219 results about "Collaborative filtering" patented technology

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 relates to a big language model enhanced knowledge adaptive learning planning system, and belongs to the field of intelligent education. The system comprises a knowledge center module, a learner portrait module, a path planning module and an intelligent learning guiding module, and the knowledge center module extracts entities and relationships from a multi-modal data source by using a large language model to construct a knowledge graph; the learner portrait module collects multi-dimensional learning data of the user and maps the multi-dimensional learning data to corresponding nodes of a knowledge graph, and dynamically deduces a learner portrait through a Bayesian knowledge tracking model; the path planning module generates an initial learning path based on the knowledge graph and the learner portrait, establishes a collaborative filtering analysis model, predicts and optimizes the expected effect of the current learner following the initial learning path in combination with a Bayesian knowledge tracking model, and finally generates a target learning path; and the intelligent learning guiding module generates a standardized knowledge card for each knowledge node on the target learning path through a security retrieval enhancement generation technology.

The invention belongs to the technical field of nuclear power station file management, and particularly relates to an intelligent nuclear power station file classification and retrieval system based on artificial intelligence. The file acquisition module is responsible for uniformly acquiring and accessing multi-modal files in the nuclear power station; the preprocessing and natural language processing module is used for carrying out deep processing and semantic analysis on the collected multi-modal document; the dynamic classification module is used for carrying out multi-dimensional automatic classification on the multi-modal document based on a deep learning model and a nuclear power industry knowledge graph; the dynamic knowledge updating module is used for continuously monitoring new documents and user feedback and keeping adaptability to industry changes; the intelligent retrieval module is used for providing a query function for a user through semantic vectorization and a large-scale rapid retrieval technology; and the intelligent recommendation module is used for providing targeted document recommendation for the user through association analysis and collaborative filtering. According to the method, the problems of low classification efficiency, insufficient retrieval accuracy, weak dynamic adaptive capacity and limited multi-modal data processing capacity are solved.

The invention provides a class incremental learning method based on prompt guidance and multi-modal fusion, and relates to the technical field of artificial intelligence and computer vision. The method comprises the following steps: firstly, performing semantic extension on a category label, and constructing semantic enhanced text representation through a text encoder; then block embedding and hierarchical feature extraction are carried out on the input image by using a pre-trained visual encoder, a cross-modal unified embedding space is constructed, a bimodal prompt gating fusion module is introduced into the unified embedding space, and adaptive weighting is carried out on text prompt and image prompt according to gating weight to generate fusion prompt; through a bimodal prompt collaborative filtering module, screening out a prompt set most relevant to the current task according to the similarity of the semantic features of the image and the text; the pre-training backbone network is frozen in the increment stage, only prompt parameters and fusion layer weights are optimized, a joint loss function is used for parameter updating, finally, image and text data are input in the reasoning stage, cross-modal similarity is calculated, and a classification prediction result is output.

The invention provides a multi-modal data joint query analysis method and system supporting natural language interaction, and relates to the technical field of data query analysis. Historical operation data, audio data and text data of a user on an intelligent search platform supporting natural language interaction are collected; modeling the historical operation data to obtain a user preference vector, performing voice recognition and text standardization processing on the audio data to obtain standard query data, and integrating the standard query data and the text data into context information; an attention mechanism-based algorithm is used for semantic understanding and intention recognition to obtain an intention feature vector, and the intention feature vector is fused with a user preference vector to obtain a classification result; and finally, based on the result, querying in a preset multi-modal database through a collaborative filtering algorithm to obtain a joint query result, so that personalized accurate query of the multi-modal data under natural language interaction can be realized, and the result fits the intention and long-term preference of the user.

The invention provides an employee work assessment recommendation method and system based on a collaborative filtering algorithm, and relates to the technical field of employee work assessment, and the method comprises the steps: extracting a task dependency relationship and time consumption data from a project management system, and constructing an employee task dependency graph; secondly, collecting and analyzing work interaction information containing document operation and task marking features, and identifying task switching critical points to generate quantitative indexes; constructing a similar employee group and extracting an efficient working mode by means of a collaborative filtering algorithm after establishing an evaluation data set through space-time alignment of the two; by taking the task dependency graph as an environment state and the efficient mode as a reference, generating a task sequence optimization scheme by using a reinforcement learning model; according to the scheme, the staff group composition is adjusted according to the scheme execution effect, the optimization model is fed back, finally, a comprehensive evaluation result is generated based on the scheme and actual task sequence matching degree and the quantitative index stability, the task sequence optimization and physiological state adjustment ability can be accurately evaluated, and the work assessment scientificity and the task execution efficiency are improved.

