clustering algorithm;
further, a sample set of past consumption data of individual consumers is defined as an object set Q, wherein each consumption data is set as an object, and any one cluster represents a group PQ (i=1, 2,., k) a set of similar objects, k representing the total number of clusters;
s21, randomly selecting an initial clustering center;
selecting K objects from the data set as initial cluster centers based on the following formula;
wherein,representing an initial cluster center, f representing the feature number of the parameter cluster, max representing a set composed of the maximum value of each feature, and min representing a set composed of the minimum value of each feature;
s22, associating each object with the cluster center point closest to the object based on a distance algorithm, and gathering all points associated with the same cluster center into a cluster;
distance algorithm:
wherein d P Representing the distance of the object in the P cluster from the center of the original cluster,Q i which represents the i-th cluster and,representing an initial cluster center;
s23, calculating attribute characteristic average values in the clusters to form a new average value object, and taking the object as a next clustering center;
s24, repeating the steps until the obtained clustering center is not changed, and stopping iteration;
s3, presuming evolution group consumers based on a random graph model;
s31, constructing a group consumer co-purchase network;
setting a common purchase network g= { N, E }, where N is a set of product nodes and E is a set of common purchase links interconnecting them;
further, by an n×n-dimensional adjacency matrix y= { y ij I, j=1, 2,..n represents a co-purchase network g= { N, E };
when y is ij =1, indicating that a link exists from product i to product j, when y ij =0, indicating that no link exists from product i to product j;
setting the weight of each product in the adjacency matrix by means of a binary group based on the price of the product, wherein 1 in the binary group (1, z) represents the existence of a link from product i to product j, and z is the price of product i;
s32, setting the topology attribute of the network G according to the link sequence of the co-purchase network G= { N, E }, and presuming the consumption concept of the evolution consumer based on the topology attribute in the network G;
product yield: representing a co-purchase network y= { y ij The number of the link edges is purchased jointly from the outside of the starting product i;
wherein,representing the product output of the product i;
product input degree: representing a co-purchase network y= { y ij The number of inward co-purchase link edges in the product i from other product streams;
wherein,representing the product input degree of the product i;
outward strength of the product: representing a weighted adjacency matrix w= { W ij A cumulative purchase amount of the outward co-purchase link from the origin product i in the co-purchase network;
wherein,representing the outward strength of product i;
product internal guiding strength: representing a weighted adjacency matrix w= { W ij Other product streams in co-purchase network represented by }Cumulative purchase amount of inward co-purchase links to product i;
wherein,product internal lead strength of product i;
further, based on a continuous power law modelFitting->，/>，/>，；
Further, based on the fitting result, a product with high product income represents that the product has high popularity, and the product is a universal product, so that more flow can be brought to a product page of the product, and the product can be sold together with a large number of other commodities;
further, products with high product yields indicate that individual consumers can directly jump to more other products from the detail web pages of the products, so that complementary relations exist between the products and the other products;
s4, carrying out decision modeling according to the sequence of the dynamic electronic shopping carts of the individual consumers;
setting interaction options of individual consumers in one shopping browse, including clicking, collecting, purchasing, deleting and purchasing five interaction behaviors of Shop epsilon { click, tag, add, remove, pubose };
further, setting a shopping browsing threshold value accessed by the individual consumer once, when the individual consumer starts browsing entering an e-commerce website in one shopping browsing process, if the time after one interaction occurs exceeds 30 minutes, the individual consumer does not send any webpage request, and the completion of the shopping browsing process is marked; the next web page request will be marked as the initial point of the new shopping browsing process;
further, in the shopping browsing process of one-time access of the individual consumer, constructing a time sequence product path diagram based on the browsing behavior of the individual consumer;
further, setting the weight of the individual consumer behavior, setting deletion < click < collection < purchase < payment, setting the scoring values of five behaviors to be [ -1, 2,3,4] based on the weight of the individual consumer behavior, and adding corresponding points to the corresponding products when the individual consumer has the five behaviors in the shopping browsing process of one visit;
further, set upScoring individual consumer clicks, num is the number of corresponding actions of the product,for the number of clicks, +.>For the reward factor of the corresponding click behavior +.>Is a click-to-active factor;
=/>×/>×/>
