Intelligent object pushing method and system based on big data analysis

Abstract

The invention discloses an intelligent object pushing method and system based on big data analysis, and relates to the technical field of intelligent object pushing, and the method comprises the following steps: collecting investment behavior data, market dynamic data and user preference data of a user in real time, carrying out the dynamic analysis of the investment preference of the user in combination with historical data, and obtaining an intelligent object pushing result; and generating a real-time investment preference model of the user. According to the method, user behaviors and market data are collected in real time and dynamically analyzed, a model capable of reflecting the newest investment preference of the user in time is generated, and it is ensured that the recommended target is matched with the market trend in combination with multi-dimensional market monitoring and user preference comparison. And the model is trained and optimized online, so that the model adapts to market changes and user requirements, and recommendation accuracy and satisfaction are improved through personalized pushing. Self-adaptive optimization is achieved through user feedback, recommendation accuracy and investment income are remarkably improved, user requirements and market changes are dynamically responded, and system effectiveness and user satisfaction are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of target intelligent push, and in particular to a target intelligent push method and system based on big data analysis. Background Art

[0002] The intelligent push system for targets based on big data analysis is an innovative platform that combines advanced data analysis and machine learning technology. The system builds an evaluation model for investment targets by collecting and integrating multi-source data such as market data, corporate operating data, and financial statements. Using machine learning algorithms, the system can conduct in-depth analysis of the data to identify targets with high investment potential. The identified targets will be intelligently pushed to users based on their personalized needs, such as investment preferences and risk tolerance. This intelligent push not only helps users quickly find targets that meet their investment strategies, but also optimizes the recommended results through real-time data updates and feedback, thereby significantly improving the accuracy and efficiency of investment decisions. By providing these accurate and personalized target recommendations, the system effectively supports users' investment decision-making process, reduces investors' time costs and information screening pressure, and enables them to make more informed investment choices in a complex market environment.

[0003] The prior art has the following deficiencies:

[0004] The target recommendation systems in the existing technology are usually not responsive enough when dealing with changes in user preferences. Traditional systems rely on static models to analyze users' historical data and investment preferences, but this method is difficult to capture and adapt to users' dynamic adjustment needs as the market changes in real time. As a result, the recommendation system is prone to lags or inaccuracies, especially in situations where the market fluctuates frequently. The targets recommended by the system may no longer meet the user's latest investment intentions, seriously affecting the quality of investment decisions and the maximization of returns. This problem is particularly prominent in the big data environment, because the amount of data is large and changes rapidly. The existing technology lacks a flexible model adjustment mechanism and cannot optimize the recommendation results in a timely manner, which greatly reduces the effectiveness of the recommendation and user satisfaction.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a method and system for intelligently pushing targets based on big data analysis. By real-time collection and dynamic analysis of user behavior and market data, a model that can timely reflect the user's latest investment preferences is generated, and multi-dimensional market monitoring and user preference comparison are combined to ensure that the recommended targets match market trends. The model is trained and optimized online to adapt to market changes and user needs, and personalized push improves recommendation accuracy and satisfaction. Adaptive optimization is achieved through user feedback, which significantly improves recommendation accuracy and investment returns, dynamically responds to user needs and market changes, and improves system effectiveness and user satisfaction to solve the problems in the above-mentioned background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a method for intelligently pushing targets based on big data analysis, comprising the following steps:

[0008] By collecting users' investment behavior data, market dynamics data and user preference data in real time, and combining historical data to dynamically analyze users' investment preferences, a real-time investment preference model for users is generated;

[0009] Conduct multi-dimensional real-time monitoring and processing of market information from multiple data sources, build a market dynamics database, identify market trends and hot spots through market data analysis, and compare this information with the user's real-time investment preference model to evaluate the impact of current market conditions on the user's potential investment targets;

[0010] Based on user and market data, we use machine learning algorithms to train and optimize the recommendation model online. According to real-time feedback from users and changes in market data, we update and adjust the machine learning model in real time to ensure that the recommendation algorithm is always consistent with users' current investment preferences and market dynamics.

[0011] Based on the real-time optimized recommendation model, a list of targets that meet the user's current investment preferences is generated and sorted by priority. During the recommendation process, the targets are personalized and customized based on the user's risk tolerance and investment strategy to ensure that the recommended targets match the user's needs;

[0012] Establish a user interactive feedback mechanism, collect user satisfaction with recommendation results and actual investment behavior data, and automatically adjust and optimize user investment preference models and recommendation models based on user feedback information to further improve the system's recommendation accuracy and user satisfaction and achieve adaptive optimization of the system.

