CN112116387A - Information prediction method based on causal relationship - Google Patents

Information prediction method based on causal relationship Download PDF

Info

Publication number
CN112116387A
CN112116387A CN202010972803.9A CN202010972803A CN112116387A CN 112116387 A CN112116387 A CN 112116387A CN 202010972803 A CN202010972803 A CN 202010972803A CN 112116387 A CN112116387 A CN 112116387A
Authority
CN
China
Prior art keywords
business
causal
causal relationship
data
prediction method
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.)
Pending
Application number
CN202010972803.9A
Other languages
Chinese (zh)
Inventor
刘立丰
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.)
Beijing Second Computing Information Technology Co ltd
Original Assignee
Beijing Second Computing Information Technology 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 Beijing Second Computing Information Technology Co ltd filed Critical Beijing Second Computing Information Technology Co ltd
Priority to CN202010972803.9A priority Critical patent/CN112116387A/en
Publication of CN112116387A publication Critical patent/CN112116387A/en
Pending legal-status Critical Current

Links

Images

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
    • G06Q30/02—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
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G06Q30/0203—Market surveys; Market polls

Landscapes

  • Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Engineering & Computer Science (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • Finance (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Data Mining & Analysis (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention provides an information prediction method based on causal relationship, belonging to the technical field of information processing and comprising the steps of collecting multi-source data from business behavior; taking multi-source data as independent variables and a business target as dependent variables, constructing a causal relationship model of the independent variables and the dependent variables, and establishing a relation between a business behavior and the business target; and selecting the business strategy according to the causal relationship model. The information prediction method based on the causal relationship can remove pseudo correlation caused by interference factors; the limitation problems of less information and small quantity of research data are solved, and global insight is formed; the causal relationship model generated in the method can highly restore the real market environment, output the result capable of practically guiding business decision, and directly assist the enterprise decision.

