328 results about "Content security" patented technology

Automatically investigating security incidents and generating security incident reports using a Large Language Model (LLM). A computerized system receives an incoming Security Alert Message pertaining to a possible security-related incident. The system automatically feeds into the LLM at least: the content of the Security Alert Message; the metadata of the Security Alert Message; context information describing a security domain; and organization context information pertaining to users and machines of that organization. The system automatically prompts the LLM to automatically investigate the Security Alert Message and to automatically generate a detailed Incident Report pertaining to the Security Alert Message.

The invention discloses a video stream dynamic fragmentation encryption and block chain evidence storage method, and relates to the technical field of video content security, and the method comprises the steps: calculating a color histogram difference value and an optical flow vector change rate between adjacent frames of an input video, marking the difference value as a scene switching point when the difference value exceeds a preset threshold value, and storing the scene switching point; the method comprises the following steps: preliminarily dividing a video into a plurality of scene segments according to scene switching points, performing content complexity evaluation on the scene segments, calculating gray level co-occurrence matrix characteristics of each frame of image through texture density analysis, calculating edge complexity to extract the number and distribution of Canny edges, and performing motion vector statistics to analyze the size and direction of inter-frame object displacement. The change rate between adjacent pixels in the color space is measured according to the color change gradient; the video stream dynamic fragment encryption and block chain evidence storage method is suitable for video contents of different types and complexities, and has relatively high detection accuracy and robustness.

According to the technical scheme, the method comprises the steps that input risk control detection is conducted on input questions of a user, if risks exist, a default risk control answer template is inquired, and all characters of the risk control answer template are sequentially returned in a streaming mode; if the risk does not exist, the problem is transmitted to a large model to generate streaming output; generating a to-be-detected text based on streaming output, performing content security detection on the to-be-detected text, determining a content security detection result, and if the detection result is no risk, continuing to generate and output a large model; and if the detection result is that the risk exists, stopping the generation of the large model, replacing the currently buffered content with the preset prompt information, and then outputting the preset prompt information. According to the scheme of the invention, through risk control detection of input and output stages in a large model dialogue, the detection efficiency is improved, and the problems of long user waiting time and poor experience in full-buffer detection during large model streaming output, frequent detection interface calling and low generation speed in streaming output real-time detection are solved.

The invention provides a security risk detection method and device for text graph large model generation content, and the method comprises the steps: selecting a test category according to a test parameter based on a preset test database, and matching a corresponding data set from the test database according to the test category; inputting the keywords in the data set into the tested large model in batches according to the format of forward cue words; performing content security detection on a corresponding picture or video image generated by the detected large model according to the input cue word; performing watermark identification detection on a corresponding picture or video image generated by the detected large model according to the input cue word; and analyzing to obtain a detection report according to the content security detection result and the watermark identification detection result. According to the test database, the risk of the large model system is externally assessed based on the characteristics of the input prompt word, the content annotation and the output video content, and the accuracy and effectiveness of the detection and assessment method are improved.

The invention belongs to the technical field of digital content security, and discloses a multi-mode AI content security risk traceability and identification detection system. Through obtaining multi-source heterogeneous data of content creation, editing, distribution and presentation of a full link, a content full link feature fingerprint is constructed, and cross-modal consistency analysis and link continuity verification are realized. According to the method, traceability map construction is introduced, potential tampering points are used as map nodes, content risks are quantified through tampering probability weights and risk transfer intensity, and tampering sources and propagation paths are accurately identified. The method has comprehensiveness and accuracy, deep analysis and risk assessment can be carried out on the multi-modal content, and the tampered content is effectively recognized and synthesized.

The invention discloses a big language model generation content security test system and method in a black box scene. The system comprises a jailbreak prompt word library module used for storing jailbreak prompt words for performing security test on a big language model; the violation question and answer pair module is used for storing violation question and answer pairs covering different types; the response acquisition module is used for obtaining a data request packet according to query content formed by the jailbreak prompt word and the query request; the security analysis module is used for calculating the similarity between response data corresponding to the query request and an expected violation answer, taking the similarity as a security score, and inputting the security score into the adaptive optimization module; and the adaptive optimization module is used for optimizing the jailbreak prompt words output by the jailbreak prompt word bank module by using a genetic algorithm according to the security score output by the security analysis module. According to the method and the device, the security of the large language model generation content can be effectively tested.