[0013] Preferably, the specific steps for generating the user's real-time investment preference model are as follows:

[0014] Collect users' investment behavior data, market dynamics data and user preference data in real time from multiple sources;

[0015] After data collection is completed, historical data is used as a basis to generate a preliminary user preference model;

[0016] Based on pre-built benchmark models, dynamic analysis of user behavior and market data collected in real time is performed;

[0017] After the real-time analysis and model update are completed, the model will be further optimized through the user interaction feedback mechanism.

[0018] Preferably, the impact of the current market conditions on the user's potential investment targets is evaluated, and the specific steps are as follows:

[0019] Collect market information from multiple data sources in real time;

[0020] After data cleaning, the collected market data is analyzed and modeled in multiple dimensions. Multi-dimensional analysis includes slicing and filtering market data from multiple perspectives.

[0021] Based on the results of multi-dimensional analysis, we use forecasting models and pattern recognition technology to identify market trends and hot spots in real time;

[0022] After identifying market trends and hot spots, this information is compared with the user's real-time investment preference model. During the comparison process, it is analyzed how the current market conditions affect the user's potential investment targets.

[0023] Preferably, based on user and market data, a machine learning algorithm is used to train and optimize the recommendation model online. The machine learning model is updated and adjusted in real time according to real-time feedback from users and changes in market data to ensure that the recommendation algorithm is always consistent with the user's current investment preferences and market dynamics. The specific steps are as follows:

[0024] Extract key features from users’ investment behavior data and market dynamics data;

[0025] After feature extraction is completed, the recommendation model is trained online using the extracted feature data;

[0026] A real-time feedback mechanism has been established to collect users’ actual behavior and feedback information immediately after they make investment decisions;

[0027] Based on the online model training and feedback adjustment, model evaluation and adaptive optimization are performed regularly.

[0028] Preferably, based on the real-time optimized recommendation model, a list of targets that meet the user's current investment preferences is generated and sorted by priority. During the recommendation process, the targets are personalized and customized based on the user's risk tolerance and investment strategy to ensure that the recommended targets match the user's needs. The specific steps are as follows:

[0029] Filter out targets that match the user's current investment preferences and market trends from a massive library of investment targets;

[0030] After generating a preliminary list of targets, the system prioritizes these targets based on the user's investment objectives, risk tolerance, and market conditions;

[0031] After the priority sorting is completed, the target will be screened more deeply and personalized, and the screening process is based on the user's specific investment strategy and personalized needs;

[0032] After the target is pushed, we continuously monitor the user's response and actual investment behavior, and collect relevant feedback data. By analyzing the feedback data, we evaluate the effectiveness of the recommendation. Based on the evaluation results, we continuously optimize the recommendation model, adjust the recommendation strategy, re-evaluate the target priority or improve the screening algorithm to ensure that user needs are accurately met in subsequent recommendations.

[0033] Preferably, a user interactive feedback mechanism is established to collect user satisfaction with recommendation results and actual investment behavior data. According to user feedback information, the user investment preference model and recommendation model are automatically adjusted and optimized to further improve the system's recommendation accuracy and user satisfaction, and realize the system's adaptive optimization. The specific steps are as follows:

[0034] During the interaction between users and the recommendation system, continuously collect user feedback data on the recommendation results;

[0035] After collecting user feedback data, we will conduct a detailed analysis of the data to identify users’ preferences and dissatisfaction with the recommendation results, and compare the feedback data with the output of the recommendation model to evaluate the accuracy and effectiveness of the current model;

[0036] Based on the analysis results of feedback data, automatically adjust and optimize the user investment preference model and recommendation model, use machine learning algorithms to retrain and adjust model parameters, and correct model deviations exposed by feedback data;

[0037] After the model is adjusted and optimized, new user feedback is continuously monitored to form an adaptive optimization loop. The system regularly or in real time evaluates the performance of the adjusted model, observes changes in user satisfaction and improvements in recommendation accuracy. If the positivity of user feedback increases after model optimization, the current adjustment strategy will be maintained; if the problem still exists, further iterative optimization will be performed until the ideal recommendation effect is achieved.

[0038] A target intelligent push system based on big data analysis, including a user preference modeling module, a market dynamic monitoring module, a recommendation model optimization module, a personalized recommendation module, and a feedback and adaptive optimization module;

[0039] The user preference modeling module collects the user's investment behavior data, market dynamics data and user preference data in real time, combines historical data to dynamically analyze the user's investment preferences, and generates a real-time investment preference model for the user;

[0040] The market dynamics monitoring module monitors and processes market information from multiple data sources in real time in multiple dimensions, builds a market dynamics database, identifies market trends and hot spots through market data analysis, and compares this information with the user's real-time investment preference model to assess the impact of current market conditions on the user's potential investment targets;

[0041] The recommendation model optimization module uses machine learning algorithms to train and optimize the recommendation model online based on user and market data. It updates and adjusts the machine learning model in real time according to real-time user feedback and changes in market data to ensure that the recommendation algorithm is always consistent with the user's current investment preferences and market dynamics.

