CN118037332A - Data processing method and system for managing marketing data - Google Patents

Data processing method and system for managing marketing data Download PDF

Info

Publication number
CN118037332A
CN118037332A CN202410276842.3A CN202410276842A CN118037332A CN 118037332 A CN118037332 A CN 118037332A CN 202410276842 A CN202410276842 A CN 202410276842A CN 118037332 A CN118037332 A CN 118037332A
Authority
CN
China
Prior art keywords
marketing
data
historical
real
time
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
CN202410276842.3A
Other languages
Chinese (zh)
Inventor
黄以槐
袁红艳
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.)
Hubei Zhongxuntai Technology Co ltd
Original Assignee
Hubei Zhongxuntai 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 Hubei Zhongxuntai Technology Co ltd filed Critical Hubei Zhongxuntai Technology Co ltd
Priority to CN202410276842.3A priority Critical patent/CN118037332A/en
Publication of CN118037332A publication Critical patent/CN118037332A/en
Pending legal-status Critical Current

Links

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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
    • 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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2337—Non-hierarchical techniques using fuzzy logic, i.e. fuzzy clustering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0475—Generative networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/094—Adversarial learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/02—Knowledge representation; Symbolic representation
    • G06N5/022—Knowledge engineering; Knowledge acquisition
    • G06N5/025—Extracting rules from data
    • 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)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Computing Systems (AREA)
  • Business, Economics & Management (AREA)
  • Evolutionary Biology (AREA)
  • Computational Linguistics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Development Economics (AREA)
  • Biophysics (AREA)
  • Strategic Management (AREA)
  • Finance (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Accounting & Taxation (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Probability & Statistics with Applications (AREA)
  • Fuzzy Systems (AREA)
  • Automation & Control Theory (AREA)
  • Game Theory and Decision Science (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention belongs to the technical field of data processing, and discloses a data processing method and a system for managing marketing data. The method comprises the following steps: collecting historical marketing big data, and performing data dimension reduction and label addition on the historical marketing big data to obtain a marketing effect classification sample set containing a historical marketing service type label and a historical marketing effect grade label; constructing a plurality of marketing service databases, a plurality of marketing effect sub-databases and a marketing effect classification model; and acquiring real-time marketing data, performing data dimension reduction according to the key marketing index set, acquiring a real-time marketing service type label and a real-time marketing effect grade label of the real-time marketing data after the data dimension reduction, and storing the real-time marketing data after the data dimension reduction into a corresponding marketing effect sub-database of a corresponding marketing service database. The invention solves the problems of simple steps, poor effect and unfavorable data management in the prior art.