Description

Information prediction method based on causal relationship
Technical Field
The invention belongs to the technical field of information processing, and particularly relates to an information prediction method based on causal relationship.
Background
The Net Promoter Score (NPS) is an index for measuring possible recommendation of a product or service to others by a client, is a key index for measuring loyalty of the product client, is subject to the american benne consulting company, and aims at benign income and real growth of enterprises, information recommendation is performed on the basis of the NPS, so that the actual conditions of the client can be better met, the recommendation success rate is improved, and the user experience is improved. The existing process of collecting net recommendation values mostly collects user opinions by means of issuing questionnaires to users and analyzes the opinions by using a traditional statistical method. The user attitudes are collected through questionnaires, the invalidity is poor, all answers are the attitudes that users claim, the situation that visitors give unreal opinions inevitably exists, and the authenticity of the user attitudes and the causes of attitudes cannot be completely supported by the same source data is avoided.
The traditional method for analyzing the business influence factors mainly comprises the following steps: correlation analysis, regression analysis, causal analysis. The disadvantages of the conventional data analysis techniques: 1. the strong correlation is taken as a cause and effect, and a real reason cannot be found; 2. The traditional statistical method has more obvious limitation in the big data era; 3. the traditional statistical method has poor efficiency and effect in the aspects of simulation and optimization.
The traditional statistical method has certain bottlenecks in the aspects of causal analysis, simulation efficiency and optimization effect, and the traditional data analysis method is influenced by a large number of interference factors, only can see the correlation among target factors, cannot judge the real causal relationship, cannot influence the result from dependent variables, is easy to make an invalid business strategy, and has the limitations that the high fault-tolerant cost loses market opportunity and the like.
Disclosure of Invention
The invention aims to provide an information prediction method based on causal relationship, and aims to solve the technical problems that a traditional statistical method has certain bottlenecks in causal analysis, simulation efficiency and optimization effect, is influenced by interference factors, only can see the correlation among target factors, cannot judge the real causal relationship, cannot influence the result from dependent variables and is easy to make an invalid business strategy.
In order to achieve the purpose, the invention adopts the technical scheme that: provided is a causal relationship-based information prediction method, including:
collecting multi-source data from business activities;
taking the multi-source data as independent variables and the business target as dependent variables, constructing a causal relationship model of the independent variables and the dependent variables, and establishing a relation between the business behavior and the business target;
and selecting a business strategy according to the causal relationship model.
Further, collecting multi-source data from business activities includes collecting data from enterprise or/and third party collection research platforms.
Further, the construction of the causal relationship model comprises: determining a research problem; selecting factors contained in the problem, wherein the factors comprise independent variables and dependent variables, and constructing a structure cause and effect model; determining a causal relationship parameter; establishing a relation between the observation parameters and the causal model; and evaluating the identifiability of the average causal formula and determining the causal relationship of the multiple paths.
Further, selecting factors contained in the problem, wherein the factors include independent variables and dependent variables, and constructing a structural causal model includes:
determining an endogenous node: x ═ W1, W2, a, Y, where W1: basic miscellaneous elements, W2: non-observable confounding elements; a: independent variable, Y: a dependent variable; determining an exogenous node: u ═ U (U)W1,UW2，UY, UA)～P*Wherein, UW1,UW2，UY,UAHave correlations with W1, W2, A, Y, respectively, P*A pointer variable representing a data type; constructing a structural equation F:
W1＝fW1(UW1)
W2→fW2(W1,UW2)
A→fA(W1,W2,UA)
Y→fY(W1,W2,A,UY)。
further, determining the causal relationship parameter includes:
ACE (Y1) -E (Y0) ═ P (Y1 ═ 1) -P (Y0 ═ 0), where Y1 represents the business objective obtained using the argument; y0 denotes the business objective that is obtained without using arguments.
Further, the associating of the observation parameters with the causal model comprises:
u Observation＝(W1,W2,A,Y)～P
u ACE＝E[Y1]-E[Y0]。
further, the multi-source data includes: one or more of subjective factors of the consumer, objective factors of the consumer, product sales, user behavior data, survey data of satisfaction, investments in various media channels, and characteristics of historical promotional activities.
Further, the business objectives include: product pricing, product profits, net recommendation values for non-investigated users, business outcome indicators, promotional strategies, purchase outcomes.
The information prediction method based on the causal relationship has the advantages that: compared with the prior art, the information prediction method based on the causal relationship can remove pseudo correlation caused by interference factors; the limitation problems of less information and small quantity of research data are solved, and global insight is formed; the causal relationship model generated in the method can highly restore the real market environment, so that the data relationship reliability is improved, a nest mortar of the traditional data analysis technology is eliminated, data islands of different layers are communicated, the result capable of practically guiding business decision is output, and enterprise decision is directly assisted.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed for the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a directed acyclic graph provided in a causal information prediction method according to an embodiment of the present invention;
fig. 2 is a pricing model with different dimensions provided in a causal information prediction method according to embodiment 2 of the present invention;
FIG. 3 is a first flowchart of a causal information prediction method according to an embodiment of the present invention;
fig. 4 is a second flowchart of a causal information prediction method according to an embodiment of the present invention.
Detailed Description
