Information processing method and device for content filtering and information classification, equipment and medium

Abstract

The invention relates to an information processing method and device for content filtering and information classification, equipment and a medium, and the method comprises the following steps: calling a preset multi-dimensional classification model which automatically adjusts the classification weight according to the features of input data, and carrying out the multi-level feature extraction based on a GNN (Graph Neural Network) and a CNN (Convolutional Neural Network); and mapping the classification path to a processing tool Apache Kafka by using an automatic mapping mechanism, calling a signal processing module, a cross-domain ontology mapping module and a data flow analysis module, and deeply processing information. Through an automatic mapping mechanism, the system can dynamically map filtered information classification paths to a proper processing tool set, signal processing, cross-domain ontology mapping and data flow analysis modules are called, and the multi-dimensional depth information analysis capability ensures that the system not only can filter information, but also can analyze data flow. In addition, hidden association and deep semantic relationships among the information can be mined, and deeper insight is given to the information processing process.

Description

Technical Field

[0001] The present invention relates to the technical field of information processing, and in particular to an information processing method, device, equipment and medium for content filtering and information classification. Background Art

[0002] In the modern information society, with the rapid development of the Internet and communication technology, content filtering and information classification technology are particularly important.

[0003] Traditional information processing methods, due to their reliance on static rules and limited computing power, often have greatly reduced efficiency and accuracy when faced with diverse and changing information types. For example, early filtering technologies were mostly based on keyword matching or simple Boolean rules, but these methods have difficulty processing information with complex semantics or diverse structures. For example, on social media, user-generated content comes in many forms, involving not only text but also pictures, videos, and audio content. The semantic relationships between these contents are complex, and even the same keyword may require different processing strategies in different contexts. In addition, the timeliness of information also places higher demands on processing efficiency, because timely content filtering and classification are crucial to the effectiveness of decision-making.

[0004] In this context, the development of machine learning and natural language processing technologies has brought new opportunities to solve these problems. Machine learning algorithms, especially deep learning, can understand the deep structure and potential patterns of data by training on large data sets. Among them, convolutional neural networks (CNN) have achieved remarkable success in image classification and feature extraction, while recurrent neural networks (RNN) and its variants such as long short-term memory networks (LSTM) have excelled in processing time series data. In addition, the graph neural network (GNN) that has emerged in recent years has shown powerful capabilities in modeling graph structure associations between data. These technologies provide technical support for more intelligent and automated information filtering and classification.

[0005] The advancement of data stream processing technology is also a major driving force. With the popularity of high-throughput, low-latency stream processing platforms such as Apache Kafka, real-time data processing has become possible. These platforms support distributed computing and can collect, analyze and process information streams in real time, making the system not only more responsive, but also able to perform complex real-time analysis of flowing data. This feature is especially important in application scenarios that require fast response (such as financial transactions, online content recommendations, etc.).

[0006] However, advances in computing technology bring not only opportunities but also challenges. For example, how to combine efficient machine learning models with flexible data flow technology, and how to achieve efficient processing while ensuring data privacy and security are issues that need to be addressed. In addition, user preferences and behavior patterns are constantly changing. How to dynamically adapt and optimize strategies during data processing to ensure long-term, efficient and stable operation of the system is still an open research topic.

[0007] In summary, the field of modern information processing is undergoing a transformation. Faced with the diversity of information types, the growing size of data, and the need to improve system responsiveness, content filtering and information classification technologies are moving from traditional static rules to intelligent, automated, and real-time development. This requires not only improvements to existing technologies, but also cross-domain technology integration and the ability to adapt to the ever-changing information environment. Summary of the invention

[0008] The main purpose of the present invention is to provide an information processing method, device, equipment and medium for content filtering and information classification.