Description

Data processing method and system for managing marketing data
Technical Field
The invention belongs to the technical field of data processing, and particularly relates to a data processing method and system for managing marketing data.
Background
With the rapid development of the internet and big data technology, marketing data of enterprises show explosive growth. Marketing data has important significance for enterprises to accurately marketing, and the marketing data needs to be properly managed so that the enterprises can analyze and apply the marketing data subsequently.
In the prior art, only simple data processing steps such as format conversion, compression, decompression and the like are performed on marketing data, so that the problems of high dimension, complexity, noise and the like of the marketing data cannot be solved, the data processing effect is poor, and the data value of the marketing data is low and management is difficult.
Disclosure of Invention
In order to solve the problems of simple steps, poor effect and unfavorable data management in the prior art, the invention aims to provide a data processing method and system for managing marketing data.
The technical scheme adopted by the invention is as follows:
a data processing method of managing marketing data, comprising the steps of:
Collecting historical marketing big data, performing data dimension reduction on the historical marketing big data to obtain a corresponding key marketing index set and a data dimension-reduced historical marketing data set, and setting a corresponding historical marketing service type label and a corresponding historical marketing effect grade label for each data dimension-reduced historical marketing data to obtain a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels;
Constructing a plurality of corresponding marketing service databases and a plurality of marketing effect sub-databases according to a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels, storing each marketing effect classification sample into the corresponding marketing effect sub-database of the corresponding marketing service database, and constructing a corresponding marketing effect classification model by using a deep learning algorithm according to a marketing effect classification sample set;
And acquiring real-time marketing data, carrying out data dimension reduction on the real-time marketing data according to the key marketing index set, obtaining corresponding real-time marketing data after the data dimension reduction, acquiring a real-time marketing service type label of the real-time marketing data after the data dimension reduction, using a marketing effect classification model, acquiring a real-time marketing effect grade label of the real-time marketing data after the data dimension reduction, and storing the real-time marketing data after the data dimension reduction into a corresponding marketing effect sub-database of a corresponding marketing service database according to the real-time marketing service type label and the real-time marketing effect grade label.
Further, the historical marketing big data comprises a plurality of historical marketing data, and the historical marketing data comprises historical marketing strategy text data and historical marketing index data;
The historical marketing index data comprises a historical marketing cost index, a historical marketing profit index, a historical marketing website flow index, a historical marketing social media index, a historical marketing conversion rate index, a historical marketing client acquisition cost index, a historical marketing performance index, a historical marketing competition analysis index and a historical client satisfaction index;
the real-time marketing data comprises real-time marketing strategy text data and real-time marketing index data;
the real-time marketing index data comprises a real-time marketing cost index, a real-time marketing profit index, a real-time marketing website flow index, a real-time marketing social media index, a real-time marketing conversion rate index, a real-time marketing client acquisition cost index, a real-time marketing performance index, a real-time market competition analysis index and a real-time client satisfaction index.
Further, collecting historical marketing big data, performing data dimension reduction on the historical marketing big data to obtain a corresponding key marketing index set and a data dimension-reduced historical marketing data set, setting a corresponding historical marketing service type label and a corresponding historical marketing effect grade label for each data dimension-reduced historical marketing data, and obtaining a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels, wherein the method comprises the following steps:
Collecting historical marketing big data, and preprocessing the historical marketing big data to obtain a preprocessed historical marketing data set;
Performing data dimension reduction on the preprocessed historical marketing data set by using a PCA method to obtain a corresponding key marketing index set and the data dimension-reduced historical marketing data set;
extracting a history keyword set of the history marketing data after the dimension reduction of each data, and taking the history keyword set as a corresponding history retrieval tag;
Setting a history marketing service type label of the history marketing data after each data dimension reduction according to the history keyword set;
Performing clustering processing on the historical marketing dataset after the dimensionality reduction by using an FCM clustering algorithm to obtain a plurality of clustering centers and a plurality of corresponding clustering clusters;
Setting a corresponding historical marketing effect grade label for each cluster center, and diffusing the historical marketing effect grade labels to cluster clusters corresponding to the cluster centers to obtain a plurality of marketing effect classification samples provided with the historical marketing effect grade labels;
And integrating all marketing effect classification samples to obtain a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect class labels.
Further, the PCA method is used for carrying out data dimension reduction on the preprocessed historical marketing data set to obtain a corresponding key marketing index set and the data dimension-reduced historical marketing data set, and the method comprises the following steps:
Performing matrix conversion on all the history marketing index data in the preprocessed history marketing data set by using a PCA method to obtain a history marketing index data matrix formed by a plurality of preprocessed history marketing index data row vectors;
Performing standardization processing on the historical marketing index data matrix to obtain a standardized historical marketing index data matrix;
acquiring a covariance matrix of the standardized historical marketing index data matrix, and acquiring a corresponding conversion matrix according to the covariance matrix;
Acquiring a principal component matrix formed by a plurality of alternative principal component column vectors according to the standardized historical marketing index data matrix and the corresponding conversion matrix;
Taking the alternative principal component column vector corresponding to the first 85% of the variance accumulated contribution rate as a principal component column vector, and taking the historical marketing index corresponding to the principal component column vector as a key marketing index to obtain a key marketing index set;
According to the key marketing index set, carrying out data dimension reduction on each preprocessed historical marketing data to obtain data dimension-reduced historical marketing data composed of a plurality of historical key marketing index data;
And traversing all the preprocessed historical marketing data in the preprocessed historical marketing data set to obtain a corresponding data dimension-reduced historical marketing data set.
Further, extracting a history keyword set of the history marketing data after the dimension reduction of each data, and taking the history keyword set as a corresponding history retrieval tag, wherein the method comprises the following steps:
extracting inter-class discrete factors and intra-class discrete factors of a plurality of feature words of the historical marketing strategy text data in the historical marketing data after the data is subjected to dimension reduction by using a TF-IDF-CI algorithm;
