Crowd fund measuring method based on data intelligent label automation

Through the automation method based on intelligent data labels, the traditional test method is time-consuming and labor-intensive and inaccurate results are solved, efficient and accurate market feedback is achieved, and the product market competitiveness is enhanced.

CN119004077APending Publication Date: 2024-11-22SHANGHAI YOUMOU INFORMATION TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411211936.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-30
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The traditional manual test method is time-consuming and labor-intensive, and it is difficult to cover a wide enough population sample, resulting in inaccurate test results.

Method used

Adopt an automated method based on data intelligent labels, collect user behavior data through crawling technology, generate user tags using machine learning algorithms, perform data preprocessing and feature extraction, train machine learning models, design test experiments and use RPA technology to automate execution, analyze and provide feedback.

Benefits of technology

It greatly improves the efficiency and accuracy of the test model, reduces the influence of human factors, improves the objectivity and reliability of data, can better meet personalized market needs, and improves the market competitiveness of products.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119004077A_ABST
    Figure CN119004077A_ABST
Patent Text Reader

Abstract

The invention discloses a crowd fund measurement method based on data intelligent label automation. The crowd fund measurement method based on data intelligent label automation comprises the following steps: step A, data acquisition; step B, generating a user tag; c, preparing data; step D, feature extraction; e, model training; f, comparing and analyzing; step G, result explanation; step H, survey experiment design; step I, result analysis and feedback; according to the invention, the efficiency and accuracy of money measurement can be greatly improved, and merchants can be helped to quickly obtain valuable market feedback. The influence of human factors is reduced through an automatic process, and the objectivity and reliability of data are improved. The personalized market requirements can be better met, and the market competitiveness of the product is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of intelligent measurement of electronic commerce, and in particular to a method for measuring the amount of money among a group of people based on automated data intelligent labeling. Background Art

[0002] In the e-commerce industry, in order to better understand consumer preferences and quickly iterate products, merchants need to conduct market tests on new styles, i.e., test styles. The traditional manual testing method is not only time-consuming and labor-intensive, but also often fails to cover a wide enough population sample, resulting in inaccurate testing results. Therefore, developing a method for testing styles based on data intelligent labeling automation has become an urgent problem to be solved. Summary of the invention

[0003] The purpose of the invention is to provide an automated crowd measurement method based on data intelligent tags, which solves the problems in the background technology.

[0004] The present invention is implemented as follows: a method for measuring money among people based on data intelligent label automation, the method for measuring money among people based on data intelligent label automation comprising the following steps:

[0005] Step A: Data collection;

[0006] Step B: User tag generation;

[0007] Step C: Data preparation;

[0008] Step D: feature extraction;

[0009] Step E: Model training;

[0010] Step F: Comparative analysis;

[0011] Step G: Interpretation of results:

[0012] Step H: Test experiment design:

[0013] Step I: Result analysis and feedback.

[0014] A further technical solution of the present invention is: the data collection in step A is: using crawler technology to automatically collect user behavior data from multiple channels such as social networks and e-commerce platforms; cleaning and preprocessing the collected data to ensure data quality.

[0015] A further technical solution of the present invention is: the user tags in step B are generated by: using a machine learning algorithm to generate user tags based on multi-dimensional data such as user purchase history, browsing history, evaluation feedback, etc.; and dividing different user groups through cluster analysis of user tags.

[0016] A further technical solution of the present invention is: the data preparation in step C is: collecting data sets A and B to be compared; preprocessing the data sets, including but not limited to data cleaning, missing value processing, outlier detection, etc.

[0017] A further technical solution of the present invention is: the feature extraction in step D is: using feature engineering technology to extract meaningful features from the data set; optionally using dimensionality reduction technology (such as PCA) to reduce the number of features and reduce computational complexity.

[0018] A further technical solution of the present invention is: the model training in step E is: selecting a suitable machine learning algorithm (such as vector machine SVM, random forest RF, deep neural network DNN, etc.) as a comparison model; using a labeled data set to train the model so that the model can learn the distinguishing features between different data sets.

[0019] A further technical solution of the present invention is: the comparative analysis in step F is: inputting the data sets to be compared into the trained model, outputting the classification results or similarity scores of each data set; analyzing the results output by the model to determine the similarities and differences between the two data sets.

[0020] A further technical solution of the present invention is: the result in step G is explained as: based on the result output by the model, the specific differences between the data sets are explained in combination with domain knowledge; a visualization tool is provided to display the comparison results for easy understanding and use by users.