In order to make the technical problems, technical solutions and advantageous effects to be solved by the present invention more clearly apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Referring to fig. 1 to 4, a causal information prediction method according to the present invention will now be described. The information prediction method based on the causal relationship comprises
S1, collecting multi-source data from the business behaviors;
the implementation of this step can be:
the collection of research data and big data is performed by using free data provided by enterprises or by a third-party collection information platform (such as second calculation).
Illustratively, NPS and other business targets are analyzed by combining behavior big data (including APP use frequency and duration) used by a user APP and investigation (questionnaire investigation on preference attitude of the user for a certain product or service).
S2, taking the multi-source data as independent variables and the business target as dependent variables, constructing a causal relationship model of the independent variables and the dependent variables, and establishing a relation between the business behavior and the business target;
the implementation of this step can be:
s2.1, determining a research problem;
s2.2, selecting factors contained in the problem, wherein the factors comprise independent variables and dependent variables, and constructing a structural causal model;
the implementation of this step can be:
determining an endogenous node: x ═ W1, W2, a, Y, where W1: basic miscellaneous elements (variables having a correlation with Y include use experience, goodness, stock-keeping rate, brand impression, etc.), W2: non-observable confounding elements; a: independent variable, Y: a dependent variable;
determining an exogenous node: u ═ U (U)W1,UW2，UY,UA)～P*Wherein, UW1,UW2，UY,UAHave correlations with W1, W2, A, Y, respectively, P*Representing data typesPointer variable (see fig. 1); as shown, the exogenous nodes point to corresponding endogenous nodes, respectively. From this we construct the structural equation F:
W1＝fW1(UW1)
W2→fW2(W1,UW2)
A→fA(W1,W2,UA)
Y→fY(W1,W2,A,UY)。
wherein W1 can be UW1Formula f as a variableW1W2 can be represented by W1 and UW2Formula f as a variableW2A can be represented by W1, W2UAAnd formula f as a variableAY can be represented by W1, W2, A and UYFormula f as a variableYAnd (4) showing.
S2.3, determining a causal relationship parameter;
the implementation of this step includes:
ACE (E) (Y1) -E (Y0) P (Y1-1) -P (Y0-0), wherein ACE represents an Average cause Effect; y1 denotes the business objective obtained with the argument; y0 denotes the business objective that is obtained without using arguments. And E (Y1) and E (Y0) represent the expected values of Y1 and Y0, respectively. P (Y1 ═ 1) and P (Y0 ═ 0) are normalized distributions of Y1 and Y2.
S2.4, establishing a relation between the observation parameters and the causal model;
the implementation of this step includes:
Observation＝(W1,W2,A,Y)～P
ACE＝E[Y1]-E[Y0]。
and S2.6, evaluating the identifiability of the average causal action formula and determining the causal relationship of a plurality of paths.
The causal relationships exposed by the above disclosure may be represented as a bayesian network (a directed acyclic graph) such as fig. 1. The direction of the arrow is pushed back by the structure. When the Bayesian network is too large to be simply backward-inferred, the network can be decomposed into a plurality of sub-networks according to different conditions for learning, and the relation directions among the sub-network elements are recombined when determined. The mathematical theory of decomposition is derived as:
all the variable sets V are set to { W1, W2, A, Y }, if Xi ∈ A and Xj ∈ (A ≦ C), then
Figure BDA0002684715980000061
Figure BDA0002684715980000062
If Xi, Xj ∈ C, then
Figure BDA0002684715980000063
Figure BDA0002684715980000064
Or
Figure BDA0002684715980000065
Figure BDA0002684715980000066
Learning can then be performed by randomly intervening a variable to minimize the number of maximum possible networks and maximize entropy. The formula for calculating entropy is:
H＝-sum^(M)_(i＝1){Ii/L*log(Ii/L)}
where M is the possible orientation result; ii is the number of networks resulting from the directed result i and L is the sum of all Ii. H is the negative of the sum of li/L log (Ii/L) in all orientation results. The purpose of this equation is to determine a variable v that dares to intervene in this variable to determine the minimum range network. From this learning approach, causal relationships in complex bayesian networks can be determined.
And S3, selecting a business strategy according to the causal relationship model, continuously monitoring the strategy effect, further enriching modeling data, continuously improving analysis precision, and finally helping enterprises to realize business growth.
In this embodiment, the multi-source data includes: one or more of subjective factors of the consumer, objective factors of the consumer, product sales, user behavior data, survey data of satisfaction, investments in various media channels, and characteristics of historical promotional activities.
In this embodiment, the business objectives include: product pricing, product profits, net recommendation values for non-investigated users, business outcome indicators, promotional strategies, purchase outcomes.
Compared with the prior art, the information prediction method based on the causal relationship can remove pseudo correlation caused by interference factors; the limitation problems of less information and small quantity of research data are solved, and global insight is formed; the causal relationship model generated in the method can highly restore the real market environment, so that the data relationship reliability is improved, a nest mortar of the traditional data analysis technology is eliminated, data islands of different layers are communicated, the result capable of practically guiding business decision is output, and enterprise decision is directly assisted.
Example 2: pricing strategy for product and service
Referring to fig. 1 to 4, enterprise operation data can be used as historical data, a causal relationship model is constructed by using the subjective and objective factors and the product sales of consumers as independent variables and the product pricing as dependent variables based on the historical data, and the causal relationship model can learn the influence of the subjective and objective factors such as brand, preference and preference of consumers to products and services on the product sales, discover the linkage relationship among the product pricing, and predict the sales of the products under different pricing. Under the condition of given cost information, the optimal pricing of the product can be calculated by taking the optimal profit of the product as a target. The causal relationship model may further calculate optimal pricing for a given product portfolio targeting optimal profit for different units such as brands/categories/stores.