extracting word frequency and reverse text frequency of feature words of the corresponding historical marketing data after the data is subjected to dimension reduction;
Acquiring weights of a plurality of feature words according to the inter-class discrete factors, the intra-class discrete factors, the word frequency and the reverse text frequency;
sorting according to the weights of the feature words, selecting the first M feature words as a history keyword set of corresponding data dimension-reduced history marketing data, wherein M is the total number of preset history keywords;
splicing the history keyword sets to obtain a history retrieval tag of the corresponding data dimension-reduced history marketing data;
And traversing all the data dimension-reduced historical marketing data in the data dimension-reduced historical marketing data set to obtain a historical keyword set and a corresponding historical retrieval tag of each data dimension-reduced historical marketing data.
Further, according to the history keyword set, setting a history marketing service type label of the history marketing data after each data dimension reduction, including the following steps:
acquiring marketing knowledge big data, and preprocessing the marketing knowledge big data to obtain a preprocessed marketing knowledge data set;
constructing a corresponding marketing knowledge graph according to the preprocessed marketing knowledge data set;
and according to the historical keyword set, using a marketing knowledge graph to analyze the marketing service types to obtain a historical marketing service type label corresponding to the historical marketing data after the dimension of each data is reduced.
Further, using an FCM clustering algorithm to perform clustering processing on the historical marketing dataset after the data is subjected to dimension reduction to obtain a plurality of clustering centers and a plurality of corresponding clustering clusters, wherein the method comprises the following steps:
initializing a historical marketing data set after data dimension reduction to obtain the initial membership degree of the historical marketing data after each data dimension reduction to a clustering center;
Clustering the historical marketing data set after the data dimension reduction by using an FCM clustering algorithm according to the initial membership degree to obtain a plurality of initial clustering centers;
Acquiring a Lagrangian multiplier method according to the current membership degree, acquiring a merging function value and a change value, if the merging function value is larger than a merging function value threshold or the change value is larger than a change value threshold, carrying out iterative updating on the current clustering center and the current membership degree to obtain an updated clustering center and an updated membership degree, and repeating the step, otherwise, taking the current clustering center as a final clustering center, and entering the next step;
And dividing the historical marketing data set after the dimension reduction according to Euclidean distance between the historical marketing data after the dimension reduction of each data and a plurality of final clustering centers to obtain each final clustering center and a corresponding clustering cluster.
Further, according to a plurality of historical marketing service type labels and a plurality of historical marketing effect class labels, a plurality of corresponding marketing service databases and a plurality of marketing effect sub-databases are constructed, each marketing effect classification sample is stored in the corresponding marketing effect sub-database of the corresponding marketing service database, and according to a marketing effect classification sample set, a deep learning algorithm is used to construct a corresponding marketing effect classification model, comprising the following steps:
Constructing corresponding A marketing service databases according to the A-type historical marketing service type labels, wherein A is the type number of the historical marketing service type labels;
constructing corresponding B marketing effect sub-databases in each marketing service database according to the B-type historical marketing effect grade labels, wherein B is the type number of the historical marketing effect grade labels;
According to the historical marketing service type label and the historical marketing effect grade label of each marketing effect classification sample, storing the marketing effect classification samples into corresponding marketing effect sub-databases of the corresponding marketing service databases;
and constructing a corresponding marketing effect classification model by using an N-GAN-Attention-MLP algorithm according to the marketing effect classification sample set.
Further, collecting real-time marketing data, performing data dimension reduction on the real-time marketing data according to a key marketing index set to obtain corresponding real-time marketing data after the data dimension reduction, obtaining a real-time marketing service type label of the real-time marketing data after the data dimension reduction, using a marketing effect classification model to obtain a real-time marketing effect grade label of the real-time marketing data after the data dimension reduction, and storing the real-time marketing data after the data dimension reduction into a corresponding marketing effect sub-database of a corresponding marketing service database according to the real-time marketing service type label and the real-time marketing effect grade label, wherein the method comprises the following steps:
Collecting real-time marketing data, and performing data dimension reduction on the real-time marketing data according to the key marketing index set to obtain corresponding data dimension-reduced real-time marketing data;
Extracting a real-time keyword set of the real-time marketing data after the data is subjected to dimension reduction by using a TF-IDF-CI algorithm, and taking the real-time keyword set as a corresponding real-time retrieval tag;
According to the real-time keyword set, matching corresponding real-time marketing service type labels by using a marketing knowledge graph;
according to the real-time marketing service type label, transmitting the real-time marketing data after the data is reduced in size to a corresponding marketing service database;
using a marketing effect classification model to classify the marketing effect of the real-time marketing data after the data is subjected to dimension reduction, and obtaining a corresponding real-time marketing effect grade label;
and storing the real-time marketing data after the data is reduced in size into a corresponding marketing effect sub-database of the marketing service database according to the real-time marketing effect grade label.
A data processing system for managing marketing data is used for realizing a data processing method, and comprises a data initialization processing unit, a database construction unit and a real-time marketing data processing unit which are connected in sequence;
The data initialization processing unit is used for collecting the historical marketing big data, carrying out data dimension reduction on the historical marketing big data to obtain a corresponding key marketing index set and a data dimension-reduced historical marketing data set, setting a corresponding historical marketing service type label and a historical marketing effect grade label for each data dimension-reduced historical marketing data, and obtaining a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels;
The database construction unit is used for constructing a plurality of corresponding marketing service databases and a plurality of marketing effect sub-databases according to a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels, storing each marketing effect classification sample into the corresponding marketing effect sub-database of the corresponding marketing service database, and constructing a corresponding marketing effect classification model by using a deep learning algorithm according to a marketing effect classification sample set;