[0042] The personalized recommendation module generates a list of targets that meet the user's current investment preferences based on a real-time optimized recommendation model and sorts them by priority. During the recommendation process, the module conducts personalized screening and customized push of targets based on the user's risk tolerance and investment strategy to ensure that the recommended targets match the user's needs.

[0043] The feedback and adaptive optimization module establishes a user interactive feedback mechanism, collects user satisfaction with recommendation results and actual investment behavior data, and automatically adjusts and optimizes user investment preference models and recommendation models based on user feedback information, further improving the system's recommendation accuracy and user satisfaction, and achieving system adaptive optimization.

[0044] Preferably, the user preference modeling module is specifically used to: collect user investment behavior data, market dynamics data and user preference data in real time from multiple sources; after data collection is completed, use historical data as a basis to generate a preliminary user preference model; based on a pre-built benchmark model, dynamically analyze the user behavior and market data collected in real time; after the real-time analysis and model update are completed, the model will be further optimized through a user interaction feedback mechanism.

[0045] Preferably, the market dynamics monitoring module is specifically used to: collect market information from multiple data sources in real time; after completing data cleaning, perform multi-dimensional analysis and modeling on the collected market data, and the multi-dimensional analysis includes slicing and screening the market data from multiple perspectives; based on the results of the multi-dimensional analysis, use predictive models and pattern recognition technology to identify trends and hot spots in the market in real time; after identifying market trends and hot spots, compare this information with the user's real-time investment preference model, and during the comparison process, analyze how the current market conditions affect the user's potential investment targets.

[0046] Preferably, the recommendation model optimization module is specifically used to: extract key features from the user's investment behavior data and market dynamics data; after completing the feature extraction, use the extracted feature data to perform online training on the recommendation model; establish a real-time feedback mechanism to immediately collect the user's actual behavior and feedback information after the user makes an investment decision; based on the model's online training and feedback adjustment, regularly perform model evaluation and adaptive optimization.

[0047] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0048] The present invention generates a model that promptly reflects the user's latest investment preferences by real-time collection and dynamic analysis of user behavior and market data. At the same time, it monitors market information in real time from multiple dimensions and compares it with user preferences to ensure that the recommended target is highly matched with the current market trend. Online training and optimization of the recommendation model enables it to continuously adapt to market changes and user needs. In addition, personalized screening and customized push further improve the accuracy of recommendations and user satisfaction. Through the user interactive feedback mechanism, the system realizes adaptive optimization, continuously adjusts and improves the recommendation strategy, thereby significantly improving the accuracy of the recommendation results and the user's investment returns, thereby enabling the recommendation system to dynamically respond to user needs and market changes, providing more accurate and efficient investment target recommendation services, and greatly improving the effectiveness of the system and user satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0050] Figure 1 The present invention is a method flow chart of the target intelligent push method based on big data analysis.

[0051] Figure 2 It is a module schematic diagram of the target intelligent push system based on big data analysis of the present invention. DETAILED DESCRIPTION

[0052] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0053] The present invention provides Figure 1 The target intelligent push method based on big data analysis shown includes the following steps:

[0054] By collecting users' investment behavior data, market dynamics data and user preference data in real time, and combining historical data to dynamically analyze users' investment preferences, a real-time investment preference model for users is generated;

[0055] The specific steps for generating a user's real-time investment preference model are as follows:

[0056] Collect users' investment behavior data, market dynamics data and user preference data in real time from multiple sources;

[0057] These data sources include but are not limited to users' transaction records, browsing history, portfolio adjustments on the trading platform, as well as real-time financial data in the market and preference data expressed by users on other channels. The collected data will be pre-processed to remove noise, fill in missing values, and be integrated in a unified format so that the data can be effectively used in subsequent steps. The key to this step is to ensure the accuracy, completeness and timeliness of the data in order to provide a reliable basis for dynamic analysis.

[0058] After data collection is completed, historical data is used as a basis to generate a preliminary user preference model;

[0059] This benchmark model reflects the user's investment behavior and preference trends over the past period of time, including investment style, risk preference, industry preference, etc. The system will analyze multiple dimensions such as the user's trading frequency, holding time, and rate of return, and combine these historical data with market fluctuations to form a preliminary user preference model. This model provides a reference framework for subsequent real-time updates and dynamic adjustments to ensure that changes in user preferences can be accurately tracked.