setting upScore for individual consumer collection,/->For the reward factor of the corresponding collection behavior +.>Add purchase score for individual consumer, +.>For the reward factor corresponding to the purchasing behavior +.>Pay a score for individual consumers->For a reward factor corresponding to a payment behaviour +.>Deletion of scores for individual consumers, < >>To correspond to the bonus factor of the deletion behavior,for the number of times of product collection->For the number of times of product purchase->Pay for product number>Product deletion times;
=/>×/>
=/>×/>
=/>×/>
=/>×/>
collecting and quantifying the behaviors of individual consumers on the medical product, setting different weights for different behaviors based on different influence degrees of the five behaviors on the final score, and finally quantifying the score of the medical product based on the weights;
Ru,i=×/>+/>×/>+/>×/>+/>×/>+/>×/>
+/>+/>+/>+/>=1
wherein R is u,i Scoring individual consumer u after quantifying medical product i,as a weight factor for the click-through,for the weight factor of collection->For weighting factors purchased additionally, +.>Weight factor for payment->Is a deleted weight factor.
S5, realizing decision analysis of the intelligent business based on real-time detection of the intention of the consumer;
and selecting L commodities before scoring and L commodities before the interestingness according to the scoring of each product and the interestingness of the consumer for popularization.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (9)
1. The full-link intelligent business decision analysis method based on the dynamic consumption data is characterized by comprising the following steps of:
s1, collecting dynamic interests of individual consumers based on real-time browsing conditions and consumption data of the individual consumers to obtain dynamic interest data of the individual consumers;
s2, clustering the dynamic interest data of the individual consumers based on a clustering algorithm;
s3, presuming evolution group consumers based on a random graph model;
s4, carrying out decision modeling according to the sequence of the dynamic electronic shopping carts of the individual consumers;
s5, realizing decision analysis of the intelligent business based on real-time detection of the intention of the consumer.
2. The method for dynamic consumption data-based all-link intelligent business decision analysis according to claim 1, wherein the individual consumer dynamic interest collection based on the individual consumer real-time browsing situation and the consumption data comprises:
s11, carrying out structural representation on browsing behaviors of individual consumers, and carrying out quantitative indexes on the browsing behaviors of the individual consumers;
s12, constructing an individual consumer dynamic interest model based on the individual consumer access data and the set individual consumer browsing behavior representation index.
3. The full-link intelligent business decision analysis method based on dynamic consumption data according to claim 2, wherein quantitative indexes of browsing behaviors of individual consumers are set from three aspects of selection interestingness, access time preference and page interestingness;
setting U as a set of all page categories in the system, and H as a set of sub-paths of all browsing page categories;
setting an individual consumer browsing set H as a set H, and setting a finite sequence group s, s E H, s representing a sequence formed by the individual consumer browsing page set;
setting the ith browsing page in the limited sequence group s as the ith position in the limited sequence group s, setting n choices of individual consumers when selecting the next page category, wherein n represents that n different page categories exist;
the specific steps of the individual consumer selecting the interestingness index quantification include:
after the individual consumer leaves the page category q, browsing the page with higher relative selection times in the page category in the next step, and setting the page as the page with high interest level selected by the individual consumer;
setting the selection interest degree of the ith page as the probability of transferring the browsing page for the individual consumer based on the collected individual consumer data;
；
wherein N is i Frequency of individual consumers leaving page q to next step i-th page is represented, N k The frequency of the kth page is indicated,representing the probability that an individual consumer leaves page q to the next step, page i;
further, the specific step of quantifying the access time preference index includes:
quantifying the individual consumer's access time preference for pages based on the individual consumer's duration in the q-th page:
；
time (q) represents the duration of the individual consumer on the q-th page, T i Represents the time of the start of the ith page, T q A time representing the start of the q-th page;
setting page relative browsing time rate to accurately quantify access time preference degree;
；
wherein m represents a total of m pages visited by an individual consumer in one use, timeratio (i) represents a relative browsing Time rate of the individual consumer in the m pages visited by the individual consumer, and Time (i) represents a duration of the individual consumer in the ith page;
the specific steps of quantifying the page interest index include:
quantifying page interestingness based on individual consumer page access rates;
；
wherein, visit (i) represents the number of accesses of individual consumers to the ith page type in one use,representing the page interest level of the individual consumer in the ith page type.