The application discloses a feature editing method for large model content security, which compares and analyzes the sparse coding features of a chat assistant constructed based on a large language model under positive user input and negative user input, extracts the internal response differences of the model to different semantic directions, and the mechanism can automatically and accurately identify the key feature dimensions highly related to the semantic direction of the target attribute. The model activation is mapped to a sparse feature space by using a sparse autoencoder, and each dimension of the feature has independent and interpretable semantic meaning. By injecting a feature guide vector in the space, the interference of the control process on the text grammar, fluency and information density is significantly reduced. The sparse representation mechanism is introduced to structure the intermediate activation features in the reasoning process of the large language model and to intervene in a targeted manner, so that the reply of the chat assistant to the user input conforms to the preset safety specification, and the safety and controllability of the chat assistant in the interaction with the user are improved.

The present invention is directed to enhanced content security mechanisms for containers. Further, the invention is directed to containers comprising these enhanced content security mechanisms. In particular, the containers of the present invention comprising the enhanced content security mechanism of the invention offer advanced anchoring of the contents (e.g., baked goods or sensitive devices such as medical devices) of the container by preventing: shifting, tipping, breaking, crumbling, rolling or other movement of the contents; thus securing the contents and preventing damage to the contents, i.e., preserving the appearance and integrity of contents contained within.

The invention relates to the technical field of computers, and discloses a model content security control method and system, electronic equipment and a storage medium, and the method comprises the steps: combining a received user input with a historical dialogue record of a current session to form a dialogue context; performing risk identification based on the dialogue context, generating at least one risk category and a corresponding initial risk value, and generating a risk integral corresponding to the current input according to a preset integral strategy; accumulating the risk points to accumulated risk points of the current session; comparing the accumulated risk integral with a dynamic risk threshold value obtained by calculation according to a reference threshold value, a logarithmic function attenuation item of the dialogue round and a user historical behavior adjustment item; and when the accumulated risk integral reaches or exceeds a dynamic risk threshold value, triggering a preset risk management and control measure. According to the method, the risk content can be effectively identified and controlled, attacks bypassing a security mechanism through slow induction can be defended, and the security boundary is not easy to detect.

The invention discloses a webpage content security processing method based on a knowledge graph, and relates to the technical field of webpage content security. The webpage content security processing method based on the knowledge graph comprises the following specific steps: data collection: collecting webpage related internet content data covering texts, images and videos from an internal network of a network service provider, an enterprise or an organization and a website server. According to the method, comprehensive monitoring and analysis of webpage content security are realized, potential security threats, abnormal modes or association relationships in webpage content can be deeply mined by utilizing the reasoning capability of the knowledge graph, including representation learning, association analysis, event tracing, behavior prediction and the like, so that the security threat identification accuracy is improved, and the security threat identification efficiency is improved. And a visual and easy-to-understand visual interface is provided for safety personnel, so that the safety personnel can respond quickly and take corresponding safety measures.

The invention provides a multi-modal large model content security assessment method and device, and the method comprises the steps: inputting a multi-modal harmful data set into a to-be-assessed multi-modal large model, and obtaining an output response of the multi-modal large model; initializing a plurality of large-scale language models, and endowing the plurality of large-scale language models with different role features to form a heterogeneity evaluation system; performing independent evaluation on the multi-modal large model in the aspect of content security by utilizing a plurality of large language models; a plurality of large language models are used for jointly reviewing independent evaluation results, and consistent evaluation suggestions are achieved through cooperative discussion among the models; and obtaining a final evaluation result by using a hard voting mechanism and a soft voting mechanism. According to the method, a plurality of large language models with different role features are introduced for cooperative assessment, and multi-view and multi-level security risk assessment is realized. And through a mode of combining independent evaluation and collaborative review, diversified view angles are effectively integrated, and the reliability and consistency of evaluation results are remarkably improved.

