1378 results about "Recommendation model" patented technology

The invention discloses a multi-modal visual arrangement recommendation method, belongs to the technical field of artificial intelligence and data visualization crossing, and realizes visual arrangement recommendation based on multi-modal input analysis, a dynamic mixed recommendation model and an intelligent optimization algorithm. Comprising the following steps: multi-modal intention analysis: realizing intelligent analysis of multi-modal input through combined use of a base model and a fine tuning model, realizing high-precision intention classification in combination with a pre-training language model and a domain adaptation fine tuning technology, and triggering dynamic prompt word recommendation; performing intelligent layout generation: performing global optimization of component space allocation by adopting a genetic algorithm, performing business rule adaptation by combining a constraint solver, and modeling an interaction relationship between components by utilizing a graph neural network; and dynamic mixed recommendation: constructing a three-level recommendation architecture including collaborative filtering, content matching and reinforcement learning. According to the method, a closed-loop recommendation process of user intention-intelligent recommendation-feedback optimization is realized, and the intelligent level of visual arrangement and the user experience are remarkably improved.

The invention discloses a teaching data management method and system based on artificial intelligence, and the method comprises the steps: obtaining a standardized time series data stream according to a heterogeneous data stream generated by a multi-source teaching platform in real time; based on the standardized time sequence data stream, performing classified encryption on the teaching data through a dynamic hierarchical storage strategy based on attribute-based encryption to obtain a security hierarchical storage topological structure; according to a user query request and a teaching scene label, extracting a target data set from the security hierarchical storage topological structure to obtain an enhanced multi-modal teaching data set; based on the enhanced multi-modal teaching data set, generating an interpretable teaching mode graph through a dynamic sub-graph evolution algorithm; and according to the teaching mode map and the real-time teaching feedback data, generating a personalized teaching recommendation strategy through a course-learner dual-channel adaptive recommendation model. According to the embodiment of the invention, the utilization efficiency of teaching resources can be improved, and personalized and intelligent teaching recommendation and decision can be realized.

The invention relates to an article sequence recommendation method and system based on collaborative filtering and an LLM perspective, and the method comprises the steps: obtaining a user historical data set and an article data set, and carrying out the preprocessing, and obtaining an article title similarity matrix, a user behavior sequence, and historical collaborative filtering interaction information; historical collaborative filtering interaction information is enhanced through a large language model; constructing a sequence recommendation model, inputting the user behavior sequence into the sequence recommendation model for training, and correcting the deviation of the user behavior sequence through comparative learning according to the item title similarity matrix and the enhanced historical collaborative filtering interaction information to obtain a trained sequence recommendation model; and inputting the user behavior sequence of the user into the trained sequence recommendation model for prediction, calculating prediction scores, and generating a recommendation list according to score sorting, thereby completing article sequence recommendation. According to the method, the performance of the sequence recommendation system is greatly improved by solving the cold start problem.

An inclusive prompt recommendation system for generative AI utilizes an inclusive prompt recommendation model to provide recommendations of inclusive language to include in a prompt in order to promote inclusivity and diversity of generated content. The inclusive prompt recommendation model is trained to analyze input text to identify situations, such as gaming, storytelling, social media, projects or presentations for work / school, and like, where the user's intent is to generate an image or description of a person. The model is trained to identify patterns associated with ways users have historically incorporated inclusive terminology intext. The system can include an ethical filtering mechanism for ensuring that prompt recommendations do not have language that directly or indirectly promotes bias and / or stereotypes.

The invention discloses a cross-domain privacy protection method and system for advertisement recommendation, equipment and a medium, and the method specifically comprises the steps: carrying out the encryption matching of user behavior data and anonymized equipment data, and generating an initial cross-domain joint feature vector; fusing the initial cross-domain joint feature vector and the disturbance feature vector to form a target cross-domain joint feature vector; splitting a pre-trained advertisement recommendation model into a feature coding sub-module and a reasoning sub-module, deploying the feature coding sub-module to a user adjacent edge node, and retaining the reasoning sub-module in a user local device; and based on the target cross-domain joint feature vector, performing calculation of the feature coding sub-module and calculation of the reasoning sub-module, and uploading the encrypted hidden layer feature vector to a federated learning aggregation server for global model updating. According to the method, cross-domain data utilization and user privacy protection in an advertisement recommendation process are realized, and effective feature vectors are generated for personalized advertisement recommendation while data are guaranteed not to be out of a domain.