4. The method for full link intelligent business decision analysis based on dynamic consumption data according to claim 2, wherein the constructing an individual consumer dynamic interest model based on individual consumer access data and set individual consumer browsing behavior representation indicators comprises:
based on the three quantized interest variables in the S11, carrying out clustering modeling on the webpage through a K-Means clustering algorithm;
data set x= { X of N browsing pages of one V dimension is set 1 ,x 2 ,x 3 ,...,x N }∈ϴ N×V Setting each page to contain V interest variables, and dividing N browsing pages into K surrounding cluster centers C= { C through a clustering algorithm 1 ,c 2 ,...,c k Non-empty disjoint clusters s= { S 1 ,s 2 ,...,s k The specific steps include:
；
wherein W is k Is the sum of the errors of all pages within K clusters from the respective cluster center, each cluster center c k Uniquely representing a cluster s k ，d(x i ,c k ) Representing page x i To cluster center c k Is a distance of (2);
verifying the clustered clusters by setting profile coefficients;
for a certain page x, the contour coefficients are:
；
wherein a (x i ) Representing the calculation x i Average distance from other pages in the cluster, b (x i ) Represents x i Minimum average distance to all pages in other clustersThe value of the sum of the values,representing the contour coefficients of page xi;
s is obtained from the contour coefficient formula i At [ -1,1]When s approaches-1, it represents page x i The more toward other clusters than the own cluster, x i Assigned to the wrong cluster;
if s i When the trend is 0, the page data set X does not have any cluster, and the pages in the cluster are randomly allocated;
if s i Trend towards 1; indicate page x i Away from other clusters, towards the cluster in which it is located;
for the set K cluster clusters of the page data set X, the average profile coefficient thereof is set as:
；
where N represents the number of pages in the dataset.
5. The full-link intelligent business decision analysis method based on dynamic consumption data according to claim 1, wherein the clustering of individual consumer past consumption data based on a clustering algorithm comprises the following steps:
defining a sample set of past consumption data of an individual consumer as an object set Q, wherein each consumption data is set as an object, and any one cluster represents a group PQ (i=1, 2,., k) a set of similar objects, k representing the total number of clusters;
s21, randomly selecting an initial clustering center;
selecting K objects from the data set as initial cluster centers based on the following formula;
；
wherein,representing an initial cluster center, f representing the feature number of the parameter cluster, max representing a set composed of the maximum value of each feature, and min representing a set composed of the minimum value of each feature;
s22, associating each object with the cluster center point closest to the object based on a distance algorithm, and gathering all points associated with the same cluster center into a cluster;
distance algorithm:
；
wherein d P Representing the distance of the object in the P cluster from the center of the original cluster,Q i which represents the i-th cluster and,representing an initial cluster center;
s23, calculating attribute characteristic average values in the clusters to form a new average value object, and taking the object as a next clustering center;
s24, repeating the steps until the obtained clustering center is not changed, and stopping iteration.
6. The method for dynamic consumption data-based all-link intelligent business decision analysis according to claim 1, wherein the step of estimating evolving group consumers based on the stochastic graph model comprises the steps of:
s31, constructing a group consumer co-purchase network;
s32, setting the topology attribute of the network G according to the link sequence of the co-purchase network G= { N, E }, and presuming the consumption concept of the evolving consumer based on the topology attribute in the network G.