The invention provides a large model dynamic protection method and system based on a zero-trust architecture. The method comprises the steps that a security proxy gateway receives an access request; authenticating an initiating main body of the access request, collecting context information and transmitting the context information to a strategy decision point; the strategy decision point calculates a trust score in real time based on a dynamic trust evaluation model and performs real-time evaluation in combination with an access control strategy to generate a dynamic authorization judgment result; if the access is allowed, forwarding the access request to a large language model server, and performing input security filtering; the large language model server generates response content and performs output security filtering; and returning the final response subjected to the output security filtering to the initiating main body through the security proxy gateway. According to the method, a multi-layer protection framework is constructed, a dynamic trust evaluation model is introduced, and a content filtering layer is deployed, so that continuous permission verification, risk adaptive control and full-link content security protection are realized, and the service security of a large model is effectively guaranteed.

The present application discloses a content security recognition method based on the integration of multiple large language models, belonging to the technical field of data processing. The method includes: receiving a query text, performing word segmentation processing on the query text to generate query text word segments, obtaining sensitive data texts matching the query text word segments in a sensitive database, splicing the query text and the sensitive data texts based on a preset prompt word template to generate a security recognition prompt word for the query text, inputting the security recognition prompt word into at least two large language models, obtaining the output results of the large language models, integrating the output results, and determining the security recognition result of the query text. The present application recalls sensitive data texts through a sensitive database, enhances the security recognition prompt words of the large language models, and improves the accuracy of sensitive word recognition through the integrated processing of multiple large language models.

The invention relates to the technical field of virtual reality interaction, and discloses a virtual reality multi-person interaction method, which comprises the following steps: realizing environment synchronous initialization based on space anchor point calibration and dynamic object loading, and configuring value physical rules and interaction logic; executing a data synchronization mechanism by establishing connection and session management; performing real-time action synchronization and interaction logic processing according to user input processing and action capture; user data privacy and security are protected, and content security and user behaviors are supervised. According to the virtual reality multi-user interaction method and system, by adopting a distributed server architecture and dynamically distributing the nearest edge node according to the geographic position of the user, dynamic load balancing can be realized, QoS priority division is started, low-delay transmission is realized by adopting a UDP + RUDP protocol, communication data is encrypted by adopting an AES-256-GCM, and a key is dynamically updated by adopting a dual-ratchet protocol; and the biological characteristics are converted into irreversible hash values, so that the security of privacy data of the user can be effectively improved.

The invention relates to the technical field of multi-modal content security detection, in particular to a video generation content credibility detection method and system. The method comprises the following steps: dividing an input video into non-overlapping segments, independently modeling through three paths of spatial appearance, motion and semantics, and constructing a unified dynamic representation tensor after alignment and fusion; a cross-fragment bidirectional residual evolution mechanism is introduced to analyze time sequence splitting abnormity, Gaussian perturbation is injected to quantify disturbance response, and fragment vulnerability is evaluated in combination with a residual stability index; quantizing a fragile region by combining confrontation disturbance and designing a multi-scale consistency kernel mechanism to generate a credible score, and judging the authenticity of the video by combining an adaptive threshold value; secondary disturbance is adopted to enhance confidence for fuzzy samples, continuous self-learning optimization is realized through pseudo tag guidance, counterfeit style clustering and model fine tuning, and generalization ability for new counterfeit samples is improved. According to the invention, through multi-dimensional dynamic modeling and a self-learning mechanism, the accuracy and robustness of video credible detection are significantly improved.

The invention discloses a multi-modal video content security examination method, which comprises the following steps of: performing multi-modal data extraction on a to-be-detected video, and performing multi-modal feature alignment; inputting each mode into a pre-trained cross attention and MoE mechanism-based fusion model, wherein the fusion model is used for linking the current single mode with information of other modes through a bidirectional cross attention mechanism and obtaining fusion features of the current mode; and carrying out pooling or clustering operation on the fused features of each mode, then carrying out feature splicing, inputting the spliced features into a classification network model for classification to obtain a classification result, and judging whether the content is illegal or not according to the classification result. According to the method, bad information of visual, text, sound and other modes contained in the video is detected through the multi-mode fusion large model, and feedback is given, so that propagation of bad content is prevented.