The invention discloses a personalized recommendation method driven by user intention recognition, which comprises the following steps of: collecting multi-dimensional information such as browsing records, click behaviors and comment data of a user, and constructing a user behavior data set; preprocessing the collected user behavior data to obtain a session sequence; the current demand of the user is speculated by analyzing the input text and the behavior mode of the user, and the intention of the user is recognized; constructing a user interest preference prediction model according to the identified user intention; and in combination with real-time feedback of the user, the recommendation model is dynamically updated, and the response speed and the individuation degree of the recommendation system are improved. The problem that a traditional recommendation system only depends on static data and does not comprehensively consider real-time feedback of users is solved, self-adaptive ability is injected for personalized recommendation, intention behaviors of different users in multiple scenes can be analyzed and understood, accurate recommendation service is provided, and recommendation accuracy and user satisfaction are improved.

The invention provides an anesthesia virtual simulation training system fusing knowledge, skills and a thinking closed loop. The anesthesia virtual simulation training system comprises a medical knowledge base module, a clinical thinking module, a skill training module, an examination question brushing module, a knowledge graph module and an intelligent platform bottom layer framework. The intelligent platform underlying architecture comprises a data middle platform, an AI engine and a 3D engine, collects student behavior data of each module, constructs a dynamic student ability portrait through a gradient boosting tree algorithm and a collaborative filtering recommendation model, analyzes knowledge blind areas and skill shortages, plans a personalized learning path and pushes targeted training content, and provides a personalized learning result. A closed-loop process of evaluation, learning, practice and re-evaluation is formed; and deep fusion of theoretical knowledge, clinical thinking and skill operation is realized through a cross-module collaboration mechanism. The problems that traditional anesthesia teaching is high in practical operation risk, scattered in resource and insufficient in individuation are solved, and the clinical comprehensive ability and teaching quality of anesthetists are effectively improved.

The invention belongs to the technical field of article recommendation, and provides an article recommendation method and system based on big language model enhanced graph representation learning, and the method comprises the steps: obtaining a user-article graph recommendation data set; extracting semantic knowledge features of the obtained graph recommendation data set; according to the extracted semantic knowledge features and an article recommendation model, completing article recommendation learning; wherein the article recommendation model adopts a big language model to enhance a graph representation learning model, embeds semantic similarity through the big language model, adopts an adaptive graph structure learning mechanism to identify a semantic boundary so as to determine structure information, and performs bidirectional knowledge interaction transmission on semantic and structure information; interactive noise is filtered and relieved in combination with information bottleneck regularization, optimization of an article recommendation model is carried out with the purpose of minimizing a comprehensive multi-objective loss function, and article recommendation is completed by calculating preference scores of articles.

The embodiment of the invention relates to the technical field of artificial intelligence, and provides an intelligent recommendation method and system based on the plasticizing industry, and the method comprises the steps: obtaining multi-source heterogeneous data from a plurality of third-party data sources of the plasticizing industry; constructing the multi-source heterogeneous data into a dynamic knowledge graph in combination with a plasticizing industry knowledge base; performing demand prediction on the dynamic knowledge graph based on a knowledge-guided hybrid particle swarm algorithm to obtain a first plasticizing recommendation strategy; enhancing a recommendation model through causal reasoning, and performing anti-fact prediction on the dynamic knowledge graph in combination with industry real-time change data to obtain a second plasticizing recommendation strategy; the first plasticizing recommendation strategy and the second plasticizing recommendation strategy are dynamically fused based on Bayesian optimization, a plasticizing industry recommendation report is generated, the plasticizing industry recommendation report comprises the industry flow relation change trend and the corresponding industry hotspot recommendation, and a user is assisted in marketing decision making. The method can shorten the time consumption of the whole supply-demand docking process, effectively alleviates the information asymmetry problem, and remarkably improves the industry operation efficiency.