The real-time marketing data processing unit is used for collecting real-time marketing data, carrying out data dimension reduction on the real-time marketing data according to the key marketing index set, obtaining real-time marketing data after the corresponding data dimension reduction, obtaining real-time marketing service type labels of the real-time marketing data after the data dimension reduction, using a marketing effect classification model, obtaining real-time marketing effect grade labels of the real-time marketing data after the data dimension reduction, and storing the real-time marketing data after the data dimension reduction into a corresponding marketing effect sub-database of a corresponding marketing service database according to the real-time marketing service type labels and the real-time marketing effect grade labels.
The beneficial effects of the invention are as follows:
The invention discloses a data processing method and a system for managing marketing data, which are used for carrying out data dimension reduction on the marketing data, solving the problems of high dimension, complexity and noise of the marketing data and reducing the volume of the data and the occupied space of a memory; the marketing service database and the marketing effect sub-database are constructed to store marketing data in a classified mode according to the marketing service type and the marketing effect grade, so that data exchange among the marketing data is enhanced, management and retrieval of the marketing data are facilitated, and the data value of the marketing data is improved; the marketing effect classification model automatically and efficiently classifies marketing data, and the practicability and the accuracy of marketing data classification and storage are improved.
Other advantageous effects of the present invention will be further described in the detailed description.
Drawings
FIG. 1 is a flow chart of a data processing method of managing marketing data in the present invention.
FIG. 2 is a block diagram of a data processing system for managing marketing data in the present invention.
Detailed Description
The invention is further illustrated by the following description of specific embodiments in conjunction with the accompanying drawings.
Example 1:
As shown in fig. 1, the present embodiment provides a data processing method for managing marketing data, including the steps of:
S1: the method comprises the steps of collecting historical marketing big data, carrying out data dimension reduction on the historical marketing big data to obtain a corresponding key marketing index set and a data dimension-reduced historical marketing data set, setting a corresponding historical marketing service type label and a corresponding historical marketing effect grade label for each data dimension-reduced historical marketing data, and obtaining a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels, and comprises the following steps:
s1-1: collecting historical marketing big data, and preprocessing the historical marketing big data to obtain a preprocessed historical marketing data set;
The historical marketing big data comprises a plurality of historical marketing data, and the historical marketing data comprises historical marketing strategy text data and historical marketing index data;
The historical marketing index data comprises a historical marketing cost index, a historical marketing profit index, a historical marketing website flow index, a historical marketing social media index, a historical marketing conversion rate index, a historical marketing client acquisition cost index, a historical marketing performance index, a historical marketing competition analysis index and a historical client satisfaction index;
preprocessing comprises the steps of carrying out data format conversion, repeated data screening and error data rejection on the historical marketing big data;
S1-2: performing data dimension reduction on the preprocessed historical marketing dataset by using a principal component analysis (PRINCIPAL COMPONENT ANALYSIS, PCA) method to obtain a corresponding key marketing index set and the data dimension-reduced historical marketing dataset, wherein the method comprises the following steps of:
S1-2-1: using PCA method to perform matrix conversion on all the history marketing index data in the preprocessed history marketing data set to obtain a history marketing index data matrix composed of a plurality of preprocessed history marketing index data row vectors Wherein/>The method comprises the steps that (1) row vectors of historical marketing index data after pretreatment are the ith row vector indication quantity, and n is the total number of the historical marketing data after pretreatment;
s1-2-2: performing standardization processing on the historical marketing index data matrix to obtain a standardized historical marketing index data matrix;
The formula is:
In the method, in the process of the invention, A standardized historical marketing index data matrix; /(I)The average value of the historical marketing index data matrix; /(I)Variance of the historical marketing index data matrix;
S1-2-3: acquiring a covariance matrix of the standardized historical marketing index data matrix, and acquiring a corresponding conversion matrix according to the covariance matrix;
The formula is:
In the method, in the process of the invention, A covariance matrix of the historical marketing index data matrix after the standardization processing; /(I)Is a principal component matrix; Is a conversion matrix; /(I) Is a unit feature vector matrix; n is the total number of the history marketing data after pretreatment;
S1-2-4: according to the standardized historical marketing index data matrix and the corresponding conversion matrix, a principal component matrix formed by a plurality of alternative principal component column vectors is obtained ; Wherein/>For/>Alternate principal component column vector,/>For the data dimension indication of post-preprocessing historical marketing data,/>A total number of data dimensions;
The formula is:
In the method, in the process of the invention, Is a principal component matrix; /(I)Is a conversion matrix; /(I)A standardized historical marketing index data matrix;
S1-2-5: taking the alternative principal component column vector corresponding to the first 85% of the variance accumulated contribution rate as a principal component column vector, and taking the historical marketing index corresponding to the principal component column vector as a key marketing index to obtain a key marketing index set;
The formula is:
In the method, in the process of the invention, Accumulating contribution rates for variances; /(I)First/>Individual principal component alternatives/>Is a variance of (2); /(I)Indicating the quantity for the data dimension; /(I)A total number of data dimensions; /(I)Is the total number of main components;
Principal component column vector The corresponding historical marketing index is used as a key marketing index;
S1-2-6: according to the key marketing index set, carrying out data dimension reduction on each preprocessed historical marketing data to obtain data dimension-reduced historical marketing data composed of a plurality of historical key marketing index data;
S1-2-7: traversing all the preprocessed historical marketing data in the preprocessed historical marketing data set to obtain a corresponding data dimension-reduced historical marketing data set;
S1-3: extracting a history keyword set of the history marketing data after the dimension reduction of each data, and taking the history keyword set as a corresponding history retrieval tag, wherein the method comprises the following steps of:
S1-3-1: extracting inter-class discrete factors and intra-class discrete factors of a plurality of feature words of the historical marketing strategy text data in the historical marketing data after the data is subjected to dimension reduction by using a (TF-IDF-CI, term Frequency-Inverse Document Frequency-Class Information) algorithm;
The formula is:
In the method, in the process of the invention, Is an inter-class discrete factor; /(I)Is characterized by word/>Standard deviation of (2); /(I)Is characterized by word/>Is a category of (2); Is the total number of categories; /(I) Is characterized by word/>And in category/>Is a frequency of occurrence in the first and second embodiments; /(I)For/>Frequency of occurrence in each class; /(I)Indicating the quantity for the feature words; /(I)Indicating the quantity for the category;
In the method, in the process of the invention, Is an intra-class discrete factor; /(I)For category/>Commonly occurring feature words/>Is the number of (3); To include feature words/> Category/>Is a number of documents; /(I)To include category/>Is a total number of documents;