The invention discloses an AI-driven convergence media digital audio and video content security auditing method, which comprises the following steps: acquiring audio and video contents published by a platform and user information of a publisher in real time, the audio and video contents comprising live streaming, video files, audio files and image-text contents; feature extraction is carried out on the video file, the text data, the video data and the audio data, and then fusion and vectorization processing are carried out to obtain a comprehensive feature vector; and inputting the comprehensive feature vector into an AI model for detection, determining whether illegal content exists or not, and processing according to a detection result. The AI technology is introduced, intelligent auditing of various modal contents such as videos, audios and images and texts is achieved, the auditing efficiency and accuracy are improved, meanwhile, the auditing strategy is dynamically adjusted according to different scenes and user behaviors, the diversified content auditing requirements of the convergence media platform are met, and the safety and compliance of digital audio and video contents are ensured.

The invention relates to the technical field of natural language processing, in particular to an AIGC content safety monitoring system and method based on dynamic reasoning and context awareness, and the system comprises a semantic graph construction module which is used for extracting entity nouns and predicate verbs in a text according to a received AIGC interaction text flow, generating a semantic concept node set, and sending the semantic concept node set to a database; and performing directed connection and hierarchical nesting on the concepts in the semantic concept node set according to a logic direction according to a subject-predicate-object dependency relationship rule. According to the method, the curvature value of the semantic track is calculated, and the similarity between the direction vector and the center of the sensitive semantic cluster is combined for double verification, so that sudden turning of an intention in a dialogue process or progressive induction to a sensitive field can be perceived, and abnormal mutation can be recognized through a curvature pulse form; therefore, hostile attack behaviors are accurately captured in real time in dynamic interaction, and the defense capability for context dependent attacks and implicit induction behaviors is improved.

The invention relates to the field of large model security testing, and particularly provides a large model output content security testing method and device, and the method comprises the following steps: S1, preparing and managing a test set, a sensitive word library and a regular expression which are required by testing; s2, reading a test set, and obtaining a large model output result according to the test set and the large model interface information; s3, judging whether the output content of the large model is safe or not according to the sensitive lexicon and the regular expression; s4, extracting semantic risk features according to the output content of the large model by using the large model and the oriented Prompt, and automatically storing the semantic risk features after confidence verification; and S5, storing the information result of each request in a file. Compared with the prior art, the test time can be shortened, and the evaluation efficiency can be improved; and the security of the output content of the large model can be effectively evaluated by using a method for dynamically constructing the sensitive word bank by using the output result of the large model.

The invention discloses a large model content security-oriented feature editing method, which comprises the following steps of: comparing and analyzing sparse coding features of a chat assistant constructed on the basis of a large language model under positive user input and negative user input, and extracting internal response differences of the model in different semantic directions; the mechanism can automatically and accurately identify key feature dimensions highly related to the semantic direction of the target attribute. A sparse auto-encoder is utilized to activate and map the model to a sparse feature space, and each dimension of feature has an independent and interpretable semantic meaning. By injecting a feature guide vector into the space, the interference of a control process on text grammar, fluency and information density is remarkably reduced. A sparse representation mechanism is introduced, structural modeling and targeted intervention are carried out on intermediate activation features in a big language model reasoning process, replies input by a chat assistant to a user are guided to conform to a preset safety specification, and the safety and controllability of the chat assistant in interaction with the user are improved.

The invention provides a large-model content security multi-level defense method. The method comprises the steps that an input-processing-output full-link multi-stage cooperative defense framework is constructed, and the full-link multi-stage cooperative defense framework receives to-be-detected user input data and transmits sensitive word detection passing content to an intention recognition module; the intention identification module judges the risk level of the user input data, and routes the user input data identified as medium and high risks to the risk label classification module; the risk label classification module performs risk category classification on the user input data, and transmits the user input data subjected to risk category classification to the security enhancement module; and the security enhancement module generates compliant system response content through a domain fine tuning and reinforcement learning strategy, and returns the system response content to the user after performing security interception processing on the system response content. Through hierarchical defense, dynamic routing and security enhancement module training, a large model security enhancement scheme covering the whole process of input, reasoning and output is constructed.