[0060] Based on pre-built benchmark models, dynamic analysis of user behavior and market data collected in real time is performed;

[0061] The analysis process uses machine learning and data mining technology to identify the changing trends of users' investment preferences and update the user preference model in real time. For example, when the market fluctuates, users may adjust their investment strategies. The system can identify these behavioral changes and update the model parameters accordingly to adjust the user's risk preference and investment direction. The core of this step is to achieve real-time data analysis and dynamic adjustment of the model to ensure that the model can always reflect the user's latest investment intentions.

[0062] After the real-time analysis and model update are completed, the model will be further optimized through the user interaction feedback mechanism;

[0063] The system will analyze the actual investment behavior and feedback data of users after using the recommendation system to determine the accuracy and effectiveness of the current model. If it is found that there is a large deviation between the recommendation results and the user's actual investment behavior, the system will adjust the model's parameters and re-evaluate the model's weight distribution to ensure that the model can better adapt to the user's investment style. In addition, the system will conduct regular self-inspection and evaluation, and optimize the model based on long-term feedback data, so that it gradually stabilizes and accurately reflects the user's investment preferences and strategy adjustments.

[0064] Conduct multi-dimensional real-time monitoring and processing of market information from multiple data sources, build a market dynamics database, identify market trends and hot spots through market data analysis, and compare this information with the user's real-time investment preference model to evaluate the impact of current market conditions on the user's potential investment targets;

[0065] Evaluate the impact of current market conditions on your potential investment targets. The specific steps are as follows:

[0066] Collect market information from multiple data sources in real time;

[0067] These data sources include stock markets, bond markets, commodity futures markets, macroeconomic indicators, industry reports, news media, and social media. The types of data collected are diverse, including price changes, trading volumes, economic data releases, industry trends, and public opinion. In order to ensure the reliability and consistency of the data, the system will clean the data after collection, remove redundant information, correct data errors, unify the data format, and mark the data timestamp. The focus of this step is to ensure the quality of the data so that subsequent analysis is based on accurate and timely market information.

[0068] After data cleaning, the collected market data is analyzed and modeled in multiple dimensions. Multi-dimensional analysis includes slicing and filtering market data from multiple perspectives.

[0069] Slice and filter market data from multiple perspectives, such as industry, region, time, and volume. These analyses not only help the system identify short-term market fluctuations, but also discover long-term market trends and potential hot spots. By performing correlation analysis on market data of different dimensions, the system can build comprehensive market dynamic models, which will become the basis for identifying market trends.

[0070] Based on the results of multi-dimensional analysis, we use forecasting models and pattern recognition technology to identify market trends and hot spots in real time;

[0071] These trends may include the rise or fall of a certain industry, the increase in volatility in a specific market, the impact of economic data releases on the market, etc. By identifying these trends and hot spots, the system can predict the direction of the market in advance and classify and prioritize this information. The key to this step is to convert complex market data into clear trend information for further comparison with the user's investment preference model.

[0072] After identifying market trends and hot spots, this information is compared with the user's real-time investment preference model. During the comparison process, it analyzes how the current market conditions affect the user's potential investment targets;

[0073] For example, whether a hot spot in the market is related to the industry or asset class that the user is concerned about, or whether the market trend is in line with the user's risk tolerance and investment strategy. Through this comparison, the system can evaluate the impact of the current market environment on the user's investment decision, and then generate personalized target recommendations. This step ensures that the targets recommended by the system are in line with the current market conditions and meet the user's investment needs, thereby improving the accuracy and effectiveness of the recommendations.

[0074] Based on user and market data, we use machine learning algorithms to train and optimize the recommendation model online. According to real-time feedback from users and changes in market data, we update and adjust the machine learning model in real time to ensure that the recommendation algorithm is always consistent with users' current investment preferences and market dynamics.

[0075] Based on user and market data, we use machine learning algorithms to train and optimize the recommendation model online. According to real-time feedback from users and changes in market data, we update and adjust the machine learning model in real time to ensure that the recommendation algorithm is always consistent with users' current investment preferences and market dynamics. The specific steps are as follows:

[0076] Extract key features from users’ investment behavior data and market dynamics data;

[0077] These features include the user's trading history, positions, investment preferences, market volatility, and other variables related to investment decisions. In order to ensure that the data can be effectively used by the machine learning model, the system will perform data preprocessing, such as normalization, missing value processing, and data enhancement. In addition, the system will select the features that best reflect the user's investment behavior and market status based on business needs and model requirements, establish feature engineering, and extract a multi-dimensional feature set as the basis for model training. The core of this step is to provide accurate, rich, and meaningful data input for the machine learning model to improve the model's training effect.