S1-3-2: extracting word frequency and reverse text frequency of feature words of the corresponding historical marketing data after the data is subjected to dimension reduction;
S1-3-3: acquiring weights of a plurality of feature words according to the inter-class discrete factors, the intra-class discrete factors, the word frequency and the reverse text frequency;
The formula is:
In the method, in the process of the invention, Is characterized by word/>Weights of (2); /(I)Is word frequency; /(I)Is the reverse text frequency; /(I)Is a discrete factor;
S1-3-4: sorting according to the weights of the feature words, selecting the first M feature words as a history keyword set of corresponding data dimension-reduced history marketing data, wherein M is the total number of preset history keywords;
s1-3-5: splicing the history keyword sets to obtain a history retrieval tag of the corresponding data dimension-reduced history marketing data;
S1-3-6: traversing all the data dimension-reduced historical marketing data in the data dimension-reduced historical marketing data set to obtain a historical keyword set and a corresponding historical retrieval tag of each data dimension-reduced historical marketing data;
S1-4: according to the historical keyword set, setting a historical marketing service type label of the historical marketing data after each data dimension reduction, comprising the following steps:
s1-4-1: acquiring marketing knowledge big data, and preprocessing the marketing knowledge big data to obtain a preprocessed marketing knowledge data set;
s1-4-2: constructing a corresponding marketing knowledge graph according to the preprocessed marketing knowledge data set, wherein the method comprises the following steps of:
s1-4-2-1: extracting a plurality of named entities of each preprocessed marketing knowledge data;
s1-4-2-2: extracting a plurality of entity relations of each preprocessed marketing knowledge data;
s1-4-2-3: carrying out knowledge fusion according to a plurality of named entities and a plurality of entity relations, and constructing a corresponding marketing knowledge graph;
S1-4-3: according to the historical keyword set, using a marketing knowledge graph to analyze the marketing service types to obtain a historical marketing service type label corresponding to the historical marketing data after the dimension of each data is reduced;
s1-5: clustering the historical marketing dataset after the data is reduced in size by using a Fuzzy C-Means (FCM) clustering algorithm to obtain a plurality of clustering centers and a plurality of corresponding clustering clusters, wherein the clustering method comprises the following steps of:
s1-5-1: initializing a historical marketing data set after data dimension reduction to obtain the initial membership degree of the historical marketing data after each data dimension reduction to a clustering center;
S1-5-2: clustering the historical marketing data set after the data dimension reduction by using an FCM clustering algorithm according to the initial membership degree to obtain a plurality of initial clustering centers;
S1-5-3: acquiring a Lagrangian multiplier method according to the current membership degree, acquiring a merging function value and a change value, if the merging function value is larger than a merging function value threshold or the change value is larger than a change value threshold, carrying out iterative updating on the current clustering center and the current membership degree to obtain an updated clustering center and an updated membership degree, and repeating the step, otherwise, taking the current clustering center as a final clustering center, and entering the next step;
The formula is:
In the method, in the process of the invention, For/>Combining function values at the moment; /(I)Is a variation value; /(I)For/>Characteristic values of historical marketing data after data dimension reduction; /(I)For/>Historical marketing data to the/>, after dimension reduction of the individual dataThe distance of the current cluster center; /(I)For/>Historical marketing data after data dimension reduction for the/>The current membership of the cluster center; /(I)Is a super parameter; /(I)Data total; /(I)Is the total number of clustering centers; /(I)Indicating the quantity for the clustering center; /(I)Indicating the quantity of the historical marketing data after the data is subjected to dimension reduction;
In the method, in the process of the invention, For/>Historical marketing data after data dimension reduction for the/>Updated membership of the cluster center; For/> Historical marketing data to the/>, after dimension reduction of the individual data、/>The distance of the updated cluster centers; /(I)Indicating the quantity for the clustering center;
In the method, in the process of the invention, For/>An updated cluster center; /(I)For/>Historical marketing data after data dimension reduction; /(I)For/>Historical marketing data after data dimension reduction for the/>Updated membership of the cluster center;
s1-5-4: dividing the historical marketing data set after the dimension reduction according to Euclidean distance between the historical marketing data after the dimension reduction of each data and a plurality of final clustering centers to obtain each final clustering center and a corresponding clustering cluster;
s1-6: setting a corresponding historical marketing effect grade label for each cluster center, and diffusing the historical marketing effect grade labels to cluster clusters corresponding to the cluster centers to obtain a plurality of marketing effect classification samples provided with the historical marketing effect grade labels;
S1-7: integrating all marketing effect classification samples to obtain a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect class labels;
S2: constructing a plurality of corresponding marketing service databases and a plurality of marketing effect sub-databases according to a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels, storing each marketing effect classification sample into the corresponding marketing effect sub-database of the corresponding marketing service database, and constructing a corresponding marketing effect classification model by using a deep learning algorithm according to a marketing effect classification sample set, comprising the following steps:
s2-1: constructing corresponding A marketing service databases according to the A-type historical marketing service type labels, wherein A is the type number of the historical marketing service type labels;
s2-2: constructing corresponding B marketing effect sub-databases in each marketing service database according to the B-type historical marketing effect grade labels, wherein B is the type number of the historical marketing effect grade labels;
S2-3: according to the historical marketing service type label and the historical marketing effect grade label of each marketing effect classification sample, storing the marketing effect classification samples into corresponding marketing effect sub-databases of the corresponding marketing service databases;
S2-4: according to the marketing effect classification sample set, using an N-generated countermeasure Network (GAN) -Attention-multilayer perceptron (Multilayer Perceptron, MLP) algorithm to construct a corresponding marketing effect classification model, comprising the steps of:
s2-4-1: classifying the plurality of marketing effect classification samples in the marketing effect classification sample set according to 7:3, dividing the ratio into a model training sample set and a model testing sample set;
S2-4-2: an initial marketing effect classification model is built by using an N-GAN-Attention-MLP algorithm, a model training sample set is input into the initial marketing effect classification model, optimization training is carried out, and an optimized marketing effect classification model is obtained, wherein N is the total number of key marketing indexes;
s2-4-3: inputting the model test sample set into an optimized marketing effect classification model, and performing model test to obtain a plurality of corresponding predictive marketing effect grade labels;
s2-4-4: comparing and counting according to each predicted marketing effect grade label and the corresponding historical marketing effect grade label to obtain model test accuracy;
s2-4-5: if the model test accuracy is greater than a preset model test accuracy threshold, outputting an optimal marketing effect classification model, otherwise, continuing to perform optimization training;