When the causal relationship model is constructed, data of environment, passenger flow, sales, competitive products and related products are fused, so that the real market environment is highly restored.
Optimal pricing for new items that did not appear in the historical data:
a causal relationship model of price and sales is built through a causal algorithm, and then how to price the maximum total sales can be realized in a business means allowable range is explored through a mode of solving the maximum value of a function, and the pricing is the optimal pricing.
Different targets can also be set: such as the desire to optimize the profit for a single product, or the desire to optimize the total profit for a product portfolio, or the desire to optimize the total profit for all products in a store. Because a causal relationship model of price and sales is constructed, the problem is converted into a multivariate function extreme value problem, and the extreme value can be calculated by carrying out quadratic derivation on the obtained multivariate function. Solving the corresponding multidimensional equations according to different objectives can realize that price points with the maximum profit of the single product, the maximum profit of the product combination and the maximum profit of the store are respectively called 'single product optimal pricing' (as shown in a diagram in fig. 2), 'product combination optimal pricing' (as shown in B diagram in fig. 2) and 'store optimal pricing' (as shown in C diagram in fig. 2).
Example 3: promote the satisfaction
Referring to fig. 1, 3 and 4, the research data based on the user behavior data and the satisfaction degree, wherein the user behavior data includes the browsing/clicking/posting behaviors of the user in the website and the mobile App. And after the model is established, the net recommendation value of the user who does not receive investigation can be predicted according to the behavior data collected by the service, and a recommender, a passive person and a derogator are identified.
For example, in an investigation aiming at the customer satisfaction of an operator, after taking the behaviors of a user such as call time, network usage, service function subscription and the like as independent variables and taking a net user recommendation value as a dependent variable (obtained through research), a causal relationship model is established, the causal relationship model can be used for predicting: for the users who do not receive investigation, the net recommendation value of the user can be predicted by using the model through the use behavior data (such as call condition, service subscription, network use and the like) of the client at the operator, so that service intervention can be made in advance.
The causal relationship model can analyze the reasons influencing the satisfaction degree on an individual level, and supports personalized marketing scheme customization.
The causal algorithm fuses the investigation data and a BI analysis platform of an internal CRM system, and the overall attitude of a user can be predicted only by inputting sample attitude data, so that the effects of saving cost and preempting market first opportunity are achieved, early warning is performed on crises in advance, along with the accumulation of data, the model precision is improved, the prediction is more and more accurate, and the value maximization of the data is realized.
Example 4: optimizing return on investment for marketing
Referring to fig. 1, fig. 3 and fig. 4, a causal relationship model is constructed by using a business result index as a target variable, wherein the business result index includes purchasing behavior, consumption amount and the like; outputting a causal path relation graph between releases in different media channels and target variables and contributions of the releases in the different media channels to business results; and calculating the return on investment of each media by combining the investment of each media channel. And changing the value of the manipulated variable according to the output causal relationship model, and simulating the effect of different media delivery combinations. The causal relationship model gets through data islands of different channels, CRM data, market research data and behavior big data are fused to conduct fusion understanding on behaviors of consumers, and the value of existing data of an enterprise is maximized; the input of different forms of media can be measured by a uniform index, namely the contribution to the business result, so that cross-channel comparison is realized; the problem of the relationship between brand construction investment and short-term promotion activities which troubles enterprises for a long time is solved, the positions and the functions of different investments in the overall strategy are determined, and guidance is provided for budget allocation of all departments; the influence of different levels of investment on the business result is predicted and simulated, and the setting of the business KPI can be guided under the limitation of the existing resources; by utilizing the characteristic of simulating the intervention effect, the method helps to evaluate the resource level required to be input to reach the preset business target.
Example 5: specifying promotional policies
Referring to fig. 1, 3 and 4, a causal relationship model between sales volume/sales and promotional activity characteristics, product characteristics, and competitive products is established using historical sales information and sales data of the products. The influence of different factors on sales volume/sales volume is observed; and predicting sales volume/sales amount under different promotion forms, and preferentially implementing according to promotion targets. The causal algorithm outputs a corresponding optimal promotion strategy according to different targets such as sales volume, profit and the like; after a given target is obtained, the optimal promotion strategy of the single product can be output, and the optimal promotion strategy of the appointed product combination can also be output; extracting the characteristics of historical promotion activities, such as promotion time, promotion single products, promotion duration, promotion and propaganda contents and the like, and generating an optimal promotion activity; and performing an effect model on the selectable promotion strategies, and selecting the optimal promotion strategy according to the target.
Example 6: online shopping Path analysis
Referring to fig. 1, fig. 3, and fig. 4, the user path stored in the e-commerce background is integrated to obtain the user behavior data. The causal relationship model is applied to ascertain the typed cause of various user actions directed to the purchase outcome. The development of the hysteresis factor is increased, and the volume of the transaction is continuously improved.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.