[0078] After feature extraction is completed, the recommendation model is trained online using the extracted feature data;

[0079] Online training is different from traditional batch training. It can update model parameters in real time as new data is continuously input. The system inputs the user's latest investment behavior and market data into the machine learning model, and adjusts the model's weights and parameters in real time, thereby ensuring that the model can quickly adapt to market changes and dynamic adjustments of user preferences. Commonly used online training algorithms include stochastic gradient descent (SGD), online Bayesian update, etc. The purpose of online training is to enable the model to reflect the latest data changes in a timely manner and improve the accuracy and real-time performance of recommendation results.

[0080] In order to ensure that the recommendation model is always consistent with the user's current investment preferences and market dynamics, the system has established a real-time feedback mechanism to collect users' actual behavior and feedback information immediately after they make investment decisions;

[0081] Collect users' actual behaviors and feedback information, such as whether users accept the recommended targets, whether the performance after investment meets expectations, etc. These feedback data will be immediately sent back to the model as additional input for online learning to further optimize and adjust model parameters. Through this closed-loop feedback mechanism, the system can quickly adjust the recommendation model according to the actual usage of users and real-time market fluctuations, ensuring that the recommendation results always meet user needs.

[0082] Based on the online model training and feedback adjustment, regular model evaluation and adaptive optimization are performed;

[0083] Model evaluation mainly uses a series of performance indicators, such as accuracy, recall, F1 score, user satisfaction, etc., to determine whether the performance of the current model meets expectations. If the model performance is found to have degraded or deviated, the system will automatically trigger the adaptive optimization mechanism to adjust the model architecture, feature selection or training parameters. At the same time, the system may use methods such as ensemble learning to combine the prediction results of multiple models to improve the stability and reliability of the overall recommendation. The goal of adaptive optimization is to ensure that the recommendation model remains efficient and accurate during long-term use, and to continuously improve the recommendation effect as the market and user behavior change.

[0084] Based on the real-time optimized recommendation model, a list of targets that meet the user's current investment preferences is generated and sorted by priority. During the recommendation process, the targets are personalized and customized based on the user's risk tolerance and investment strategy to ensure that the recommended targets match the user's needs;

[0085] Based on the real-time optimized recommendation model, a list of targets that meet the user's current investment preferences is generated and sorted by priority. During the recommendation process, the targets are personalized and customized based on the user's risk tolerance and investment strategy to ensure that the recommended targets match the user's needs. The specific steps are as follows:

[0086] Filter out targets that match the user's current investment preferences and market trends from a massive library of investment targets;

[0087] This process uses a real-time optimized recommendation model, combined with the user's historical investment behavior, market dynamics data and personalized investment preferences, to filter out a group of preliminary matching targets. The screening is based on key factors such as industry, region, asset class, and historical performance. The system generates a list of preliminary matching targets by weighted scoring of these factors, laying the foundation for subsequent priority sorting and further screening.

[0088] After generating a preliminary list of targets, the system prioritizes these targets based on the user's investment objectives, risk tolerance, and market conditions;

[0089] When sorting, the system will consider factors such as the user's risk preference (such as conservative, aggressive), expected returns, investment period, etc., and assign weights based on the current market environment (such as volatility, liquidity). The ranking of targets usually adopts a weighted scoring method, which sums up the scores of various factors. The higher the score, the higher the priority. This step ensures that users can first see the targets that best match their current investment needs, which improves the pertinence and effectiveness of the recommendation.

[0090] After the priority sorting is completed, the target will be screened more deeply and personalized according to the user's specific investment strategy and personalized needs;

[0091] This screening process further screens and adjusts the target based on the user's specific investment strategy and personalized needs, such as preference for certain industries or risk aversion needs. The system will analyze the specific investment strategy provided by the user, such as diversified investment, concentrated investment, short-term speculation, etc., and compare it with the attributes of the target to screen out the target that best matches the user's strategy. The screened targets will be pushed to the user in a customized manner, such as through notifications, recommendation lists, or personalized investment advice.

[0092] After the target is pushed, we continuously monitor the user's response and actual investment behavior, and collect relevant feedback data. By analyzing this data, we evaluate the effectiveness of the recommendation. Based on the evaluation results, we continuously optimize the recommendation model, adjust the recommendation strategy, re-evaluate the target priority, or improve the screening algorithm to ensure that user needs are accurately met in subsequent recommendations;

[0093] By analyzing this data, the system can evaluate the effectiveness of recommendations, such as user click-through rate, investment success rate, user satisfaction and other indicators. Based on the evaluation results, the system will continuously optimize the recommendation model, adjust the recommendation strategy, re-evaluate the priority of the target or improve the screening algorithm to ensure that user needs can be more accurately met in subsequent recommendations. This closed-loop feedback process enables the system to continuously learn and improve, ensuring that the recommended targets are always highly matched with user needs and can adapt to changes in user preferences and fluctuations in market conditions.