The invention provides a text classification method and a deep learning model training method and device, and relates to the technical field of artificial intelligence, in particular to the technical field of deep learning, natural language processing, content security and intelligent search. According to the specific implementation scheme, the method comprises the steps of generating semantic features of an input text by using a shared base network; according to the semantic features and a splicing weight matrix of the plurality of tower networks, determining respective classification results of a plurality of text classification tasks respectively corresponding to the plurality of tower networks, the splicing weight matrix being obtained by splicing weight matrixes of the same processing layer of the plurality of tower networks; and according to respective classification results of the plurality of text classification tasks, determining a target classification result of the input text.

The invention relates to the technical field of Internet video content distribution and distribution, in particular to a micro and short play content distribution and distribution system. The system comprises a content management module, a user portrait module, a copyright protection module, a commercial realization module and a data analysis module, the content management module realizes efficient uploading, auditing, storage and intelligent distribution of a movie, and content retrieval and popularization efficiency is improved through accurate classification and tagging management; the user portrait module utilizes data preprocessing, feature extraction and clustering analysis to construct an accurate user model, and user experience is improved based on a personalized recommendation algorithm; the copyright protection module comprehensively guarantees content security through request verification, URL encryption, anti-theft link and a dynamic access control mechanism; the commercial realization module adopts various strategies to improve the platform profitability; and the data analysis module optimizes a content recommendation and distribution strategy through big data collection and mining, and improves the overall operation effect. According to the invention, high-quality experience is provided for the user while the rights and interests of the content creator are ensured.

The invention discloses a digital display safety early warning method based on real-time content monitoring, and belongs to the technical field of digital display system information safety, and the method comprises the steps: carrying out the real-time content capturing based on a content collection assembly embedded in a digital display; generating a content object data set; activating an embedded edge calculation module of the digital display, performing content identification on the character data sequence and the image data sequence, and performing behavior pattern identification on the operation data sequence; in combination with a preset content safety early warning rule and the content recognition result, content safety judgment is carried out, the behavior pattern recognition result is compared with a standard behavior pattern, and operation safety judgment is carried out; and generating a safety early warning instruction to perform safety early warning response. The technical problems that in the prior art, a digital display lacks a real-time content monitoring and safety early warning mechanism, and therefore potential safety hazards such as content violation, sensitive information leakage and operation abnormity cannot be found and responded in time are solved.

The invention relates to the technical field of video analysis, in particular to a video stream multi-mode content security dynamic auditing system and method. According to the invention, through the conjoint analysis of the hue ratio change of the continuous image frames and the focus area heat map, the hue sudden change and concentrated concern area in the video image can be accurately identified, the potential violation fragment can be effectively positioned, and on this basis, the interaction behavior analysis mechanism of the image and audio bimodal data is introduced, so that the accuracy of the image and audio bimodal data interaction behavior analysis is improved. Therefore, the recognition of the sudden behavior is not limited to a single visual or auditory clue any more, the modal jump behavior segment is constructed when the image action and the voice change are synchronously highlighted, and then the time sequence overlapping point location in the modal sudden change process is further captured, so that the positioning precision of the abnormal interruption behavior is enhanced, and the positioning accuracy of the abnormal interruption behavior is improved. And multi-dimensional measurement and scoring of the burst area are realized in combination with trend information of frame-level dominant weight change, so that the recognition coverage rate and the response capability of dynamic content auditing on violation behaviors are improved.