[0009] To achieve the above object, the present invention provides an information processing method for content filtering and information classification, comprising the following steps: Calling a preset multidimensional classification model, which automatically adjusts the classification weight according to the characteristics of the input data, performs multi-level feature extraction based on the graph neural network GNN and the convolutional neural network CNN, and converts the multimodal information into a feature vector that can be used for classification; Based on the feature vector, create context-aware filtering rules, combine historical data to learn user preferences, generate a personalized filtering rule library, and based on the filtering rules, use association filtering and hierarchical content filtering algorithms to analyze hidden associations and semantic structures of information; Use the data stream processing tool Apache Kafka to perform real-time information filtering after analysis. Based on the real-time information filtering results, use the automated mapping mechanism to map the classification path to the processing tool Apache Kafka, call the signal processing, cross-domain ontology mapping and data stream analysis modules, and deeply process the information; Analyze the information processing results, adjust the filtering parameters through the embedded algorithm, and pass the adjustment results back to step 1 and step 2 to update the multidimensional classification model and filtering rule base; Generate visual reports of the processing process in real time, and use the report results to identify processing tools that need to be updated and expanded based on the microservice architecture.

[0010] Furthermore, the calling of a preset multidimensional classification model, the multidimensional classification model automatically adjusts the classification weight according to the characteristics of the input data, performs multi-level feature extraction based on the graph neural network GNN and the convolutional neural network CNN, and converts the multimodal information into a feature vector that can be used for classification; including: Calling a preset multidimensional classification model, the multidimensional classification model receives and analyzes the characteristics of the input data, and automatically optimizes the classification weight through an adaptive dynamic weight adjustment mechanism; Use graph neural network GNN to model complex associations in data; Use convolutional neural network (CNN) to extract spatial features and local patterns of data, and extract key patterns and features in the data through multi-layer convolution operations; After being processed by GNN and CNN, the extracted deep features are integrated to generate a comprehensive feature vector.

[0011] Furthermore, the process of creating context-aware filtering rules based on the feature vectors, learning user preferences in combination with historical data, and generating a personalized filtering rule library includes: Analyzing the feature vector to identify key attributes related to the user context, including but not limited to time, location, and behavior pattern; Combined with historical data, machine learning algorithms are applied to capture and learn user preferences and behavior patterns, and identify users' specific information filtering needs in different situations; The information is converted into personalized filtering rules and compiled into a rule base.

[0012] Furthermore, based on the filtering rules, the hidden associations and semantic structures of the information are analyzed by using association filtering and hierarchical content filtering algorithms; including: Standardize the input information, including format adjustment and basic cleaning; Apply association filtering algorithms to analyze potential connections between information elements and construct relationship graphs to identify relevant data; Use hierarchical content filtering algorithms to extract key features of information according to semantic depth, and analyze it step by step from the surface to the deep layer; In combination with the multi-stage filtering process, non-priority information is eliminated under the guidance of the filtering rule base, and the rule base is optimized through a feedback mechanism.

[0013] Furthermore, the data stream processing tool Apache Kafka is used to perform post-analysis real-time information filtering, and according to the real-time information filtering results, the classification path is mapped to the processing tool Apache Kafka using an automated mapping mechanism, and the signal processing, cross-domain ontology mapping and data stream analysis modules are called to deeply process the information; including: Start a real-time data pipeline to receive and analyze the continuously flowing streaming information, and quickly perform preliminary filtering through the stream processing capabilities of the processing tool Apache Kafka; Based on the results of real-time information filtering, an automated mapping mechanism is used to associate the identified classification paths with a preconfigured set of processing tools; Map the signal processing module, the cross-domain ontology mapping module and the further data flow analysis module to the current data flow task.

[0014] Furthermore, the analysis information processing results, adjusting the filtering parameters through the embedded algorithm, and transmitting the adjustment results back to step 1 and step 2 to update the multidimensional classification model and the filtering rule base; including: Call embedded analysis algorithms to evaluate the output results of signal processing, cross-domain ontology mapping, and data flow analysis modules; Based on the evaluation results, automatically identify and calculate the optimization space of each filtering parameter and classification weight; Dynamically adjust the filtering parameters to ensure that the filtering strategy is highly matched with the real-time data characteristics; It is transmitted back to the front multidimensional classification model and filtering rule base through the feedback mechanism.

[0015] Furthermore, the real-time generation of a visual report of the processing process, and identifying the processing tools that need to be updated and expanded based on the microservice architecture through the report results, include: collecting key metadata and performance indicators from the signal processing module, the cross-domain ontology mapping module, and the data flow analysis module using an embedded visualization component; using the metadata and performance indicators to present the status and results of information processing in the form of charts and graphs in a real-time dashboard; Based on the microservice architecture, interpret the conclusions of the visualization report and identify the processing tools and modules that need to be updated or expanded through the API interface; Automatically schedule update tasks based on identified needs, perform modular upgrades or dynamically expand services.