Claims (8)

1. A causal relationship-based information prediction method is characterized by comprising the following steps:
collecting multi-source data from business activities;
taking the multi-source data as independent variables and the business target as dependent variables, constructing a causal relationship model of the independent variables and the dependent variables, and establishing a relation between the business behavior and the business target;
and selecting a business strategy according to the causal relationship model.
2. The method of claim 1, wherein collecting multi-source data from business operations comprises collecting data from enterprise or/and third party collection research platforms.
3. The causal information prediction method of claim 1, wherein said causal model is constructed by:
determining a research problem; selecting factors contained in the problem, wherein the factors comprise independent variables and dependent variables, and constructing a structure cause and effect model; determining a causal relationship parameter; establishing a relation between the observation parameters and the causal model; and evaluating the identifiability of the average causal formula and determining the causal relationship of the multiple paths.
4. The method of claim 1, wherein the factors included in the problem are selected, wherein the factors include independent variables and dependent variables, and wherein constructing the structural causal model comprises:
determining an endogenous node: x ═ W1, W2, a, Y, where W1: basic miscellaneous elements, W2: non-observable confounding elements; a: independent variable, Y: a dependent variable; determining an exogenous node: u ═ U (U)W1,UW2，UY,UA)～P*Wherein, UW1,UW2，UY,UAHave correlations with W1, W2, A, Y, respectively, P*A pointer variable representing a data type; constructing a structural equation F:
W1＝fW1(UW1)
W2→fW2(W1,UW2)
A→fA(W1,W2,UA)
Y→fY(W1,W2,A,UY)。
5. the causal information prediction method of claim 1, wherein determining the causal parameters comprises: ACE (Y1) -E (Y0) ═ P (Y1 ═ 1) -P (Y0 ═ 0), where Y1 represents the business objective obtained using the argument; y0 denotes the business objective that is obtained without using arguments.
6. The causal information prediction method of claim 1, wherein associating the observation parameters with the causal model comprises:
Observation＝(W1,W2,A,Y)～P；
ACE＝E[Y1]-E[Y0]。
7. the causal information prediction method of claim 1, wherein the multi-source data comprises: one or more of subjective factors of the consumer, objective factors of the consumer, product sales, user behavior data, survey data of satisfaction, investments in various media channels, and characteristics of historical promotional activities.
8. The causal information prediction method of claim 1, wherein said business objectives comprise: product pricing, product profits, net recommendation values for non-investigated users, business outcome indicators, promotional strategies, purchase outcomes.
CN202010972803.9A 2020-09-16 2020-09-16 Information prediction method based on causal relationship Pending CN112116387A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010972803.9A CN112116387A (en) 2020-09-16 2020-09-16 Information prediction method based on causal relationship