7. The method for dynamic consumption data-based all-link intelligent business decision analysis according to claim 6, wherein said constructing a group of consumers co-purchase network comprises:
setting a common purchase network g= { N, E }, where N is a set of product nodes and E is a set of common purchase links E interconnecting them;
by an n x n-dimensional adjacency matrix y= { y ij I, j=1, 2,..n represents a co-purchase network g= { N, E };
when y is ij =1, indicating that a link exists from product i to product j, when y ij =0, indicating that no link exists from product i to product j;
based on the price of the product, the weight of each product in the adjacency matrix is set by a binary group, where 1 in the binary group (1, z) indicates the existence of a link from product i to product j, and z is the price of product i.
8. The method of claim 6, wherein the estimating the consumer concept of the evolving consumer based on topology attributes in the network G comprises:
product yield: representing a co-purchase network y= { y ij The number of the link edges is purchased jointly from the outside of the starting product i;
；
wherein,representing the product output of the product i;
product input degree: representing a co-purchase network y= { y ij The number of inward co-purchase link edges in the product i from other product streams;
；
wherein,representing the product input degree of the product i;
outward strength of the product: representing a weighted adjacency matrix w= { W ij A cumulative purchase amount of the outward co-purchase link from the origin product i in the co-purchase network;
；
wherein,representing the outward strength of product i;
product internal guiding strength: representing a weighted adjacency matrix w= { W ij Cumulative purchase amount of inward co-purchase links in the co-purchase network from other product streams to product i;
；
wherein,product internal lead strength of product i;
based on continuous power law modelFitting->，/>，/>，/>；
Based on the fitting result, products with high product income represent high popularity, and represent that the products are general products, and can bring more flow to the product pages of the products so as to be sold together with a large number of other commodities;
products with high product output degree indicate that individual consumers can directly jump to more other products from the detail web pages of the products, and the complementary relationship between the products and the other products is indicated.
9. The method of claim 1, wherein the modeling decision from a sequence of individual consumer dynamic electronic shopping carts comprises:
setting interaction options of individual consumers in one shopping browse, including clicking, collecting, purchasing, deleting and purchasing five interaction behaviors of Shop epsilon { click, tag, add, remove, pubose };
setting a shopping browsing threshold value which is accessed by an individual consumer once, when the individual consumer starts browsing entering an e-commerce website in one shopping browsing process, if the time after one interaction occurs exceeds 30 minutes, the individual consumer does not send any webpage request, and the completion of the shopping browsing process is marked; the next web page request will be marked as the initial point of the new shopping browsing process;
in the shopping browsing process of one-time access of the individual consumer, constructing a time sequence product path diagram based on the browsing behavior of the individual consumer;
setting the weight of the individual consumer behaviors, setting deletion < click < collection < purchase < payment, setting the scoring values of five behaviors to be [ -1, 2,3,4] based on the weight of the individual consumer behaviors, and adding corresponding points to corresponding products when the individual consumer has the five behaviors in the shopping browsing process of one visit;
setting upClicking for individual consumersIs the number of times the product corresponds to the behavior, +.>In order to determine the number of clicks,for the reward factor of the corresponding click behavior +.>Is a click-to-active factor;
；
=/>×/>×/>；