[0016] The present invention also provides an information processing device for content filtering and information classification, comprising: Extraction unit; used to call a preset multidimensional classification model, the multidimensional classification model automatically adjusts the classification weight according to the characteristics of the input data, performs multi-level feature extraction based on the graph neural network GNN and the convolutional neural network CNN, and converts the multimodal information into a feature vector that can be used for classification; An analysis unit; used to create context-aware filtering rules based on the feature vectors, learn user preferences in combination with historical data, generate a personalized filtering rule library, and analyze the hidden associations and semantic structures of information based on the filtering rules using association filtering and hierarchical content filtering algorithms; Processing unit: used to perform real-time information filtering after analysis using the data stream processing tool Apache Kafka. According to the real-time information filtering results, the classification path is mapped to the processing tool Apache Kafka using the automatic mapping mechanism, and the signal processing, cross-domain ontology mapping and data stream analysis modules are called to deeply process the information. Feedback unit: used to analyze the information processing results, adjust the filtering parameters through the embedded algorithm, and transmit the adjustment results back to the front-end multi-dimensional classification model and filtering rule library; Identification unit: used to generate visual reports of the processing process in real time, and identify the processing tools that need to be updated and expanded based on the microservice architecture through the report results.

[0017] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and wherein the processor implements the steps of the information processing method for content filtering and information classification as described in any one of claims 1 to 7 when executing the computer program.

[0018] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the information processing method for content filtering and information classification described in any one of claims 1 to 7 are implemented.

[0019] The information processing method, device, equipment and medium for content filtering and information classification provided by the present invention have the following beneficial effects: (1) The present invention calls a preset multi-dimensional classification model. The method can automatically adjust the classification weight according to the characteristics of the input data and use graph neural network (GNN) and convolutional neural network (CNN) to perform multi-level feature extraction. The application of this deep learning technology enables the system to extract more representative feature vectors from complex and multimodal information, thereby greatly improving the accuracy and efficiency of classification.

[0020] (2) The context-aware filtering rule library created in the present invention learns user preferences by combining historical data. It uses machine learning algorithms to capture users' specific information filtering needs in different situations and generate a personalized filtering rule library. The system can dynamically adapt to users' changing preferences and more accurately filter out information content that they are concerned about from massive amounts of information, providing a customized user experience.

[0021] (3) The present invention adopts data stream processing tools such as Apache Kafka, and the system can realize real-time information filtering and analysis. The instant processing capability of data stream enables the system to respond quickly to changes in input information. It is particularly suitable for coping with rapidly changing data environments, such as financial market analysis, real-time news filtering and other scenarios, ensuring the high timeliness of information.

[0022] (4) Through the automated mapping mechanism, the system can dynamically map the filtered information classification path to an appropriate processing tool set, calling the signal processing, cross-domain ontology mapping and data flow analysis modules. This multi-dimensional deep information analysis capability ensures that the system can not only filter information, but also discover the hidden associations and deep semantic relationships between information, giving deeper insights into the information processing process.

[0023] (5) The present invention analyzes the information processing results through an embedded algorithm, automatically adjusts the filtering parameters, and feeds back the adjustment results to update the multidimensional classification model and filtering rule library. The system has a built-in closed loop of continuous learning and self-optimization. This capability ensures that the system can gradually improve its filtering and classification accuracy over time, thereby adapting to the ever-changing environment and user needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a flow chart of an information processing method for content filtering and information classification in one embodiment of the present invention; Figure 2 It is a structural block diagram of an information processing device for content filtering and information classification in one embodiment of the present invention; Figure 3 It is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0025] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0027] The information processing method for content filtering and information classification proposed by the present invention is executed by a computer device as an execution subject.