S3: collecting real-time marketing data, carrying out data dimension reduction on the real-time marketing data according to a key marketing index set to obtain corresponding real-time marketing data after the data dimension reduction, obtaining a real-time marketing service type label of the real-time marketing data after the data dimension reduction, using a marketing effect classification model to obtain a real-time marketing effect grade label of the real-time marketing data after the data dimension reduction, and storing the real-time marketing data after the data dimension reduction into a corresponding marketing effect sub-database of a corresponding marketing service database according to the real-time marketing service type label and the real-time marketing effect grade label, wherein the method comprises the following steps:
S3-1: collecting real-time marketing data, and performing data dimension reduction on the real-time marketing data according to the key marketing index set to obtain corresponding data dimension-reduced real-time marketing data;
the real-time marketing data comprises real-time marketing strategy text data and real-time marketing index data;
The real-time marketing index data comprises a real-time marketing cost index, a real-time marketing profit index, a real-time marketing website flow index, a real-time marketing social media index, a real-time marketing conversion rate index, a real-time marketing client acquisition cost index, a real-time marketing performance index, a real-time market competition analysis index and a real-time client satisfaction index;
S3-2: extracting a real-time keyword set of the real-time marketing data after the data is subjected to dimension reduction by using a TF-IDF-CI algorithm, and taking the real-time keyword set as a corresponding real-time retrieval tag;
s3-3: according to the real-time keyword set, matching corresponding real-time marketing service type labels by using a marketing knowledge graph;
s3-4: according to the real-time marketing service type label, transmitting the real-time marketing data after the data dimension reduction to a corresponding marketing service database, comprising the following steps:
S3-4-1: if the historical marketing service type label matched with the real-time marketing service type label exists, transmitting the real-time marketing data after the data is reduced in size to a marketing service database corresponding to the matched historical marketing service type label, otherwise, entering the next step;
S3-4-2: constructing a newly added marketing service database according to the real-time marketing service type label, and transmitting real-time marketing data after the data is reduced in size to the newly added marketing service database;
S3-4-3: constructing corresponding B marketing effect sub-databases in the newly added marketing service database according to the B-class historical marketing effect grade labels;
S3-5: and classifying the marketing effect of the real-time marketing data after the dimension reduction by using a marketing effect classification model to obtain a corresponding real-time marketing effect grade label, wherein the method comprises the following steps of:
s3-5-1: inputting the real-time marketing data after the data is reduced in size into a marketing effect classification model;
s3-5-2: n generators based on the GAN network of the marketing effect classification model are used for extracting data characteristics of real-time marketing data at N angles after the data is reduced in dimension;
The converter structure is used as a generator of the GAN network and is used for extracting data characteristics of real-time marketing data after data dimension reduction at a plurality of angles, and the converter structure has generalization capability of extracting characteristics of information with different scales; the attribute mechanism dynamically adjusts the weights of a plurality of multi-scale generators of the GAN network, so that the multi-scale generators have different weights;
s3-5-3: feature fusion is carried out on the data features of the N angles, and corresponding fusion data features are obtained;
The MLP network is used as a full-connection layer, and is characterized in that all neurons among the layers are connected with each other and used for carrying out feature fusion on the data features of the N-angle dimension-reduced real-time marketing data to obtain multi-angle fusion features and improve the data representation capability;
s3-5-4: classifying the marketing effect according to the fusion data characteristics to obtain corresponding real-time marketing effect grade labels;
The classifier is constructed based on the softmax function and is used for predicting the real-time marketing effect grade label according to the multi-angle fusion characteristics to obtain the real-time marketing effect grade label;
S3-6: and storing the real-time marketing data after the data is reduced in size into a corresponding marketing effect sub-database of the marketing service database according to the real-time marketing effect grade label.
Example 2:
As shown in fig. 2, the present embodiment provides a data processing system for managing marketing data, for implementing a data processing method, where the system includes a data initialization processing unit, a database construction unit, and a real-time marketing data processing unit that are sequentially connected;
The data initialization processing unit is used for collecting the historical marketing big data, carrying out data dimension reduction on the historical marketing big data to obtain a corresponding key marketing index set and a data dimension-reduced historical marketing data set, setting a corresponding historical marketing service type label and a historical marketing effect grade label for each data dimension-reduced historical marketing data, and obtaining a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels;
The database construction unit is used for constructing a plurality of corresponding marketing service databases and a plurality of marketing effect sub-databases according to a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels, storing each marketing effect classification sample into the corresponding marketing effect sub-database of the corresponding marketing service database, and constructing a corresponding marketing effect classification model by using a deep learning algorithm according to a marketing effect classification sample set;
The real-time marketing data processing unit is used for collecting real-time marketing data, carrying out data dimension reduction on the real-time marketing data according to the key marketing index set, obtaining real-time marketing data after the corresponding data dimension reduction, obtaining real-time marketing service type labels of the real-time marketing data after the data dimension reduction, using a marketing effect classification model, obtaining real-time marketing effect grade labels of the real-time marketing data after the data dimension reduction, and storing the real-time marketing data after the data dimension reduction into a corresponding marketing effect sub-database of a corresponding marketing service database according to the real-time marketing service type labels and the real-time marketing effect grade labels.
The invention discloses a data processing method and a system for managing marketing data, which are used for carrying out data dimension reduction on the marketing data, solving the problems of high dimension, complexity and noise of the marketing data and reducing the volume of the data and the occupied space of a memory; the marketing service database and the marketing effect sub-database are constructed to store marketing data in a classified mode according to the marketing service type and the marketing effect grade, so that data exchange among the marketing data is enhanced, management and retrieval of the marketing data are facilitated, and the data value of the marketing data is improved; the marketing effect classification model automatically and efficiently classifies marketing data, and the practicability and the accuracy of marketing data classification and storage are improved.
The invention is not limited to the alternative embodiments described above, but any person may derive other various forms of products in the light of the present invention. The above detailed description should not be construed as limiting the scope of the invention, which is defined in the claims and the description may be used to interpret the claims.