The invention discloses a person and post matching method and system based on a bidirectional collaborative filtering algorithm, and the method comprises the steps: assisting a job seeker to clearly input job-seeking basic information and transmit job-seeking appeals in a flexible manner through an intelligent interaction interface integrating voice input and large language model analysis; the method comprises the following steps: collecting data through interaction information collection, resume information analysis and recruitment demand decomposition, carrying out data cleaning and label mining, generating a job seeker dominant label, a job seeker implicit label, a recruiter dominant label and a recruiter implicit label based on an ontology, preliminarily constructing an interestingness preference model of the job seeker and the recruiter, and establishing an interestingness preference model of the job seeker and the recruiter; and then calculating a job-seeking label data set and a recruitment position label data set based on an improved collaborative filtering algorithm, calculating a person-position bidirectional matching coefficient by integrating dominant demands and implicit demands of both parties, recommending a working position with a higher bidirectional matching degree for a job-seeker position, and improving the accuracy of position recommendation.

The present disclosure relates to systems, methods, and non-transitory computer readable media that utilize collaborative filtering and a reinforcement learning model having an actor-critic framework to provide digital content items across client devices. In particular, in one or more embodiments, the disclosed systems monitor interactions of a client device with one or more digital content items to generate item embeddings (e.g., utilizing a collaborative filtering model). The disclosed systems further utilize a reinforcement learning model to generate a recommendation (e.g., determine one or more additional digital content items to provide to the client device) based on the user interactions. In some implementations, the disclosed systems utilize the reinforcement learning model to analyze every negative and positive interaction observed when generating the recommendation. Further, the disclosed systems utilize the reinforcement learning model to analyze item embeddings, which encode the relationships among the digital content items, when generating the recommendation.

A computer-readable medium stores instructions that cause one or more processors to perform mental wellness activities. The activities can include providing a user interface to receive inputs about a user's challenges and strengths, generating a query to a large language model (LLM) to create a narrative where challenges are addressed based on strengths, and displaying the narrative. The activities can include providing an interface for a therapeutic exercise, showing the exercise's intention, guiding the user through the exercise, and collecting feedback afterward. The activities can include recommending therapeutic activities by analyzing user data to identify preferences or emotional states. Based on the amount of aggregate user data, either predefined rules or collaborative filtering is used to make personalized activity recommendations.

The application discloses a collaborative filtering recommendation method and system based on adaptive noise adding privacy protection, and the method comprises the following steps: suspected abnormal users are screened out by using a DBSCAN algorithm, and true abnormal users are determined by using a box chart, so that abnormal user data in a user data set is screened out; the similarity between each user is calculated, different sizes of noise are added according to the similarity value, and a user similarity noise matrix is constructed; an initial centroid is selected according to the user similarity noise matrix by using a k-means algorithm, and iterative updating is performed, so that user data clustering is realized; a recommendation list is obtained according to a neighbor set of a target user; and items are recommended to the target user according to the scores of the items in the recommendation list. Through the technical scheme of the application, the accuracy of the recommendation result is improved, the user privacy is protected, the problem that the clustering centroid deviates greatly is avoided, and the problem that noise is continuously accumulated in the iteration process to cause the final recommendation result to be inaccurate is avoided to a certain extent.

The invention discloses an electronic manufacturing industry production line UPH intelligent prediction method, system, device and medium, and the method can learn a deep interaction relationship between a line body and a product through adversarial training through a generator of a collaborative filtering generative adversarial network, so that even if a new product lacks enough historical data, the production line can be predicted more accurately. The generator can also find a mode similar to other line bodies in a potential space and generate a relatively reasonable UPH, and for an old product with a theoretical UPH and an actual UPH, the relatively reasonable UPH can be generated only by correspondingly inputting a current line body feature vector and the theoretical UPH of the product in the corresponding line body; according to the method, the difficulty that in the initial production stage of a new product, due to lack of reference bases, UPH can be generally formulated only according to the same industry standard or the average yield per hour of the production line can be solved; and meanwhile, the problem of UPH updating hysteresis caused by personnel or equipment can be solved.