The invention relates to the field of traditional Chinese medicine clinical auxiliary diagnosis, and discloses an acupuncture point recommendation method, an acupuncture point recommendation model obtaining method, an acupuncture point recommendation model obtaining device, acupuncture point recommendation equipment and a medium, and the acupuncture point recommendation model obtaining method comprises the following steps: collecting patient medical record data to form a medical record data sample set; preprocessing the medical record data sample set to obtain a standard medical record data sample set; inputting the standard medical record data sample set into a preset language model, performing preliminary training on the language model based on the standard medical record data sample set, so that the language model establishes a semantic mapping relationship between the patient symptom information and the acupuncture points, and performing parameter fine tuning and strategy optimization on the preliminarily trained language model in sequence, obtaining an acupuncture point recommendation model; according to the acupuncture point recommendation model obtaining method, the accuracy and individual adaptability of acupuncture clinical treatment are improved, the diagnosis and treatment efficiency is optimized, and a scientific and systematic treatment scheme is provided for a patient.

A dynamic marketplace system leveraging store and warehouse mobility features a neuroevolution (NE) engine, an electronic device, and a request handler facilitating communication between the electronic device and the NE engine. The NE engine interfaces with a data storage system and an event handler receiving real-time event data from a public cloud services processor. An intentions handler interprets user intention data to predict user behavior. The NE engine, integrated with a processor, generates a predictive evolutionary model for the marketplace based on request, event, and intention data. An AI agent processor within the NE engine creates a recommendation model for mobile retail vendors, devises route plans, and deploys vendors to strategic locations.

The invention provides a commodity multimedia recommendation method and system combined with RPA and AI.The method comprises the steps that firstly, a current interaction behavior flow of a user and a commodity multimedia interface is recorded in real time through an RPA interaction capture module, the current interaction behavior flow comprises an operation triggering time sequence and an attention staying track, then an intention evolution track is extracted based on a preset historical interaction mode library, and the intent evolution track is extracted; the method comprises the following steps of: generating a dynamic matching rule set according to an intention evolution track, including an association constraint condition and a priority ranking logic, inputting the dynamic matching rule set into a pre-trained AI recommendation model, performing rule matching screening on a candidate commodity multimedia content set, generating a screening result, and outputting the screening result to a user. And finally, a recommendation sequence is rendered in real time according to a screening result through an RPA display arrangement module, and the display size and the text typesetting style are adjusted, so that the individuation degree of commodity multimedia recommendation and the user experience are improved.

In various examples, session-based recommender model systems and applications are disclosed. Systems and methods are disclosed that use a cosine similarity loss during the training of a machine learning model to train the model to generate an item recommendation based on predicting a next item from a sequence of prior items selected within a session. A recommendation model is trained based on training data that represent an ordered sequence of user interactions with the set of items. A set of item embeddings is generated for the set of items. The recommendation model is trained to predict a session embedding that represents a user behavior pattern from a sequence of item embeddings. A cosine similarity loss computed from the session embedding and the item embeddings is used to train the recommendation model. The cosine similarity loss may include both positive and negative cosine similarity components.

The invention discloses a sequence recommendation method based on a time-aware hierarchical attention network, and the method comprises the steps: obtaining a user interaction sequence of a target user, and obtaining a time interval sequence and a time context sequence of the interaction of the target user based on the user interaction sequence; inputting the user interaction sequence, the time interval sequence and the time context sequence into a trained sequence recommendation model to obtain a user interest representation based on a time interval and a user interest representation fused with time context information; and determining a correlation score with the user interaction sequence according to the user interest representation, obtaining a final correlation score through weighted fusion, determining a recommended item list according to the final correlation score, and recommending the recommended item list to a target user. According to the method, the problem of insufficient noise interference and context fusion in time information modeling of a traditional method is solved.

The invention provides a goods source recommendation method and device and electronic equipment, and belongs to the technical field of logistics transportation. In the method, a real-time search request of a target driver is received, and a first preset number of first historical click behavior data corresponding to the target driver is obtained; performing generalization processing on each piece of first historical click behavior data and the real-time search request to calculate a total generalization score, screening the first historical click behavior data according to a preset generalization score to obtain second historical click behavior data with higher relevancy with the real-time search request, and determining a second behavior sequence; and finally, inputting the second behavior sequence into the target recommendation model, and generating a recommendation result sorted according to the relevancy. According to the method provided by the invention, the long-term interest and periodicity rule of the driver can be modeled by effectively utilizing the second historical click behavior data with a longer coverage range and higher correlation with the real-time search request under the condition that the online reasoning performance is limited, and the accuracy of a goods source recommendation result is effectively improved.