Claims (10)

1. A data processing method for managing marketing data, characterized by: the method comprises the following steps:
Collecting historical marketing big data, performing data dimension reduction on the historical marketing big data to obtain a corresponding key marketing index set and a data dimension-reduced historical marketing data set, and setting a corresponding historical marketing service type label and a corresponding historical marketing effect grade label for each data dimension-reduced historical marketing data to obtain a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels;
Constructing a plurality of corresponding marketing service databases and a plurality of marketing effect sub-databases according to a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels, storing each marketing effect classification sample into the corresponding marketing effect sub-database of the corresponding marketing service database, and constructing a corresponding marketing effect classification model by using a deep learning algorithm according to a marketing effect classification sample set;
And acquiring real-time marketing data, carrying out data dimension reduction on the real-time marketing data according to the key marketing index set, obtaining corresponding real-time marketing data after the data dimension reduction, acquiring a real-time marketing service type label of the real-time marketing data after the data dimension reduction, using a marketing effect classification model, acquiring a real-time marketing effect grade label of the real-time marketing data after the data dimension reduction, and storing the real-time marketing data after the data dimension reduction into a corresponding marketing effect sub-database of a corresponding marketing service database according to the real-time marketing service type label and the real-time marketing effect grade label.
2. The data processing method for managing marketing data of claim 1, wherein: the historical marketing big data comprises a plurality of historical marketing data, and the historical marketing data comprises historical marketing strategy text data and historical marketing index data;
the historical marketing index data comprises a historical marketing cost index, a historical marketing profit index, a historical marketing website flow index, a historical marketing social media index, a historical marketing conversion rate index, a historical marketing client acquisition cost index, a historical marketing performance index, a historical market competition analysis index and a historical client satisfaction index;
the real-time marketing data comprises real-time marketing strategy text data and real-time marketing index data;
The real-time marketing index data comprises a real-time marketing cost index, a real-time marketing profit index, a real-time marketing website flow index, a real-time marketing social media index, a real-time marketing conversion rate index, a real-time marketing client acquisition cost index, a real-time marketing performance index, a real-time market competition analysis index and a real-time client satisfaction index.
3. The data processing method for managing marketing data of claim 2, wherein: the method comprises the steps of collecting historical marketing big data, carrying out data dimension reduction on the historical marketing big data to obtain a corresponding key marketing index set and a data dimension-reduced historical marketing data set, setting a corresponding historical marketing service type label and a corresponding historical marketing effect grade label for each data dimension-reduced historical marketing data, and obtaining a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels, and comprises the following steps:
Collecting historical marketing big data, and preprocessing the historical marketing big data to obtain a preprocessed historical marketing data set;
Performing data dimension reduction on the preprocessed historical marketing data set by using a PCA method to obtain a corresponding key marketing index set and the data dimension-reduced historical marketing data set;
extracting a history keyword set of the history marketing data after the dimension reduction of each data, and taking the history keyword set as a corresponding history retrieval tag;
Setting a history marketing service type label of the history marketing data after each data dimension reduction according to the history keyword set;
Performing clustering processing on the historical marketing dataset after the dimensionality reduction by using an FCM clustering algorithm to obtain a plurality of clustering centers and a plurality of corresponding clustering clusters;
Setting a corresponding historical marketing effect grade label for each cluster center, and diffusing the historical marketing effect grade labels to cluster clusters corresponding to the cluster centers to obtain a plurality of marketing effect classification samples provided with the historical marketing effect grade labels;
And integrating all marketing effect classification samples to obtain a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect class labels.
4. A data processing method of managing marketing data according to claim 3, wherein: and performing data dimension reduction on the preprocessed historical marketing data set by using a PCA method to obtain a corresponding key marketing index set and the data dimension-reduced historical marketing data set, wherein the method comprises the following steps of:
Performing matrix conversion on all the history marketing index data in the preprocessed history marketing data set by using a PCA method to obtain a history marketing index data matrix formed by a plurality of preprocessed history marketing index data row vectors;
Performing standardization processing on the historical marketing index data matrix to obtain a standardized historical marketing index data matrix;
acquiring a covariance matrix of the standardized historical marketing index data matrix, and acquiring a corresponding conversion matrix according to the covariance matrix;
Acquiring a principal component matrix formed by a plurality of alternative principal component column vectors according to the standardized historical marketing index data matrix and the corresponding conversion matrix;
Taking the alternative principal component column vector corresponding to the first 85% of the variance accumulated contribution rate as a principal component column vector, and taking the historical marketing index corresponding to the principal component column vector as a key marketing index to obtain a key marketing index set;
According to the key marketing index set, carrying out data dimension reduction on each preprocessed historical marketing data to obtain data dimension-reduced historical marketing data composed of a plurality of historical key marketing index data;
And traversing all the preprocessed historical marketing data in the preprocessed historical marketing data set to obtain a corresponding data dimension-reduced historical marketing data set.
5. A data processing method of managing marketing data according to claim 3, wherein: extracting a history keyword set of the history marketing data after the dimension reduction of each data, and taking the history keyword set as a corresponding history retrieval tag, wherein the method comprises the following steps of:
extracting inter-class discrete factors and intra-class discrete factors of a plurality of feature words of the historical marketing strategy text data in the historical marketing data after the data is subjected to dimension reduction by using a TF-IDF-CI algorithm;
extracting word frequency and reverse text frequency of feature words of the corresponding historical marketing data after the data is subjected to dimension reduction;
Acquiring weights of a plurality of feature words according to the inter-class discrete factors, the intra-class discrete factors, the word frequency and the reverse text frequency;
sorting according to the weights of the feature words, selecting the first M feature words as a history keyword set of corresponding data dimension-reduced history marketing data, wherein M is the total number of preset history keywords;
splicing the history keyword sets to obtain a history retrieval tag of the corresponding data dimension-reduced history marketing data;
And traversing all the data dimension-reduced historical marketing data in the data dimension-reduced historical marketing data set to obtain a historical keyword set and a corresponding historical retrieval tag of each data dimension-reduced historical marketing data.
6. The data processing method for managing marketing data of claim 5, wherein: according to the historical keyword set, setting a historical marketing service type label of the historical marketing data after each data dimension reduction, comprising the following steps:
acquiring marketing knowledge big data, and preprocessing the marketing knowledge big data to obtain a preprocessed marketing knowledge data set;
constructing a corresponding marketing knowledge graph according to the preprocessed marketing knowledge data set;
and according to the historical keyword set, using a marketing knowledge graph to analyze the marketing service types to obtain a historical marketing service type label corresponding to the historical marketing data after the dimension of each data is reduced.
7. A data processing method of managing marketing data according to claim 3, wherein: performing clustering processing on the historical marketing dataset after the dimensionality reduction by using an FCM clustering algorithm to obtain a plurality of clustering centers and a plurality of corresponding clustering clusters, wherein the method comprises the following steps of:
initializing a historical marketing data set after data dimension reduction to obtain the initial membership degree of the historical marketing data after each data dimension reduction to a clustering center;
Clustering the historical marketing data set after the data dimension reduction by using an FCM clustering algorithm according to the initial membership degree to obtain a plurality of initial clustering centers;
Acquiring a Lagrangian multiplier method according to the current membership degree, acquiring a merging function value and a change value, if the merging function value is larger than a merging function value threshold or the change value is larger than a change value threshold, carrying out iterative updating on the current clustering center and the current membership degree to obtain an updated clustering center and an updated membership degree, and repeating the step, otherwise, taking the current clustering center as a final clustering center, and entering the next step;
And dividing the historical marketing data set after the dimension reduction according to Euclidean distance between the historical marketing data after the dimension reduction of each data and a plurality of final clustering centers to obtain each final clustering center and a corresponding clustering cluster.
8. A data processing method of managing marketing data according to claim 3, wherein: constructing a plurality of corresponding marketing service databases and a plurality of marketing effect sub-databases according to a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels, storing each marketing effect classification sample into the corresponding marketing effect sub-database of the corresponding marketing service database, and constructing a corresponding marketing effect classification model by using a deep learning algorithm according to a marketing effect classification sample set, comprising the following steps:
Constructing corresponding A marketing service databases according to the A-type historical marketing service type labels, wherein A is the type number of the historical marketing service type labels;
constructing corresponding B marketing effect sub-databases in each marketing service database according to the B-type historical marketing effect grade labels, wherein B is the type number of the historical marketing effect grade labels;
According to the historical marketing service type label and the historical marketing effect grade label of each marketing effect classification sample, storing the marketing effect classification samples into corresponding marketing effect sub-databases of the corresponding marketing service databases;
and constructing a corresponding marketing effect classification model by using an N-GAN-Attention-MLP algorithm according to the marketing effect classification sample set.
9. The data processing method for managing marketing data of claim 6, wherein: collecting real-time marketing data, carrying out data dimension reduction on the real-time marketing data according to a key marketing index set to obtain corresponding real-time marketing data after the data dimension reduction, obtaining a real-time marketing service type label of the real-time marketing data after the data dimension reduction, using a marketing effect classification model to obtain a real-time marketing effect grade label of the real-time marketing data after the data dimension reduction, and storing the real-time marketing data after the data dimension reduction into a corresponding marketing effect sub-database of a corresponding marketing service database according to the real-time marketing service type label and the real-time marketing effect grade label, wherein the method comprises the following steps:
Collecting real-time marketing data, and performing data dimension reduction on the real-time marketing data according to the key marketing index set to obtain corresponding data dimension-reduced real-time marketing data;
Extracting a real-time keyword set of the real-time marketing data after the data is subjected to dimension reduction by using a TF-IDF-CI algorithm, and taking the real-time keyword set as a corresponding real-time retrieval tag;
According to the real-time keyword set, matching corresponding real-time marketing service type labels by using a marketing knowledge graph;
according to the real-time marketing service type label, transmitting the real-time marketing data after the data is reduced in size to a corresponding marketing service database;
using a marketing effect classification model to classify the marketing effect of the real-time marketing data after the data is subjected to dimension reduction, and obtaining a corresponding real-time marketing effect grade label;
and storing the real-time marketing data after the data is reduced in size into a corresponding marketing effect sub-database of the marketing service database according to the real-time marketing effect grade label.
10. A data processing system for managing marketing data for implementing a data processing method as claimed in any one of claims 1 to 9, characterized in that: the system comprises a data initialization processing unit, a database construction unit and a real-time marketing data processing unit which are connected in sequence;
The data initialization processing unit is used for collecting the historical marketing big data, carrying out data dimension reduction on the historical marketing big data to obtain a corresponding key marketing index set and a data dimension-reduced historical marketing data set, setting a corresponding historical marketing service type label and a historical marketing effect grade label for each data dimension-reduced historical marketing data, and obtaining a marketing effect classification sample set containing a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels;
The database construction unit is used for constructing a plurality of corresponding marketing service databases and a plurality of marketing effect sub-databases according to a plurality of historical marketing service type labels and a plurality of historical marketing effect grade labels, storing each marketing effect classification sample into the corresponding marketing effect sub-database of the corresponding marketing service database, and constructing a corresponding marketing effect classification model by using a deep learning algorithm according to a marketing effect classification sample set;
The real-time marketing data processing unit is used for collecting real-time marketing data, carrying out data dimension reduction on the real-time marketing data according to the key marketing index set, obtaining real-time marketing data after the corresponding data dimension reduction, obtaining real-time marketing service type labels of the real-time marketing data after the data dimension reduction, using a marketing effect classification model, obtaining real-time marketing effect grade labels of the real-time marketing data after the data dimension reduction, and storing the real-time marketing data after the data dimension reduction into a corresponding marketing effect sub-database of a corresponding marketing service database according to the real-time marketing service type labels and the real-time marketing effect grade labels.
CN202410276842.3A 2024-03-12 2024-03-12 Data processing method and system for managing marketing data Pending CN118037332A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202410276842.3A CN118037332A (en) 2024-03-12 2024-03-12 Data processing method and system for managing marketing data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202410276842.3A CN118037332A (en) 2024-03-12 2024-03-12 Data processing method and system for managing marketing data