setting upScore for individual consumer collection,/->For the reward factor of the corresponding collection behavior +.>Add purchase score for individual consumer, +.>For the reward factor corresponding to the purchasing behavior +.>Pay a score for individual consumers->For a reward factor corresponding to a payment behaviour +.>Deletion of scores for individual consumers, < >>For a bonus factor corresponding to a deletion behavior +.>For the number of times of product collection->For the number of times of product purchase->Pay for product number>Product deletion times;
=/>×/>；
=/>×/>；
=/>×/>；
=/>×/>；
collecting and quantifying the behaviors of individual consumers on the medical product, setting different weights for different behaviors based on different influence degrees of the five behaviors on the final score, and finally quantifying the score of the medical product based on the weights;
Ru,i=×/>+/>×/>+/>×/>+/>×/>+/>×/>；
+/>+/>+/>+/>=1；
wherein R is u,i Scoring individual consumer u after quantifying medical product i,for click weight factor, ++>For the weight factor of collection->For weighting factors purchased additionally, +.>Weight factor for payment->Is a deleted weight factor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410039603.6A CN117557346B (en) | 2024-01-11 | 2024-01-11 | Full-link intelligent business decision analysis method based on dynamic consumption data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410039603.6A CN117557346B (en) | 2024-01-11 | 2024-01-11 | Full-link intelligent business decision analysis method based on dynamic consumption data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117557346A true CN117557346A (en) | 2024-02-13 |
| CN117557346B CN117557346B (en) | 2024-04-02 |
Family
ID=89815089
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410039603.6A Active CN117557346B (en) | 2024-01-11 | 2024-01-11 | Full-link intelligent business decision analysis method based on dynamic consumption data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117557346B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110246300A1 (en) * | 2009-12-15 | 2011-10-06 | Yarvis Mark D | Techniques to determe when an internet user is in-market for a specific product and determining general shopping preferences and habits of internet users |
| CN106548369A (en) * | 2016-10-14 | 2017-03-29 | 五邑大学 | Customers in E-commerce intension recognizing method based on ant group algorithm |
| CN106708821A (en) * | 2015-07-21 | 2017-05-24 | 广州市本真网络科技有限公司 | User personalized shopping behavior-based commodity recommendation method |
| CN107633430A (en) * | 2017-09-20 | 2018-01-26 | 哈尔滨工业大学 | A kind of Method of Commodity Recommendation based on community of colony |
| CN107944913A (en) * | 2017-11-21 | 2018-04-20 | 重庆邮电大学 | High potential user's purchase intention Forecasting Methodology based on big data user behavior analysis |
| CN112231548A (en) * | 2019-07-15 | 2021-01-15 | 北京京东尚科信息技术有限公司 | User login behavior analysis method, device and system and storage medium |
| CN112651805A (en) * | 2020-12-30 | 2021-04-13 | 广东各有所爱信息科技有限公司 | Commodity recommendation method and system for online shopping mall |
| CN116595255A (en) * | 2023-05-20 | 2023-08-15 | 车雨 | Big data analysis method and system for cloud service pushing |
-
2024
- 2024-01-11 CN CN202410039603.6A patent/CN117557346B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110246300A1 (en) * | 2009-12-15 | 2011-10-06 | Yarvis Mark D | Techniques to determe when an internet user is in-market for a specific product and determining general shopping preferences and habits of internet users |
| CN106708821A (en) * | 2015-07-21 | 2017-05-24 | 广州市本真网络科技有限公司 | User personalized shopping behavior-based commodity recommendation method |
| CN106548369A (en) * | 2016-10-14 | 2017-03-29 | 五邑大学 | Customers in E-commerce intension recognizing method based on ant group algorithm |
| CN107633430A (en) * | 2017-09-20 | 2018-01-26 | 哈尔滨工业大学 | A kind of Method of Commodity Recommendation based on community of colony |
| CN107944913A (en) * | 2017-11-21 | 2018-04-20 | 重庆邮电大学 | High potential user's purchase intention Forecasting Methodology based on big data user behavior analysis |
| CN112231548A (en) * | 2019-07-15 | 2021-01-15 | 北京京东尚科信息技术有限公司 | User login behavior analysis method, device and system and storage medium |
| CN112651805A (en) * | 2020-12-30 | 2021-04-13 | 广东各有所爱信息科技有限公司 | Commodity recommendation method and system for online shopping mall |