[0028] Reference Figure 1 A flowchart of an information processing method for content filtering and information classification proposed by the present invention includes the following steps: S1; calling a preset multidimensional classification model, which automatically adjusts the classification weight according to the characteristics of the input data, performs multi-level feature extraction based on the graph neural network GNN and the convolutional neural network CNN, and converts the multimodal information into a feature vector that can be used for classification; In one embodiment, a preset multidimensional classification model is called, the multidimensional classification model automatically adjusts the classification weight according to the characteristics of the input data, performs multi-level feature extraction based on the graph neural network GNN and the convolutional neural network CNN, and converts the multimodal information into a feature vector that can be used for classification; including: S1.1; calling a preset multidimensional classification model, wherein the multidimensional classification model receives and analyzes the characteristics of the input data, and automatically optimizes the classification weight through an adaptive dynamic weight adjustment mechanism; S1.2; Use graph neural network GNN to model complex associations in data; S1.3; Use convolutional neural network (CNN) to extract spatial features and local patterns of data, and extract key patterns and features in the data through multi-layer convolution operations; S1.4; After being processed by GNN and CNN, the extracted deep features are integrated to generate a comprehensive feature vector.

[0029] In the specific implementation process, the system first calls the preset multi-dimensional classification model, which receives and analyzes the characteristics of the input data and uses an adaptive dynamic weight adjustment mechanism to optimize the classification weight. In this process, the graph neural network (GNN) is first used to model the complex correlation relationships in the data, and the network structure information is extracted by considering the connectivity between nodes, emphasizing the contextual relationship and interaction pattern between data. Subsequently, the system uses convolutional neural network (CNN) to analyze the spatial features and local patterns of the data, and extracts the key patterns and feature details in the data through multi-layer convolution operations. Combining the results of GNN and CNN, the system integrates and fuses these deep-level features to generate a highly representative comprehensive feature vector. This feature vector fully reveals the characteristics and associations of the input data in multiple dimensions, and provides a solid foundation for subsequent classification tasks. Such a set of processes ensures that the system can not only accurately identify the static attributes of the data, but also have insights into the dynamic relationships between the data, thereby supporting information filtering and classification in complex environments.

[0030] The simplified formula for graph neural network (GNN) feature extraction is as follows: Node feature updates: H = f(W * (S / N)) in: H: Updated node features.

[0031] W: weight.

[0032] S: The sum of neighbor node features.

[0033] N: the number of neighbor nodes.

[0034] f: activation function, such as ReLU.

[0035] The simplified formula for convolutional neural network (CNN) feature extraction is as follows: F = f(P + b) in: F: Output features.

[0036] P: The result of convolution calculation.

[0037] b: Bias.

[0038] f: activation function, such as ReLU.

[0039] GNN updates features by aggregating neighbor node information, while CNN obtains new feature representations through local convolution calculations.

[0040] S2: Based on the feature vector, create context-aware filtering rules, combine historical data to learn user preferences, generate a personalized filtering rule library, and based on the filtering rules, use association filtering and hierarchical content filtering algorithms to analyze hidden associations and semantic structures of information; In one embodiment, based on the feature vector, context-aware filtering rules are created, user preferences are learned in combination with historical data, and a personalized filtering rule library is generated. Based on the filtering rules, association filtering and hierarchical content filtering algorithms are used to analyze the hidden associations and semantic structures of information; including: S2.1; analyzing the feature vector to identify key attributes related to the user context, including but not limited to time, location, and behavior pattern; S2.2; Combine historical data and apply machine learning algorithms to capture and learn user preferences and behavior patterns, and identify users' specific information filtering needs in different situations; S2.3; converting the information into personalized filtering rules and compiling them into a rule base; S2.4; Standardize the input information, including format adjustment and basic cleaning; S2.5; Apply association filtering algorithms to analyze potential connections between information elements and construct relationship graphs to identify relevant data; S2.6; Use hierarchical content filtering algorithms to extract key features of information according to semantic depth, and analyze it step by step from the surface to the deep layer; S2.7; In combination with the multi-stage filtering process, non-priority information is eliminated under the guidance of the filtering rule base, and the rule base is optimized through a feedback mechanism.