The application provides a customer call reason determination method and device, wherein the method comprises the following steps: constructing a feature vector of all customers in a customer service channel, and constructing a score vector of all IVR service nodes of the customers; obtaining transaction data and behavior data of other channels except the customer service channel; using a double-layer random forest model to filter out target features for the customer service channel from the transaction data and the behavior data; and performing customer call reason modeling according to the feature vector, the score vector and the target features, obtaining a customer call reason model, and predicting the customer call reason through the customer call reason model, so as to solve the problem that in the related art, the prediction accuracy of the model-based collaborative filtering for predicting the customer call reason is very low for customers with few features of the customer service channel.

The invention discloses a rail transit protection area engineering instrument detection method based on vision and collaborative filtering. The method comprises the steps that S1, vision, sensing, history and other multi-source heterogeneous data are collected and normalized; s2, calculating the average error rate of each data source in real time, and dynamically updating the reliability weight of each data source according to the average error rate; s3, multiplying the normalized original feature value by the corresponding real-time reliability weight, and constructing a dynamically weighted data feature matrix; s4, applying a pre-training model based on a collaborative filtering matrix decomposition principle to decompose the weighting matrix, and generating a standardized instrument state feature vector fusing the potential association of the multi-source data and the real-time reliability of the multi-source data; and S5, inputting the feature vector into a risk assessment model for analysis, and comparing the feature vector with a dynamic risk threshold to realize accurate early warning. According to the invention, the accuracy and robustness of detection are obviously improved, and the crossing from passive monitoring to active early warning is realized.

The invention relates to a Citrus Huanglongbing early collaborative sensing method based on Internet of Things, in particular to the field of Internet of Things, by performing confidence-driven local collaborative filtering at a sensor node, environmental noise and individual difference interference are effectively inhibited, the signal-to-noise ratio of early weak disease features is remarkably improved, and the accuracy of the Citrus Huanglongbing early collaborative sensing method is improved. A dynamic space-time association graph constructed by a convergence gateway accurately depicts space-time association of an abnormal mode, a cloud heterogeneous graph neural network further excavates a deep mode which is robust to global interference and sensitive to local coupling anomaly from the graph, and finally, a collaborative training mechanism based on federated learning is utilized to improve the robustness of the abnormal mode on the premise of protecting data privacy. And the group intelligence of distributed data is gathered, and a lightweight diagnosis model capable of reasoning in real time on the edge side is trained, so that the early-stage, accurate, low-power-consumption and sustainable-evolution intelligent perception of the citrus huanglongbing is realized.

The invention is applicable to the technical field of recommendation, and provides a collaborative filtering recommendation method based on double popularity constraints, which comprises the following steps of: firstly, constructing a user item score table and an item attribute relation table; secondly, obtaining normalized item popularity and normalized attribute popularity, and obtaining constraint factors by using double popularity; performing similarity calculation by using a dual popularity constraint factor and a Pearson similarity formula to obtain adjacent users of the target user; according to the scoring information of the adjacent users of the target user, performing prediction scoring on the item set which is not scored by the target user to obtain a prediction scoring table of the target user; and finally, extracting first N items with the highest score value from the prediction score table as recommendation results and providing the recommendation results to the target user. According to the method, by constraining the item popularity and the attribute popularity, the popularity deviation is relieved, and the long-tail effect is relieved.

The present application relates to a kind of multi unmanned mine car-oriented charging and battery swapping collaborative optimization method and system, belong to charging and battery swapping collaborative optimization technical field.It includes: collecting mine car state, infrastructure state and task plan data, based on the improved collaborative filtering algorithm for each mine car generation personalized energy supply recommendation list;Based on personalized energy supply recommendation list, in combination with task plan, determine overall service execution sequence, execute service execution sequence and evaluate the service capability and expected income of each charging and battery swapping facility;In combination with vehicle state data, infrastructure state data, expected service capability and income evaluation report, use deep reinforcement learning algorithm to generate mine car dispatch instruction set, and through multi-aggregator cooperation mechanism optimization as cross-regional collaborative scheduling instruction set;Calculate the comprehensive service performance index when executing cross-regional collaborative scheduling instruction set, if better than preset service performance index, then execute the scheduling scheme;Otherwise, trigger degradation strategy, readjust scheduling scheme.