The invention provides an HLS private encryption rapid play starting method and system. The system comprises a server module used for generating a secret key, encrypted video content and an m3u8 file; the client module is used for requesting and decrypting video content and executing security protection, the client module requests an encrypted m3u8 file from the server according to a video URL, and the server returns the encrypted m3u8 file after verifying the legality of the request through legal login; the client side module analyzes the m3u8 file and requests the corresponding encrypted TS slice according to the encrypted information, and the server side provides encrypted video content for the client side through the source station or the CDN; and the storage module is used for storing the encrypted TS slice and the associated m3u8 file. Through a series of innovative designs, on the premise that the video content security is guaranteed, the playing efficiency and the user experience are greatly improved, meanwhile, the complexity and the maintenance cost of the system are reduced, and remarkable technical advantages and practical value are achieved.

The invention discloses a file anti-desensitization self-learning recognition system and method based on information entropy, belongs to the technical field of intersection of natural language processing and content security recognition, and is applied to document screening and risk recognition in a multi-task scene. The implementation method comprises the following steps of: 1, performing character recognition and noise reduction processing on an original file to form a data set; 2, training labeled sample data through small samples, respectively adopting probability distribution of a data sliding window and information entropy to carry out maximum and minimum normalization screening, and further utilizing a fitted linear regression model to form an anti-desensitization word list; 3, screening the anti-desensitization degrees of the sentence segments of the data set by adopting a dictionary tree Trie structure to form an anti-desensitization sentence segment table; 4, marking the chapter-level anti-desensitization degree data set text fragments by using the large model; 5, generating an anti-desensitization report according to the anti-desensitization word and the anti-desensitization degree of the marked anti-desensitization file; compared with the prior art, the anti-desensitization file screening method and device have the advantage that the anti-desensitization file screening accuracy is improved.

The invention discloses a large model generation content risk identification and intervention method based on deep learning, and the method comprises the steps: collecting multi-modal data outputted by a content generation system, carrying out the preprocessing of the multi-modal data, and generating the preprocessed multi-modal input data; encoding and fusing the features of the multi-modal perception model to generate global risk feature representation data; inputting global risk features into a hunting optimization algorithm, and optimizing model structure parameters, risk thresholds and intervention strategy parameters; optimizing parameter-driven risk identification and hierarchical intervention, and outputting risk levels, categories and hierarchical intervention measures; an intervention effect and user feedback are collected, and continuous optimization and self-adaptive evolution of parameters are driven. According to the method, efficient risk identification and intelligent intervention on the large model generation content are realized, and the accuracy and the automation level of content security management are remarkably improved.

The invention discloses a text auditing method and device, electronic equipment, a storage medium and a product, and relates to the technical field of text auditing. The text auditing method comprises the steps that sensitive words in a to-be-audited text are acquired, and position information and suspicious types of the sensitive words are determined; generating statement embedding of the to-be-audited text and sensitive word embedding of sensitive words according to the position information, and determining suspicious type embedding of suspicious types; determining the content security probability of the to-be-audited text according to statement embedding, sensitive word embedding and suspicious type embedding based on a preset sensitive extraction model; and determining the security condition of the to-be-audited text according to the content security probability. According to the embodiment of the invention, the safety condition of the to-be-audited text is determined by combining the sentence pattern characteristics, the position where the sensitive word is embedded and the sensitive word category, misjudgment of the safety condition of the to-be-audited text caused by hitting the preset sensitive word bank is avoided, and the accuracy of text safety judgment is improved.

The invention discloses a big data-based civil and commercial law learning recommendation method and system, and the method comprises the following steps: obtaining user learning behavior data and civil and commercial law knowledge base data, and constructing a multi-dimensional learning feature data set; performing standardization processing on the multi-dimensional learning feature data set to generate a structured feature matrix; constructing a user knowledge mastery degree portrait based on the structured feature matrix, and generating a user-knowledge point association graph; constructing a mixed recommendation model according to the user-knowledge point association graph, and generating a personalized learning path; and receiving user feedback data in real time, updating recommendation model parameters, and dynamically adjusting a learning path. According to the method, data refined modeling is realized through a multi-dimensional learning feature data set and an improved TF-IDF algorithm; optimizing learning path adaptation based on a dynamic weight fusion recommendation model and a real-time feedback mechanism; high concurrent processing and content security are guaranteed by means of a modular architecture and a compliance verification unit, and finally an efficient, self-adaptive and compliance intelligent recommendation system is constructed.