[0041] In the specific implementation process, the system first conducts a detailed analysis of the generated feature vectors, identifies and extracts key attributes related to the user context, including but not limited to time, location, and behavior patterns. Then, combined with a large amount of historical data, the system applies advanced machine learning algorithms to capture and learn user preferences and behavior patterns, identify users' specific information filtering needs in different situations, and create personalized filtering rules. These rules are systematically compiled into the rule base. Subsequently, the input information initially received is standardized to ensure consistent format and remove noise. Next, through the association filtering algorithm, the system analyzes the potential connectivity between information elements and constructs a relationship graph to identify the correlation between data. After that, the system uses a hierarchical content filtering algorithm to analyze information layer by layer according to semantic depth and extract key features from the surface to the deep layer. This process combines multi-level filtering steps, based on the guidance of the filtering rule base, eliminates non-priority information, and continuously optimizes and updates the rule base through a feedback mechanism, further enhancing the system's ability to analyze hidden associations and semantic structures of information. Through this series of operations, the system can dynamically adapt to user preferences and real-time information environment to achieve more accurate and personalized information filtering and classification.

[0042] S3: Use the data stream processing tool Apache Kafka to perform real-time information filtering after analysis. Based on the real-time information filtering results, use the automatic mapping mechanism to map the classification path to the processing tool Apache Kafka, call the signal processing, cross-domain ontology mapping and data stream analysis modules, and deeply process the information; In one embodiment, the data stream processing tool Apache Kafka is used to perform post-analysis real-time information filtering. According to the real-time information filtering results, the classification path is mapped to the processing tool Apache Kafka using an automated mapping mechanism, and the signal processing, cross-domain ontology mapping and data stream analysis modules are called to deeply process the information; including: S3.1; Start the real-time data pipeline, receive and analyze the continuously influx of streaming information, and quickly perform preliminary filtering through the stream processing capabilities of the processing tool Apache Kafka; S3.2; Based on the results of real-time information filtering, use an automated mapping mechanism to associate the identified classification paths with a preconfigured set of processing tools; S3.3; Map the signal processing module, the cross-domain ontology mapping module and the further data flow analysis module to the current data flow task.

[0043] In the specific implementation process, the system first starts the real-time data pipeline of Apache Kafka to receive a large amount of streaming information, and uses Kafka's powerful stream processing capabilities to quickly and efficiently filter this information. The main goal of this preliminary filtering stage is to quickly eliminate obviously irrelevant data to ensure the efficiency of the subsequent processing stage. Then, based on the real-time filtering results, the system uses an automated mapping mechanism to accurately locate and identify the classification path of the data. This result is dynamically associated with the pre-configured processing tool set in the system. Subsequently, the system calls a dedicated signal processing module, a cross-domain ontology mapping module, and an in-depth data flow analysis module, maps them to the current data flow task, and performs deep information processing. These modules work together to not only parse the surface features of the data, but also deeply explore the potential signal features and semantic structures, so that the system can perform a more comprehensive and accurate analysis of the information and provide a reliable data foundation for decision support. Through this entire set of processes, the system ensures real-time and efficient information filtering and deep data processing in a highly dynamic environment.

[0044] The formula for real-time information filtering is as follows: I filtered ={I t ∈I|Rule(I t )=true} in: I t is the information unit at time t.

[0045] I filtered is the result set that passes the filter.

[0046] Rule(I t ) is the defined filter rule function.

[0047] The path mapping and processing call formula is as follows: Based on the filtering results, determine the classification path and map it to the processing tool: mapped =Map(I filtered ) Call signal processing, cross-domain ontology mapping and data flow analysis module: R=Process(P mapped ) in: P mapped Is the processing path of the mapping.

[0048] Map is an automatic mapping mechanism.

[0049] Process is the calling process of modules such as signal processing.

[0050] R is the final processing result.

[0051] S4; analyzing the information processing results, adjusting the filtering parameters through the embedded algorithm, and transmitting the adjustment results back to step 1 and step 2 to update the multidimensional classification model and filtering rule base; In one embodiment, analyzing the information processing results, adjusting the filtering parameters through the embedded algorithm, and transmitting the adjustment results back to step 1 and step 2 to update the multi-dimensional classification model and the filtering rule base; including: S4.1; Call the embedded analysis algorithm to evaluate the output results of the signal processing, cross-domain ontology mapping and data flow analysis modules; S4.2; Based on the evaluation results, automatically identify and calculate the optimization space of each filtering parameter and classification weight; S4.3; dynamically adjust the filtering parameters to ensure that the filtering strategy is highly matched with the real-time data characteristics; S4.4; transmitted back to the front multidimensional classification model and filtering rule library through the feedback mechanism.