The application discloses a livestock data search method and system based on multi-source heterogeneous fusion. The application adopts a four-layer architecture of data acquisition, data processing, retrieval service and user interaction: through adaptive acquisition integration of active grabbing and passive access, covering public and non-public data; using rule engine and machine learning to clean data and establish meta model, standardizing and converting structured, semi-structured and unstructured data, constructing livestock field knowledge graph and using hybrid storage of relational database, non-relational database and graph database; providing personalized recommendation through user portrait construction and collaborative filtering algorithm; using RBAC permission management and data desensitization to ensure compliance. The application realizes efficient integration, accurate retrieval and intelligent service of multi-source livestock data.

The invention relates to a personalized library book recommendation method and system based on conditional sharpening collaborative filtering, and belongs to the technical field of data processing and information recommendation. The method comprises the following steps: constructing a user-book interaction matrix according to an obtained historical borrowing record of a library user, thereby constructing a collaborative signal matrix, normalizing the collaborative signal matrix, and performing linear weighted fusion to obtain an initial guidance matrix; performing exploratory remodeling on the initial guidance matrix to obtain a remodeled guidance matrix; performing fuzzy processing on the user-book interaction matrix to generate a basic preference matrix; and sharpening the basic preference matrix based on the remodeled guidance matrix to obtain a book recommendation score matrix, and performing personalized recommendation to the user based on the book recommendation score matrix. The technical problem that in the prior art, due to the fact that recommendation results are excessively concentrated in the known interest field of a user, the information cocoon house effect is generated, and the novelty and diversity of the recommendation results are limited is solved.

The invention belongs to the technical field of internet information recommendation, and discloses a collaborative filtering recommendation method and device based on user grading, and the method comprises the following steps: S1, user grading processing: obtaining user data of a target platform, grading users based on a preset grading dimension, obtaining at least two user levels and a level weight corresponding to each user level; s2, data preprocessing and feature extraction: collecting user behavior data and content data, performing data cleaning, constructing a user-content interaction matrix, and extracting user features and content features; according to the scheme, the user level weights are deeply fused into similarity calculation and recommendation priority ranking, so that the preferences of high-value users and payment users can be remarkably amplified, and therefore, a more accurate high-quality recommendation result which better conforms to the identities and requirements of the users is obtained, and the satisfaction degree, retention rate and commercial conversion efficiency of core users are effectively improved.

The invention relates to the technical field of data processing, and particularly provides an industrial supply and demand recommendation method, system and device and a storage medium, and the method comprises the steps: constructing a case library of two-dimensional indexes based on predefined industry and link classification, and associating each index item with a service provider case set; collecting diagnosis data of a demand side enterprise, wherein the diagnosis data at least comprises industries, links and multi-dimensional diagnosis indexes; recalling an initial service provider set from an index database based on industry and link information in the diagnosis data; constructing an enterprise portrait based on the diagnosis data, calculating the matching degree between each service provider and the enterprise by using a multi-level dynamic weight model at least comprising a basic matching layer, a context awareness adjustment layer and a collaborative filtering layer, and generating a preliminary recommendation list; and filtering the preliminary list according to a preset access rule, and outputting a final recommendation list. According to the invention, efficient, accurate and reliable supply and demand matching is realized.

The invention provides a charging station recommendation method and device based on hybrid collaborative filtering and a medium, and belongs to the technical field of vehicles. The method comprises the steps of obtaining a charging event of a vehicle of each user, performing data conversion processing on the charging event, determining an interaction mapping relationship between each user and each historical charging station, and extracting charging interaction information; and determining a charging user type of each user according to the charging event, and determining a corresponding collaborative filtering recommendation strategy according to the charging user type. And determining the charging preference degree of each user according to the charging interaction information based on the interaction mapping relationship and a corresponding collaborative filtering recommendation strategy. And generating a charging station recommendation list according to the charging preference degree and each historical charging station, constructing a preferred charging station set of each user according to the charging station recommendation list, and rendering and displaying the preferred charging station set. Accurate and personalized charging station recommendation is provided, the set highly conforms to the personality of the user, potential interested charging stations can be mined for the user, and the charging experience is improved.