[0094] Establish a user interactive feedback mechanism to collect user satisfaction with recommendation results and actual investment behavior data. Based on user feedback information, automatically adjust and optimize the user investment preference model and recommendation model to further improve the system's recommendation accuracy and user satisfaction, and achieve system adaptive optimization;

[0095] Establish a user interactive feedback mechanism to collect user satisfaction with recommendation results and actual investment behavior data. Based on user feedback information, automatically adjust and optimize the user investment preference model and recommendation model to further improve the system's recommendation accuracy and user satisfaction, and achieve system adaptive optimization. The specific steps are as follows:

[0096] During the interaction between users and the recommendation system, continuously collect user feedback data on the recommendation results;

[0097] These feedback data include the click rate of users on the recommended targets, the returns after the investment decision, the user's satisfaction score on the recommended results, and the user's active feedback (such as evaluation, suggestions or complaints). The system will classify and process these feedback data into direct feedback (such as user ratings) and indirect feedback (such as actual investment behavior). The focus of this step is to ensure that the collected data is comprehensive and accurate, covering all aspects of the user's interaction with the system, and providing rich data support for subsequent analysis and optimization.

[0098] After collecting user feedback data, we will conduct a detailed analysis of the data to identify users’ preferences and dissatisfaction with the recommendation results, and compare the feedback data with the output of the recommendation model to evaluate the accuracy and effectiveness of the current model;

[0099] For example, by analyzing the user's investment success rate and satisfaction score, the system can determine whether the recommended target really meets the user's needs. For negative feedback, the system will conduct an in-depth analysis of the reasons, such as insufficient relevance of the recommended target, inaccurate risk assessment, etc. The core of this step is to find the weak links in the model through data analysis and determine the specific aspects that need to be adjusted.

[0100] Based on the analysis results of feedback data, automatically adjust and optimize the user investment preference model and recommendation model, use machine learning algorithms to retrain and adjust model parameters, and correct model deviations exposed by feedback data;

[0101] For example, if the user is dissatisfied with the recommendation result of a certain type of high-risk target, the system will reduce the weight of this type of target in the model or adjust the risk assessment strategy. The automatic adjustment process is based on the continuous input and learning of feedback data, so that the model can be gradually improved and its adaptability to user needs can be enhanced.

[0102] After the model is adjusted and optimized, new user feedback is continuously monitored to form an adaptive optimization loop. The system regularly or in real time evaluates the performance of the adjusted model, observes changes in user satisfaction and improvements in recommendation accuracy. If the user feedback becomes more positive after the model is optimized, the current adjustment strategy will be maintained; if the problem still exists, further iterative optimization will be performed until the ideal recommendation effect is achieved.

[0103] Through this adaptive feedback loop, the system can continuously evolve and always stay aligned with users’ investment preferences and market dynamics, achieving continuous improvement of the recommendation system and maximizing user satisfaction.

[0104] The present invention provides Figure 2The target intelligent push system based on big data analysis shown includes a user preference modeling module, a market dynamic monitoring module, a recommendation model optimization module, a personalized recommendation module, and a feedback and adaptive optimization module;

[0105] The user preference modeling module collects the user's investment behavior data, market dynamics data and user preference data in real time, combines historical data to dynamically analyze the user's investment preferences, and generates a real-time investment preference model for the user;

[0106] The market dynamics monitoring module monitors and processes market information from multiple data sources in real time in multiple dimensions, builds a market dynamics database, identifies market trends and hot spots through market data analysis, and compares this information with the user's real-time investment preference model to assess the impact of current market conditions on the user's potential investment targets;

[0107] The recommendation model optimization module uses machine learning algorithms to train and optimize the recommendation model online based on user and market data. It updates and adjusts the machine learning model in real time according to real-time user feedback and changes in market data to ensure that the recommendation algorithm is always consistent with the user's current investment preferences and market dynamics.

[0108] The personalized recommendation module generates a list of targets that meet the user's current investment preferences based on a real-time optimized recommendation model and sorts them by priority. During the recommendation process, the module conducts personalized screening and customized push of targets based on the user's risk tolerance and investment strategy to ensure that the recommended targets match the user's needs.

[0109] Feedback and adaptive optimization module: Establish a user interactive feedback mechanism, collect user satisfaction with recommendation results and actual investment behavior data, and automatically adjust and optimize the user investment preference model and recommendation model based on user feedback information, further improve the system's recommendation accuracy and user satisfaction, and achieve system adaptive optimization;

[0110] The target intelligent push method based on big data analysis provided in an embodiment of the present invention is realized by the above-mentioned target intelligent push system based on big data analysis. The specific methods and processes of the target intelligent push system based on big data analysis are detailed in the above-mentioned embodiment of the target intelligent push method based on big data analysis, which will not be repeated here.