[0052] During the specific implementation process, the system first calls the embedded analysis algorithm to comprehensively evaluate the output results of the signal processing, cross-domain ontology mapping, and data flow analysis modules. By deeply analyzing these results, the system automatically identifies the optimization space of the current filtering parameters and classification weights. Based on this evaluation, the system makes dynamic adjustments and accurately optimizes these filtering parameters so that the filtering strategy is highly consistent with the real-time data characteristics to adapt to the changing data information environment. After the adjustment is completed, these optimized parameters are transmitted back to the front-end multidimensional classification model and filtering rule library through the feedback mechanism constructed by the system. This feedback link not only ensures the continuous optimization and efficient operation of the model, but also ensures that the filtering rule library can reflect the latest analysis strategies and user needs in real time, so that the entire information processing process forms a closed-loop adaptive system, and continuously improves the classification and filtering efficiency of information.

[0053] S5; Generate a visual report of the processing process in real time, and use the report results to identify the processing tools that need to be updated and expanded based on the microservice architecture; S5.1; In one embodiment, a visual report of the processing process is generated in real time, and the processing tools that need to be updated and expanded are identified based on the microservice architecture through the report results; including: S5.2; using the embedded visualization component to collect key metadata and performance indicators from the signal processing module, the cross-domain ontology mapping module and the data flow analysis module; S5.3; displaying the status and results of information processing in the form of charts and graphs in a real-time dashboard using the metadata and performance indicators; S5.4; Based on the microservice architecture, interpret the conclusions of the visualization report and identify the processing tools and modules that need to be updated or expanded through the API interface; S5.5; Automatically schedule update tasks based on identified needs, perform modular upgrades or dynamically expand services.

[0054] During the specific implementation process, the system extracts key metadata and performance indicators from the signal processing module, cross-domain ontology mapping module and data flow analysis module through embedded visualization components. These data include core indicators such as processing speed, accuracy, and latency. These collected metadata and performance indicators are presented in real time in the integrated dashboard, which intuitively displays the operating status and effect analysis of the information processing system in the form of charts and graphs. Utilizing the flexibility of the microservice architecture, the system interprets the conclusions presented in the visualization report and accurately identifies the processing tools and modules that need to be updated or expanded through the API interface. According to these identified needs, the system automatically schedules and executes the corresponding update tasks, and performs modular upgrades or dynamic service expansions. This feedback and self-adjustment mechanism not only enables the system to adapt to changes in information processing needs in real time, but also ensures the efficiency and scalability of the processing tools, and continuously optimizes the overall system performance.

[0055] Reference Figure 2 The structure block diagram of an information processing device for content filtering and information classification in one embodiment of the present invention specifically discloses an information processing device for content filtering and information classification, including: Extraction unit; used to call a preset multidimensional classification model, the multidimensional classification model automatically adjusts the classification weight according to the characteristics of the input data, performs multi-level feature extraction based on the graph neural network GNN and the convolutional neural network CNN, and converts the multimodal information into a feature vector that can be used for classification; An analysis unit; used to create context-aware filtering rules based on the feature vectors, learn user preferences in combination with historical data, generate a personalized filtering rule library, and analyze the hidden associations and semantic structures of information based on the filtering rules using association filtering and hierarchical content filtering algorithms; Processing unit: used to perform real-time information filtering after analysis using the data stream processing tool Apache Kafka. According to the real-time information filtering results, the classification path is mapped to the processing tool Apache Kafka using the automatic mapping mechanism, and the signal processing, cross-domain ontology mapping and data stream analysis modules are called to deeply process the information. Feedback unit: used to analyze the information processing results, adjust the filtering parameters through the embedded algorithm, and transmit the adjustment results back to the front-end multi-dimensional classification model and filtering rule library; Identification unit: used to generate visual reports of the processing process in real time, and identify the processing tools that need to be updated and expanded based on the microservice architecture through the report results.

[0056] In this embodiment, for the specific implementation of each unit in the above device embodiment, please refer to the above method embodiment, which will not be repeated here.