Publications (1)

Publication Number Publication Date
CN118037332A true CN118037332A (en) 2024-05-14

Family

ID=90998502

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202410276842.3A Pending CN118037332A (en) 2024-03-12 2024-03-12 Data processing method and system for managing marketing data

Country Status (1)

Country Link
CN (1) CN118037332A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118297640A (en) * 2024-06-06 2024-07-05 南京信息工程大学 Product marketing management system and method based on big data

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118297640A (en) * 2024-06-06 2024-07-05 南京信息工程大学 Product marketing management system and method based on big data

Similar Documents

Publication Publication Date Title
US7844566B2 (en) Latent semantic clustering
US20060242190A1 (en) Latent semantic taxonomy generation
CN111897963B (en) Commodity classification method based on text information and machine learning
JP7499360B2 (en) Obtaining Compact Representations and Time Series Segments with Deep Learning
CN112184484A (en) Differentiated service method and system for power users
CN118037332A (en) Data processing method and system for managing marketing data
CN118261677A (en) Client information analysis management system and method based on data mining
Sawalha et al. Towards an efficient big data management schema for IoT
CN114637846A (en) Video data processing method, video data processing device, computer equipment and storage medium
CN116629258B (en) Structured analysis method and system for judicial document based on complex information item data
CN117056510A (en) Automatic collecting method for multi-element social contradiction dispute information
Laaroussi et al. A novel hybrid deep learning approach for tourism demand forecasting
He et al. Semisupervised network embedding with differentiable deep quantization
CN115719182A (en) Method and system for quickly generating power data index
KR102358357B1 (en) Estimating apparatus for market size, and control method thereof
Chen et al. LPR‐MLP: A Novel Health Prediction Model for Transmission Lines in Grid Sensor Networks
Luo et al. A comparison of som based document categorization systems
AlKarawi et al. Data Reduction Techniques: A Comparative Study
CN116932487B (en) Quantized data analysis method and system based on data paragraph division
Samarajeewa et al. An artificial intelligence framework for explainable drift detection in energy forecasting
CN118331952B (en) Financial data cleaning management system and method based on big data
CN117556118B (en) Visual recommendation system and method based on scientific research big data prediction
US20240370732A1 (en) Method and an apparatus for functional model generation
Zhang et al. [Retracted] Impact of the Validity Analysis Model and Multirelational Data Clustering Based on the Trust Probability
Mondal et al. A survey of multimodal event detection based on data fusion

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