[0111] The present invention generates a model that promptly reflects the user's latest investment preferences by real-time collection and dynamic analysis of user behavior and market data. At the same time, it monitors market information in real time from multiple dimensions and compares it with user preferences to ensure that the recommended targets are highly matched with current market trends. The recommendation model is trained and optimized online so that it can continuously adapt to market changes and user needs. In addition, personalized screening and customized push further improve the accuracy of recommendations and user satisfaction. Through the user interactive feedback mechanism, the system realizes adaptive optimization, continuously adjusts and improves the recommendation strategy, thereby significantly improving the accuracy of recommendation results and the user's investment returns, thereby enabling the recommendation system to dynamically respond to user needs and market changes, provide more accurate and efficient investment target recommendation services, and greatly improve the effectiveness of the system and user satisfaction.

[0112] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for intelligently pushing targets based on big data analysis, characterized in that: The following steps are involved: By collecting users' investment behavior data, market dynamics data and user preference data in real time, and combining historical data to dynamically analyze users' investment preferences, a real-time investment preference model for users is generated; Conduct multi-dimensional real-time monitoring and processing of market information from multiple data sources, build a market dynamics database, identify market trends and hot spots through market data analysis, and compare this information with the user's real-time investment preference model to evaluate the impact of current market conditions on the user's potential investment targets; Based on user and market data, we use machine learning algorithms to train and optimize the recommendation model online. According to real-time feedback from users and changes in market data, we update and adjust the machine learning model in real time to ensure that the recommendation algorithm is always consistent with users' current investment preferences and market dynamics. Based on the real-time optimized recommendation model, a list of targets that meet the user's current investment preferences is generated and sorted by priority. During the recommendation process, the targets are personalized and customized based on the user's risk tolerance and investment strategy to ensure that the recommended targets match the user's needs; Establish a user interactive feedback mechanism, collect user satisfaction with recommendation results and actual investment behavior data, and automatically adjust and optimize user investment preference models and recommendation models based on user feedback information to further improve the system's recommendation accuracy and user satisfaction and achieve adaptive optimization of the system.

2. The target intelligent push method according to claim 1, characterized in that: The specific steps for generating a user's real-time investment preference model are as follows: Collect users' investment behavior data, market dynamics data and user preference data in real time from multiple sources; After data collection is completed, historical data is used as a basis to generate a preliminary user preference model; Based on pre-built benchmark models, dynamic analysis of user behavior and market data collected in real time is performed; After the real-time analysis and model update are completed, the model will be further optimized through the user interaction feedback mechanism.

3. The target intelligent push method according to claim 1, characterized in that: Evaluate the impact of current market conditions on your potential investment targets. The specific steps are as follows: Collect market information from multiple data sources in real time; After data cleaning, the collected market data is analyzed and modeled in multiple dimensions. Multi-dimensional analysis includes slicing and filtering market data from multiple perspectives. Based on the results of multi-dimensional analysis, we use forecasting models and pattern recognition technology to identify market trends and hot spots in real time; After identifying market trends and hot spots, this information is compared with the user's real-time investment preference model. During the comparison process, it is analyzed how the current market conditions affect the user's potential investment targets.

4. The target intelligent push method according to claim 1, characterized in that: Based on user and market data, we use machine learning algorithms to train and optimize the recommendation model online. According to real-time feedback from users and changes in market data, we update and adjust the machine learning model in real time to ensure that the recommendation algorithm is always consistent with users' current investment preferences and market dynamics. The specific steps are as follows: Extract key features from users’ investment behavior data and market dynamics data; After feature extraction is completed, the recommendation model is trained online using the extracted feature data; A real-time feedback mechanism has been established to collect users’ actual behavior and feedback information immediately after they make investment decisions; Based on the online model training and feedback adjustment, model evaluation and adaptive optimization are performed regularly.

5. The target intelligent push method according to claim 1, characterized in that: Based on the real-time optimized recommendation model, a list of targets that meet the user's current investment preferences is generated and sorted by priority. During the recommendation process, the targets are personalized and customized based on the user's risk tolerance and investment strategy to ensure that the recommended targets match the user's needs. The specific steps are as follows: Filter out targets that match the user's current investment preferences and market trends from a massive library of investment targets; After generating a preliminary list of targets, the system prioritizes these targets based on the user's investment objectives, risk tolerance, and market conditions; After the priority sorting is completed, the target will be screened more deeply and personalized, and the screening process is based on the user's investment strategy and personalized needs; After the target is pushed, we continuously monitor the user's response and actual investment behavior, and collect relevant feedback data. By analyzing the feedback data, we evaluate the effectiveness of the recommendation. Based on the evaluation results, we continuously optimize the recommendation model, adjust the recommendation strategy, re-evaluate the target priority or improve the screening algorithm to ensure that user needs are accurately met in subsequent recommendations.