[0057] Reference Figure 3 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0058] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0059] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0060] In summary, the present invention calls a preset multidimensional classification model. The method can automatically adjust the classification weight according to the characteristics of the input data, and uses graph neural network (GNN) and convolutional neural network (CNN) to perform multi-level feature extraction. The application of this deep learning technology enables the system to extract more representative feature vectors from complex and multimodal information, thereby greatly improving the accuracy and efficiency of classification; the context-aware filtering rule library created in the present invention learns user preferences by combining historical data, and uses machine learning algorithms to capture users' specific information filtering needs in different situations, and generates a personalized filtering rule library. The system can dynamically adapt to users' changing preferences, more accurately filter out the information content they care about from massive information, and provide a customized user experience; the present invention adopts Apache Kafka and other data stream processing tools, the system can realize real-time information filtering and analysis. The instant processing capability of data stream enables the system to respond quickly to changes in input information, which is particularly suitable for dealing with rapidly changing data environments, such as financial market analysis, real-time news filtering and other scenarios, to ensure the high timeliness of information; the present invention uses an automated mapping mechanism, and the system can dynamically map the filtered information classification path to a suitable processing tool set, calling signal processing, cross-domain ontology mapping and data stream analysis modules. This multi-dimensional deep information analysis capability ensures that the system can not only filter information, but also dig out hidden associations and deep semantic relationships between information, giving the information processing process a deeper insight; the present invention analyzes the information processing results through an embedded algorithm, automatically adjusts the filtering parameters, and feeds back the adjustment results to update the multi-dimensional classification model and filtering rule library. The system has a built-in closed loop of continuous learning and self-optimization. This capability ensures that the system can gradually improve its filtering and classification accuracy over time, thereby adapting to the ever-changing environment and user needs.

[0061] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0062] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0063] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. An information processing method for content filtering and information classification, characterized in that: The following steps are involved: Calling a preset multidimensional classification model, which automatically adjusts the classification weight according to the characteristics of the input data, performs multi-level feature extraction based on the graph neural network GNN and the convolutional neural network CNN, and converts the multimodal information into a feature vector that can be used for classification; Based on the feature vector, create context-aware filtering rules, combine historical data to learn user preferences, generate a personalized filtering rule library, and based on the filtering rules, use association filtering and hierarchical content filtering algorithms to analyze hidden associations and semantic structures of information; Use the data stream processing tool Apache Kafka to perform real-time information filtering after analysis. Based on the real-time information filtering results, use the automated mapping mechanism to map the classification path to the processing tool Apache Kafka, call the signal processing, cross-domain ontology mapping and data stream analysis modules, and deeply process the information; Analyze the information processing results, adjust the filtering parameters through the embedded algorithm, and pass the adjustment results back to step 1 and step 2 to update the multidimensional classification model and filtering rule base; Generate visual reports of the processing process in real time, and use the report results to identify processing tools that need to be updated and expanded based on the microservice architecture.

2. The information processing method for content filtering and information classification according to claim 1, characterized in that: The calling of a preset multidimensional classification model, wherein the multidimensional classification model automatically adjusts the classification weight according to the characteristics of the input data, performs multi-level feature extraction based on the graph neural network GNN and the convolutional neural network CNN, and converts the multimodal information into a feature vector that can be used for classification; including: Calling a preset multidimensional classification model, the multidimensional classification model receives and analyzes the characteristics of the input data, and automatically optimizes the classification weight through an adaptive dynamic weight adjustment mechanism; Use graph neural network GNN to model complex associations in data; Use convolutional neural network (CNN) to extract spatial features and local patterns of data, and extract key patterns and features from data through multi-layer convolution operations; After being processed by GNN and CNN, the extracted deep features are integrated to generate a comprehensive feature vector.

3. The information processing method for content filtering and information classification according to claim 1, characterized in that: The method of creating context-aware filtering rules based on the feature vector, learning user preferences in combination with historical data, and generating a personalized filtering rule library includes: Analyzing the feature vector to identify key attributes related to the user context, including but not limited to time, location, and behavior pattern; Combined with historical data, machine learning algorithms are applied to capture and learn user preferences and behavior patterns, and identify users' specific information filtering needs in different situations; The information is converted into personalized filtering rules and compiled into a rule base.