The invention relates to a semantic collaboration modeling method and system for multi-modal sequence recommendation, belongs to the technical field of multi-modal sequence recommendation, and aims to solve the problems that an existing method depends on an article ID, multi-modal semantic collaboration signals are difficult to mine, the cross-scene migration capability is weak, and the representation precision is insufficient. The method comprises the steps that text information and image information of an article are extracted according to historical interaction behaviors of a user, and multi-level semantic representation is extracted through a multi-mode encoder; a multi-head attention module with a one-way mask is adopted to capture semantic collaboration signals, and the semantic collaboration signals are integrated into initial modal representation; performing fine-grained semantic focusing and optimization by using a hybrid expert structure; obtaining a final modal representation of the article through a hybrid expert fusion module; and constructing a multi-modal behavior sequence of the user based on the final modal representation of the article, and calculating preference scores after coding by a sequence recommendation model to realize next-step interaction prediction of the user. According to the method, recommendation accuracy, generalization ability and cross-scene knowledge migration efficiency can be remarkably improved.

The invention discloses an internet consumption intelligent monitoring system and method based on big data, and relates to the technical field of big data user behavior monitoring and intelligent recommendation systems, and the system comprises a behavior statistics module, a significance drift judgment module, and an update control module. According to the system disclosed by the invention, 'user interest fluctuation 'and'systematic behavior offset' are effectively distinguished, a more stable user behavior monitoring mechanism with a lower misjudgment rate is constructed, continuous dynamic description of a user behavior state is realized, a more accurate and more stable input basis is provided for subsequent offset identification, and on the premise that the robustness of the system is kept, the user behavior monitoring accuracy is improved. Important behavior changes possibly needing model response are sensitively recognized, the method is particularly suitable for user behavior collective migration scenes caused by rapid evolution of crowds or emergencies, and high-efficiency, low-cost and differentiated trigger type intelligent updating of recommendation models is achieved; and the adaptive capacity and the personalized accuracy of the recommendation system to the user interest migration are obviously improved.

The embodiment of the invention provides an intelligent recommendation method and system based on child digital picture book reading. The method comprises the following steps: constructing a recommendation model based on self-supervised learning and a convolutional neural network CNN; in the self-supervised learning module, feature representation of a comparison mechanism optimization model combined with user cognition is introduced; constructing a dynamically updated user portrait, wherein the user portrait comprises a current cognitive level label of the user; in response to a reading request of a user, self-adaptive fusion is carried out on a user portrait and the reading request, an age-worthiness evaluation factor of the picture book is introduced in the fusion process, and the age-worthiness evaluation factor is determined based on the matching degree of the character difficulty, the picture complexity and the user cognitive level of the picture book. And taking the fused label vector as the input of a trained recommendation model to obtain a recommendation list. According to the embodiment of the invention, the feature representation of the model is dynamically optimized in combination with the cognitive development stage of the user, the reading ability difference of different age groups is accurately adapted, and the recommendation accuracy and adaptability are improved.

The invention discloses a social group evaluation method and system based on double-space fusion calibration, and the method comprises the steps: firstly, extracting user preference features based on user-news interaction data and news type tags, constructing a fixed user portrait through employing a statistical analysis and maximum preference strategy, and completing the division of the user portrait; then, starting from the matching space, sorting the users in the matching space through joint sorting of the behavior activeness and the behavior coupling degree; meanwhile, starting from a semantic space, generating a portrait semantic expression by utilizing a large language model, generating a semantic center point in combination with user behaviors, and sorting the users in the semantic space based on a vector distance; on this basis, the sorting information of the matching space and the semantic space is fused, the semantic consistency and the behavioral representativeness of the users are balanced, and finally the user groups with comprehensive representativeness under each type of portraits are screened out; further, based on the high-frequency behavior record and portrait features of the representative user, constructing an injection simulation behavior sequence, selecting low-interaction news items, and generating an evaluation sample with interference features; and finally, retraining a recommendation model, evaluating recommendation response change of each portrait group on low-interaction information, and quantifying anti-interference scores of the system under different portrait consistency conditions, thereby revealing robustness difference of the recommendation system, and providing support for security enhancement and portrait recognition. According to the method, the performance difference and the anti-interference capability of different user portrait groups in a recommendation system can be described.