[0021] A further technical solution of the present invention is: the design of the product testing experiment in step H is: according to the characteristics of the user group, different product testing experiments are designed, including but not limited to tests on price, color, style, etc.; and the product testing experiment is automatically performed using RPA (Robotic Process Automation) technology.

[0022] A further technical solution of the present invention is: the result analysis feedback in step I is: collecting data generated by the style testing experiment, and using data analysis technology to evaluate the market performance of different styles; based on the analysis results, generating a report and providing improvement suggestions to the merchant.

[0023] Beneficial effects of the present invention: The present invention can greatly improve the efficiency and accuracy of product testing, and help merchants quickly obtain valuable market feedback. The automated process reduces the impact of human factors, and improves the objectivity and reliability of data. It can better meet personalized market needs and improve the market competitiveness of products. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1It is a flow chart of a method for automating crowd measurement based on data intelligent tags provided by the present invention. DETAILED DESCRIPTION

[0025] Embodiment 1: Figure 1 A method for measuring money among people based on data intelligent tag automation is shown, which includes the following steps: The method for measuring money among people based on data intelligent tag automation includes the following steps:

[0026] Step A: Data collection;

[0027] Step B: User tag generation;

[0028] Step C: Data preparation;

[0029] Step D: feature extraction;

[0030] Step E: Model training;

[0031] Step F: Comparative analysis;

[0032] Step G: Interpretation of results:

[0033] Step H: Test experiment design:

[0034] Step I: Result analysis and feedback; Data collection in step A is: using crawler technology to automatically collect user behavior data from multiple channels such as social networks and e-commerce platforms; cleaning and preprocessing the collected data to ensure data quality; User tag generation in step B is: using machine learning algorithms to generate user tags based on multi-dimensional data such as user purchase history, browsing history, evaluation feedback, etc.; Different user groups are divided through cluster analysis of user tags; Data preparation in step C is: collecting data sets A and B to be compared; preprocessing the data sets, including but not limited to data cleaning, missing value processing, outlier detection, etc.; Feature extraction in step D is: using feature engineering technology to extract meaningful features from the data set; Dimensionality reduction technology (such as PCA) is used to reduce the number of features and reduce computational complexity; Model training in step E The method comprises the following steps: selecting a suitable machine learning algorithm (such as vector machine (SVM), random forest (RF), deep neural network (DNN), etc.) as a comparison model; training the model with a labeled data set so that the model can learn the distinguishing features between different data sets; the comparison analysis in step F comprises: inputting the data set to be compared into the trained model, and outputting the classification result or similarity score of each data set; analyzing the results output by the model to determine the similarities and differences between the two data sets; the result interpretation in step G comprises: based on the results output by the model, interpreting the specific differences between the data sets in combination with domain knowledge; providing a visualization tool to display the comparison results for easy understanding and use by users; the design of the test model experiment in step H comprises: designing different test model experiments according to the characteristics of the user group, including but not limited to tests on price, color, style, etc.; automatically executing the test model experiment using RPA (Robotic Process Automation) technology; the result analysis feedback in step I comprises: collecting the data generated by the test model experiment, and using data analysis technology to evaluate the market performance of different styles; generating a report based on the analysis results and providing improvement suggestions to merchants; the present invention can greatly improve the efficiency and accuracy of test models and help merchants quickly obtain valuable market feedback. The automated process reduces the impact of human factors and improves the objectivity and reliability of data, which can better meet personalized market needs and improve the market competitiveness of products.

[0035] Data collection, characterized by: using crawler technology to capture users' shopping history, browsing behavior and other data from e-commerce platforms

[0036] 2) Intelligent label generation, which is characterized by: using machine learning algorithms (such as cluster analysis, decision trees, etc.) to classify different styles of goods according to users' shopping preferences and shopping styles, and assigning corresponding feature vectors (such as color, size, material, etc.) to each style

[0037] 3) Tag matching, which is characterized by: predicting the user's preference for different styles of goods by calculating the similarity between the user's smart tag and the feature vector of the product style;

[0038] 4) Model testing strategy formulation, which is characterized by: based on the matching results, selecting the product models that are most likely to be popular with users for promotion or generating decision results.