4. The information processing method for content filtering and information classification according to claim 1, characterized in that: Based on the filtering rules, the hidden associations and semantic structures of the information are analyzed by using association filtering and hierarchical content filtering algorithms; including: Standardize the input information, including format adjustment and basic cleaning; Apply association filtering algorithms to analyze potential connections between information elements and construct relationship graphs to identify relevant data; Use hierarchical content filtering algorithms to extract key features of information according to semantic depth, and analyze it step by step from the surface to the deep layer; In combination with the multi-stage filtering process, non-priority information is eliminated under the guidance of the filtering rule base, and the rule base is optimized through a feedback mechanism.

5. The information processing method for content filtering and information classification according to claim 1, characterized in that: The method uses the data stream processing tool Apache Kafka to perform real-time information filtering after analysis, and according to the real-time information filtering results, uses the automatic mapping mechanism to map the classification path to the processing tool Apache Kafka, calls the signal processing, cross-domain ontology mapping and data stream analysis modules, and deeply processes the information; including: Start a real-time data pipeline to receive and analyze the continuous influx of streaming information, and quickly perform preliminary filtering through the stream processing capabilities of the processing tool Apache Kafka; Based on the results of real-time information filtering, an automated mapping mechanism is used to associate the identified classification paths with a preconfigured set of processing tools; Map the signal processing module, the cross-domain ontology mapping module and the further data flow analysis module to the current data flow task.

6. The information processing method for content filtering and information classification according to claim 1, characterized in that: The analysis information processing results, adjusting the filtering parameters through the embedded algorithm, and transmitting the adjustment results back to step 1 and step 2 to update the multidimensional classification model and the filtering rule base; including: Call embedded analysis algorithms to evaluate the output results of signal processing, cross-domain ontology mapping, and data flow analysis modules; Based on the evaluation results, automatically identify and calculate the optimization space of each filtering parameter and classification weight; Dynamically adjust the filtering parameters to ensure that the filtering strategy is highly matched with the real-time data characteristics; It is transmitted back to the front multidimensional classification model and filtering rule base through the feedback mechanism.

7. The information processing method for content filtering and information classification according to claim 1, characterized in that: The real-time generation of a visual report of the processing process, through the report results, identifies the processing tools that need to be updated and expanded based on the microservice architecture; including: collecting key metadata and performance indicators from the signal processing module, the cross-domain ontology mapping module, and the data flow analysis module using an embedded visualization component; using the metadata and performance indicators to present the status and results of information processing in the form of charts and graphs in a real-time dashboard; Based on the microservice architecture, interpret the conclusions of the visualization report and identify the processing tools and modules that need to be updated or expanded through the API interface; Automatically schedule update tasks based on identified needs, perform modular upgrades or dynamically expand services.

8. An information processing device for content filtering and information classification, characterized in that: include: Extraction unit; Used to call a preset multidimensional classification model, which automatically adjusts the classification weight according to the characteristics of the input data, performs multi-level feature extraction based on the graph neural network GNN and the convolutional neural network CNN, and converts multimodal information into feature vectors that can be used for classification; Unit of analysis; Used to create context-aware filtering rules based on the feature vectors, learn user preferences in combination with historical data, generate a personalized filtering rule library, and analyze hidden associations and semantic structures of information using association filtering and hierarchical content filtering algorithms based on the filtering rules; Processing unit; It is used to perform real-time information filtering after analysis using the data stream processing tool Apache Kafka. According to the real-time information filtering results, the classification path is mapped to the processing tool Apache Kafka using the automatic mapping mechanism, and the signal processing, cross-domain ontology mapping and data stream analysis modules are called to deeply process the information; Return unit; Used to analyze the information processing results, adjust the filtering parameters through the embedded algorithm, and transmit the adjustment results back to the front-end multi-dimensional classification model and filtering rule library; Identification unit; Used to generate visual reports of the processing process in real time, and through the report results, identify the processing tools that need to be updated and expanded based on the microservice architecture.

9. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the information processing method for content filtering and information classification described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the information processing method for content filtering and information classification according to any one of claims 1 to 7 are implemented.

Images