The invention discloses a financial product recommendation method, system and device based on large language model assistance, and the method comprises the steps: obtaining user-related data and financial product-related data, and carrying out the feature extraction, and obtaining user-related features and financial product-related features; constructing an interpretation generation module, wherein the interpretation generation module comprises a large language model, an auto-encoder network and a gating network; combining the weighted explanation features, the user features, the financial product related features, the user financing preference change features, the context features and the recommendation reason features to form a feature matrix, and training a traditional recommendation pre-training model through the feature matrix to obtain an improved recommendation model; and processing the to-be-recommended data based on the improved recommendation model to obtain a financial product recommendation result. The large language model and the traditional recommendation model are deeply fused, so that the interpretability of the recommendation system is remarkably improved, and the recommendation accuracy of the model is further enhanced.

The invention discloses a surgical instrument automatic scheduling method, system and terminal based on pre-operation scheduling, and the method comprises the steps: obtaining pre-operation scheduling information which comprises an operation style, an operator, a time period and inter-operation scheduling; a standard instrument set template is obtained according to the operation mode, historical instrument use behavior data of the operator is obtained, and the historical use behavior data is used for dynamically correcting the standard instrument set template to obtain a personalized template; acquiring instrument inventory information in the time period, and intelligently scoring the standard instrument package template and the personalized template according to the instrument inventory information and the recommendation model to obtain a target template; obtaining a scheduling task list according to the target template, the operation style, the time period and the inter-operation arrangement; and inputting the scheduling task list into an automatic transfer system, generating a route task package, and packaging, delivering and delivering the instruments according to the route task package and the scheduling instruction. According to the method, the instrument set is automatically generated and optimized, and the surgical instrument configuration efficiency is remarkably improved.

The invention relates to the technical field of recommendation systems, and discloses a cross-domain recommendation and model training method and device, and the method comprises the steps: obtaining and coding source domain behavior data through a source domain user behavior log collection module, and generating a user interest representation; target domain interaction data are collected, a user mapping relation is constructed, and preference alignment between the source domain and the target domain is achieved; constructing a joint training set, taking the source domain interest vector as auxiliary input, training a cross-domain recommendation model and generating model parameters; and finally, generating a personalized recommendation result based on the interest state of the target domain user, and dynamically adjusting a recommendation strategy according to user feedback. The device comprises a source domain representation construction module, a target domain alignment module, a migration training module and a recommendation feedback module, and interest migration and cross-domain recommendation performance optimization among multi-source data can be realized.

The embodiment of the invention provides a content recommendation method and device and a double-tower content recommendation model training method and device, and the content recommendation method comprises the steps: obtaining the content information of to-be-recommended content and the user information of a target user in response to a content recommendation task; the user information and the content information are input into a double-tower content recommendation model, recommended content for the target user is obtained, the double-tower content recommendation model comprises a user tower and a content tower, the user tower is used for extracting user features based on the user information, and the content tower comprises a feature adaptation layer; the feature adaptation layer is used for obtaining target content features corresponding to a task target of the content recommendation task based on the content information, and the recommendation content is obtained by decoding based on the target content features and the user features. The content representation can be dynamically adjusted according to the task target through the feature adaptation layer, so that the same content presents differentiated feature expression under different tasks, and the adaptation precision of the recommendation result to the specific task target is improved on the premise of not changing the double-tower structure.

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.

Devices, systems, methods, and processes for recommendation and update of network policies. Existing network policy update solutions rely on human intervention in monitoring and analyzing traffic patterns in a network, checking for policy compliance, detecting any policy violations, and even updating new policies in the network. However, manual processes are prone to human error, introduce significant delays, and lack scalability and objectivity. To address these issues, an automated system is provided that monitors traffic across a network (in real-time or near real-time) and detects violations in a set of network policies associated with the network. The system utilizes one or more recommendation models to process network flow data and network inventory data, and generate one or more policy update recommendations to resolve the detected policy violations. The system further enforces the one or more policy update recommendations on various network devices within the network to resolve the detected policy violations.