[0039] Step 1: Organize the product styles and specific instructions that need to be tested and investigated;

[0040] Step 2: Through RPA, batch import of basic activity information and target groups can complete batch automatic creation, release, delivery, data tracing and other actions of smart testing activities. Capture the real preferences of consumers to guide production and stocking, appearance selection, selection of main models, main picture design and product combination and other scenario applications:

[0041] Create a test campaign

[0042] Enter the activity group name; select the name of the third-party account to be set up;

[0043] Select the corresponding project (only projects under third-party accounts and for which the current operating account has permissions can be selected)

[0044] Select a release time

[0045] Release all immediately: RPA executes activities immediately and releases them in batches

[0046] Manual batch release: After the activity group is created, it will not be released for the time being. You can manually select and release it in batches later.

[0047] Scheduled release: You can select a specific time point in the future to schedule activity releases

[0048] Step 3: Automatically upload data:

[0049] To upload the "Batch Add Smart Test Template", you need to download the template first, and then fill in the content according to the template requirements. The template has notes on filling in each field, which is divided into two sections: activity content and delivery content.

[0050] 2. Upload the compressed image package:

[0051] You need to download the compressed package template first, and name the pictures and set the folders according to the rules of the compressed package.

[0052] Operators create and name folders and images according to the instructions to facilitate the successful creation and publishing of the campaign.

[0053] Finally upload the compressed folder, the directory should be: activity group name -> activity name-group name -> picture;

[0054] One activity group corresponds to one compressed package. Smart testing activities need to be distinguished according to folders in the compressed package. Please name the folder with the main title of the activity.

[0055] Step 4: View delivery data

[0056] 1 View delivery: Click [View delivery] to view the specific delivery information of the activity, including the target group, delivery status, recovery progress, channel type and delivery time.

[0057] 2. For each crowd package of each activity, RPA will automatically generate data reports and PPT key materials, which the operator can directly download and view.

[0058] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for crowd measurement based on data intelligent label automation, characterized in that: The method for automating crowd measurement based on data intelligent tags includes the following steps: Step A: Data collection; Step B: User tag generation; Step C: Data preparation; Step D: feature extraction; Step E: Model training; Step F: Comparative analysis; Step G: Interpretation of results: Step H: Test experiment design: Step I: Result analysis and feedback.

2. According to claim 1, a method for crowd measurement based on data intelligent label automation is characterized in that: The data collection in step A is as follows: using crawler technology to automatically collect user behavior data from multiple channels such as social networks and e-commerce platforms; cleaning and preprocessing the collected data to ensure data quality.

3. The method for crowd measurement based on data intelligent label automation according to claim 1 is characterized in that: The user tags in step B are generated by: using a machine learning algorithm to generate user tags based on multi-dimensional data such as the user's purchase history, browsing history, evaluation feedback, etc.; and dividing different user groups through cluster analysis of user tags.

4. The method for crowd measurement based on data intelligent label automation according to claim 1 is characterized in that: The data preparation in step C is as follows: collecting the data sets A and B to be compared; and preprocessing the data sets, including but not limited to data cleaning, missing value processing, outlier detection, etc.

5. The method for crowd measurement based on data intelligent label automation according to claim 1 is characterized in that: The feature extraction in step D is: extracting meaningful features from the data set using feature engineering technology; Dimensionality reduction techniques are used to reduce the number of features and reduce computational complexity.

6. The method for crowd measurement based on data intelligent label automation according to claim 1 is characterized in that: The model training in step E is as follows: selecting a suitable machine learning algorithm as a comparison model; and training the model using a labeled data set so that the model can learn distinguishing features between different data sets.

7. The method for crowd measurement based on data intelligent label automation according to claim 1 is characterized in that: The comparative analysis in step F is as follows: inputting the data sets to be compared into the trained model, outputting the classification results or similarity scores of each data set; and analyzing the results output by the model to determine the similarities and differences between the two data sets.

8. The method for crowd measurement based on data intelligent label automation according to claim 1 is characterized in that: The result interpretation in step G is as follows: based on the results output by the model, the specific differences between the data sets are explained in combination with domain knowledge; a visualization tool is provided to display the comparison results for easy understanding and use by users.

9. The method for crowd measurement based on data intelligent label automation according to claim 1 is characterized in that: The design of the product testing experiment in step H is as follows: according to the characteristics of the user group, different product testing experiments are designed, including but not limited to tests on price, color, style, etc.; and the product testing experiments are automatically executed using RPA technology.

10. The method for crowd measurement based on data intelligent label automation according to claim 1 is characterized in that: The result analysis feedback in step I is: collecting data generated by the test model experiment, and using data analysis technology to evaluate the market performance of different models; based on the analysis results, generating a report and providing improvement suggestions to merchants.