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010972803.9A CN112116387A (en) 2020-09-16 2020-09-16 Information prediction method based on causal relationship

Publications (1)

Publication Number Publication Date
CN112116387A true CN112116387A (en) 2020-12-22

Family

ID=73803639

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010972803.9A Pending CN112116387A (en) 2020-09-16 2020-09-16 Information prediction method based on causal relationship

Country Status (1)

Country Link
CN (1) CN112116387A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113807886A (en) * 2021-08-30 2021-12-17 浪潮卓数大数据产业发展有限公司 Device and method for automatically predicting sales of electronic commerce
CN117557299A (en) * 2024-01-11 2024-02-13 天津慧聪科技有限公司 Marketing planning method and system based on computer assistance
CN118069930A (en) * 2024-04-17 2024-05-24 杭州高能云科技有限公司 Investment dynamic recommendation method and system based on matching analysis

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113807886A (en) * 2021-08-30 2021-12-17 浪潮卓数大数据产业发展有限公司 Device and method for automatically predicting sales of electronic commerce
CN117557299A (en) * 2024-01-11 2024-02-13 天津慧聪科技有限公司 Marketing planning method and system based on computer assistance
CN117557299B (en) * 2024-01-11 2024-03-22 天津慧聪科技有限公司 Marketing planning method and system based on computer assistance
CN118069930A (en) * 2024-04-17 2024-05-24 杭州高能云科技有限公司 Investment dynamic recommendation method and system based on matching analysis
CN118069930B (en) * 2024-04-17 2024-08-30 杭州高顾智能科技有限公司 Investment dynamic recommendation method and system based on matching analysis

Similar Documents

Publication Publication Date Title
Han et al. A fuzzy TOPSIS method for performance evaluation of reverse logistics in social commerce platforms
D’Haen et al. Model-supported business-to-business prospect prediction based on an iterative customer acquisition framework
KR102045477B1 (en) System for evaluating platform of ecommerce
Kane et al. Mining for the truly responsive customers and prospects using true-lift modeling: Comparison of new and existing methods
CN112116387A (en) Information prediction method based on causal relationship
Li et al. AI-driven customer relationship management for sustainable enterprise performance
WO2010019897A1 (en) Automatically prescribing total budget for marketing and sales resources and allocation across spending categories
Tang et al. Online-purchasing behavior forecasting with a firefly algorithm-based SVM model considering shopping cart use
Javan et al. A hybrid advertising media selection model using AHP and fuzzy-based GA decision making
Zaim et al. Multi-criteria analysis approach based on consumer satisfaction to rank B2C E-commerce websites
Orogun et al. Predicting consumer behaviour in digital market: a machine learning approach
Liébana-Cabanillas et al. Variable selection for payment in social networks: Introducing the Hy-index
AboElHamd et al. Dynamic programming models for maximizing customer lifetime value: an overview
Ray et al. Integrated approach of fuzzy multi-attribute decision making and data mining for customer segmentation
Tang et al. The impact of Instagram marketing on sale in the fashion industry
Parracho Sant'Anna et al. Statistical modeling and probabilistic composition in the prediction of the customer lifetime value
Stephen et al. An intelligent cloud-based customer relationship management system to determine flexible pricing for customer retention
Satapathy Effort estimation methods in software development using machine learning algorithms
Ramirez et al. Incorporating usage data for B2B churn prediction modeling
Tănase Predictive Marketing: Anticipating Market Demand with Proactive Action
Lu et al. E-service intelligence: An introduction
Mathew et al. Integrating AI in e-procurement of hospitality industry in the UAE
Onokoy et al. Modern approaches to building recommender systems for online stores
Tirenni Allocation of marketing resources to optimize customer equity
Іванов et al. The Advertising Prediction Model Based on Machine Learning Technologies

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20201222

RJ01 Rejection of invention patent application after publication