The invention relates to a place recommendation method based on a shared hypergraph mask. The method comprises the following steps: selecting a public data set and constructing a recommendation model; constructing a trajectory hypergraph by using long and short trajectories of a user, inputting the trajectory hypergraph into the model, calculating user characterization and trajectory characterization of the long and short trajectories through an auto-encoder module in the model, and calculating fusion characterization of the user characterization and the trajectory characterization; a loss function # imgabs0 # is constructed by adopting a hypergraph, global and local loss functions # imgabs1 # and # imgabs2 # are constructed by using user characterization of long and short tracks and track characterization through a double-level contrast learning module in a model, local tracks obtained after position information is embedded are embedded into Eu, a prediction result # imgabs3 # is obtained through a Transformer encoder, and a loss function # imgabs4 # is constructed at the same time, and four losses are constructed through the prediction result # imgabs3 # and the loss function # imgabs4 #. The loss function forms a model overall loss function # imgabs5 # for training the model, and a final trained recommendation model is obtained. By adopting the method, discrete data can be effectively processed, the characteristics of long and short term behavior modes can be accurately captured, and deeper and more accurate interest point recommendation can be provided.

The invention discloses a drug recommendation model construction method, a drug recommendation method, a drug recommendation device and drug recommendation equipment, and belongs to the technical field of drug recommendation. The method comprises the following steps: constructing an EDRMM model; then constructing a loss function including binary cross entropy loss, multi-label prediction loss, drug interaction loss and regularization constraint terms; constructing a training sample set; and then training an EDRMM model by using the training sample set and the loss function to obtain a trained EDRMM model as a drug recommendation model. The fine-grained electronic health record data, the historical electronic health record selection process and the regularization constraint term are introduced, noise of the historical attribute data is reduced, the DDI occurrence risk rate is better controlled, and the drug recommendation quality is improved.

The invention relates to the technical field of information management, and discloses an adaptive decision logistics product management method. Comprising the steps of integrating a heterogeneous data cleaning algorithm, performing feature extraction and standardization on structured and unstructured data by adopting a Transform-based cross-modal alignment model, and embedding an anomaly detection model; a self-adaptive rule generator based on reinforcement learning constructs a reward function, and dynamically optimizes a subsidy rule strategy; a time sequence prediction model is combined to predict a future work clothes application demand trend, and an inventory management system is linked to realize a pre-distribution strategy; a graph neural network is adopted to construct an organizational structure knowledge graph, and an optimal approval path is generated; an approval risk prediction model is introduced, approval risks are predicted, and a grading early warning mechanism is triggered; generating a personalized chemical clothes recommendation list based on a mixed recommendation model of collaborative filtering and a knowledge graph; integrating a computer vision model, and automatically matching the size of the work clothes; the method has the advantages of improving employee satisfaction and reducing company cost.

The invention discloses a recommendation system-oriented privacy sensitive parameter identification and accurate deletion method, and relates to the technical field of recommendation system model optimization and privacy protection. In order to solve the problems that privacy sensitive parameters in an existing recommendation system model are difficult to accurately position and the model performance loss is too large after deletion, according to the evaluation thought of parameter importance and data feature relevance, importance scores of model parameters on two types of data sets are calculated respectively by dividing a forgetting set and a reservation set, and the privacy sensitive parameters are obtained. The privacy sensitive parameters with the forgetting set importance higher than the reserved set importance are screened out, and mask zero setting processing is carried out. According to the method, the model does not need to be retrained or subjected to weight updating, sensitive parameter recognition and deletion can be completed only through single-time forward propagation, and the original recommendation performance of the model is reserved to the maximum extent while the privacy security of a recommendation system is guaranteed. According to the method, accurate identification of privacy sensitive parameters is realized on a mainstream recommendation model, the reduction range of model recommendation accuracy after deletion is controlled within an acceptable range, and the privacy leakage risk is remarkably reduced.