| CN116595255A (en) * | 2023-05-20 | 2023-08-15 | 车雨 | Big data analysis method and system for cloud service pushing |
Non-Patent Citations (4)
| Title |
|---|
| DONGYU LI等: ""Mining consumer\'s opinion target based on translation model and word representation"", 《2014 11TH INTERNATIONAL COMPUTER CONFERENCE ON WAVELET ACTIEV MEDIA TECHNOLOGY AND INFORMATION PROCESSING(ICCWAMTIP)》, 21 December 2014 (2014-12-21), pages 97 - 101, XP032755585, DOI: 10.1109/ICCWAMTIP.2014.7073369 * |
| 刘枚莲;刘同存;李小龙;: "基于用户兴趣特征提取的推荐算法研究", 计算机应用研究, no. 05, 15 May 2011 (2011-05-15), pages 70 - 73 * |
| 刘洪伟;高鸿铭;陈丽;詹明君;梁周扬;: "基于用户浏览行为的兴趣识别管理模型", 数据分析与知识发现, no. 02, 25 February 2018 (2018-02-25), pages 78 - 89 * |
| 郑天宇;吴爱华;: "基于变长马尔科夫模型的用户购物行为分析", 现代计算机(专业版), no. 21, 25 July 2016 (2016-07-25), pages 10 - 16 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117557346B (en) | 2024-04-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103377250B (en) | Top k based on neighborhood recommend method | |
| US8620746B2 (en) | Scoring quality of traffic to network sites | |
| CN108665311B (en) | Electric commercial user time-varying feature similarity calculation recommendation method based on deep neural network | |
| CN110956273A (en) | Credit scoring method and system integrating multiple machine learning models | |
| JP2004506959A (en) | Evaluation value prediction model when information is missing | |
| CN106447463A (en) | Commodity recommendation method based on Markov decision-making process model | |
| Fan et al. | Improved ML‐based technique for credit card scoring in Internet financial risk control | |
| CN114782076B (en) | Online mall consumption platform lottery integral exchange intelligent management method, system and computer storage medium | |
| JP2019215778A (en) | Service provision system, advertisement related service provision system, user side facility and user side advertisement facility | |
| CN108154311A (en) | Top-tier customer recognition methods and device based on random forest and decision tree | |
| JP2022548435A (en) | A Product Recommendation System Based on Practical High-utility Negative Array Rule Mining and Its Operation Method | |
| CN111882420A (en) | Generation method of response rate, marketing method, model training method and device | |
| CN116187808A (en) | Electric power package recommendation method based on virtual power plant user-package label portrait | |
| CN115829683A (en) | Power integration commodity recommendation method and system based on inverse reward learning optimization | |
| CN108647714A (en) | Acquisition methods, terminal device and the medium of negative label weight | |
| CN117495482B (en) | Secondhand mobile phone sales recommendation method and system based on user portrait | |
| CN112232388B (en) | Shopping intention key factor identification method based on ELM-RFE | |
| CN113450004A (en) | Power credit report generation method and device, electronic equipment and readable storage medium | |
| CN111612583B (en) | Personalized shopping guide system based on clustering | |
| CN117172825A (en) | Electronic commerce data analysis method and device based on big data and computing equipment | |
| Chou et al. | The RFM Model Analysis for VIP Customer: A case study of golf clothing brand | |
| CN117557346B (en) | Full-link intelligent business decision analysis method based on dynamic consumption data | |
| CN116452018A (en) | Visual image design system based on information acquisition | |
| CN115841345A (en) | Cross-border big data intelligent analysis method, system and storage medium | |
| CN115393098A (en) | Financing product information recommendation method and device |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Pan Dongxiao Inventor after: Wu Yucun Inventor after: Li Weidong Inventor after: Jia Shengzhong Inventor after: Liu Changlong Inventor after: Luo Kuan Inventor after: Li Yanwei Inventor after: Ma Xiaofeng Inventor after: Duan Wenxuan Inventor after: Gu Shiwen Inventor before: Pan Dongxiao Inventor before: Li Weidong Inventor before: Jia Shengzhong Inventor before: Liu Changlong Inventor before: Luo Kuan Inventor before: Li Yanwei Inventor before: Ma Xiaofeng Inventor before: Gu Shiwen Inventor before: Wu Yucun |
|
| CB03 | Change of inventor or designer information |