6. The target intelligent push method according to claim 1, characterized in that: Establish a user interactive feedback mechanism to collect user satisfaction with recommendation results and actual investment behavior data. Based on user feedback information, automatically adjust and optimize the user investment preference model and recommendation model to further improve the system's recommendation accuracy and user satisfaction, and achieve system adaptive optimization. The specific steps are as follows: During the interaction between users and the recommendation system, continuously collect user feedback data on the recommendation results; After collecting user feedback data, we will conduct a detailed analysis of the data to identify users’ preferences and dissatisfaction with the recommendation results, and compare the feedback data with the output of the recommendation model to evaluate the accuracy and effectiveness of the current model; Based on the analysis results of feedback data, automatically adjust and optimize the user investment preference model and recommendation model, use machine learning algorithms to retrain and adjust model parameters, and correct model deviations exposed by feedback data; After the model is adjusted and optimized, new user feedback is continuously monitored to form an adaptive optimization loop. The system regularly or in real time evaluates the performance of the adjusted model, observes changes in user satisfaction and improvements in recommendation accuracy. If the positivity of user feedback increases after model optimization, the current adjustment strategy will be maintained; if the problem still exists, further iterative optimization will be performed until the ideal recommendation effect is achieved.

7. A target intelligent push system based on big data analysis, used to implement a target intelligent push method based on big data analysis as described in any one of claims 1 to 6, characterized in that: It includes user preference modeling module, market dynamics monitoring module, recommendation model optimization module, personalized recommendation module, and feedback and adaptive optimization module; The user preference modeling module collects the user's investment behavior data, market dynamics data and user preference data in real time, combines historical data to dynamically analyze the user's investment preferences, and generates a real-time investment preference model for the user; The market dynamics monitoring module monitors and processes market information from multiple data sources in real time in multiple dimensions, builds a market dynamics database, identifies market trends and hot spots through market data analysis, and compares this information with the user's real-time investment preference model to assess the impact of current market conditions on the user's potential investment targets; The recommendation model optimization module uses machine learning algorithms to train and optimize the recommendation model online based on user and market data. It updates and adjusts the machine learning model in real time according to real-time user feedback and changes in market data to ensure that the recommendation algorithm is always consistent with the user's current investment preferences and market dynamics. The personalized recommendation module generates a list of targets that meet the user's current investment preferences based on a real-time optimized recommendation model and sorts them by priority. During the recommendation process, the module conducts personalized screening and customized push of targets based on the user's risk tolerance and investment strategy to ensure that the recommended targets match the user's needs. The feedback and adaptive optimization module establishes a user interactive feedback mechanism, collects user satisfaction with recommendation results and actual investment behavior data, and automatically adjusts and optimizes user investment preference models and recommendation models based on user feedback information, further improving the system's recommendation accuracy and user satisfaction, and achieving system adaptive optimization.

8. The target intelligent push system according to claim 7, characterized in that: The user preference modeling module is specifically used to: collect user investment behavior data, market dynamics data and user preference data in real time from multiple sources; after data collection is completed, use historical data as the basis to generate a preliminary user preference model; based on the pre-built benchmark model, dynamically analyze the user behavior and market data collected in real time; after the real-time analysis and model update are completed, the model will be further optimized through the user interaction feedback mechanism.

9. The target intelligent push system according to claim 7, characterized in that: The market dynamic monitoring module is specifically used to: collect market information from multiple data sources in real time; after completing data cleaning, conduct multi-dimensional analysis and modeling on the collected market data, and the multi-dimensional analysis includes slicing and screening market data from multiple perspectives; based on the results of multi-dimensional analysis, use prediction models and pattern recognition technology to identify trends and hot spots in the market in real time; After identifying market trends and hot spots, this information is compared with the user's real-time investment preference model. During the comparison process, it is analyzed how the current market conditions affect the user's potential investment targets.

10. The target intelligent push system according to claim 7, characterized in that: The recommendation model optimization module is specifically used to: extract key features from users' investment behavior data and market dynamics data; after completing feature extraction, use the extracted feature data to conduct online training of the recommendation model; establish a real-time feedback mechanism to immediately collect users' actual behavior and feedback information after they make investment decisions; and regularly conduct model evaluation and adaptive optimization based on online model training and feedback adjustment.

Images