The invention relates to the technical field of member operation and big data application, and discloses a Changbai Mountain specialty membership system point appreciation and big data consumption implementation method, which comprises the following steps of: acquiring member, supply chain, merchant and consumption scene data in a full-dimension manner, and laying a foundation through encrypted storage and tagging processing; logistic regression, an ARIMA model and a greedy algorithm are adopted to fuse and drive integral hierarchical appreciation, and dynamic balance of a fund pool is synchronously managed and controlled. The consumption demand is matched based on K-means clustering and a collaborative filtering algorithm, and a full-scene point consumption and grade differentiation order-free mechanism is achieved; redisk optimization is carried out through algorithms such as a decision tree to form an operation closed loop; according to the invention, point appreciation is controllable, consumption matching is accurate, member stickiness and merchant participation enthusiasm are effectively improved, and large-scale popularization requirements of Changbai Mountain specialties are met.

The invention provides an electric vehicle charging pile recommendation method based on a heterogeneous information network. The method comprises the steps that historical charging POI page view data, charging pile scoring data of a user and charging plan search trend data and event search trend data from a search engine are collected and subjected to data preprocessing; constructing a heterogeneous information network; on the basis of the heterogeneous information network and the preprocessed data, the charging page view of the charging pile is predicted through a charging prediction model, and a charging page view prediction result is generated; based on the charging page view prediction result, through a collaborative filtering algorithm based on matrix decomposition, generating prediction scores of all unused charging piles by a user; the prediction score is corrected according to the weighted score of the POI around the charging pile, and a final recommendation score is obtained; and generating a charging pile recommendation list according to the final recommendation score. The accuracy and practicability of charging facility recommendation are improved.

This invention discloses a personalized shop recommendation method based on multi-dimensional spatiotemporal perception and hierarchical fusion, belonging to the field of intelligent product recommendation. The method includes the following steps: system initialization and multi-source data integration; real-time passenger feature extraction and spatiotemporal modeling; execution of collaborative filtering layer processing to generate a preliminary shop recommendation list; execution of geofencing filtering layer processing; execution of dynamic time warping layer processing; execution of a hierarchical fusion mechanism based on adaptive weights to dynamically integrate scores from different processing layers and generate a final personalized recommendation list; real-time monitoring of passenger behavior feedback, and dynamic optimization of the weight configuration of each processing layer through an online learning strategy to achieve self-evolution of the recommendation strategy.

The application discloses a potential positive sample mining method for personalized recommendation. In view of the problem that existing collaborative filtering methods are prone to misjudging non-interaction items as negative samples, the application aims to mine potential positive samples in non-interaction items to alleviate data sparsity and improve model generalization ability. The method first disentangles item representation into independent semantic factors through a semantic factor routing module; secondly, potential positive samples in non-interaction items are identified based on semantic factors by using a semantic factor matching module; finally, the observed interaction pairs and the mined potential positive sample pairs are aligned simultaneously by using a semantic pair alignment module, and the uniformity of the distribution of user and item representations is constrained. The method effectively improves the generalization ability and scalability of the model under sparse data, significantly reduces the computational complexity, and greatly improves the accuracy of recommendation.

The invention relates to a sensitive topic propagation and control method based on subjective emotion and blocking tolerance of a user, and belongs to the technical field of information propagation control. Firstly, an information entropy theory is introduced to mine objective influence of sensitive information on user cognition, and meanwhile, a popularity algorithm method is introduced to track dynamic change of information flow in real time; measuring message influence based on user topic cognition and information traffic accuracy; a collaborative filtering algorithm and a Jaccard coefficient are introduced to quantify topic attention, interests and preferences of the users and credibility among the users respectively, subjective emotion influence factors of the users are designed, the effects of subjective emotion and message influence are considered, an evolutionary game theory is introduced to design an H hesitant state, and a multi-state propagation dynamic model based on subjective emotion is constructed; and proposing a comprehensive blocking index based on influence-tolerance to reveal the propagation situation of the sensitive topics, and effectively controlling the propagation of the sensitive topics.

The invention relates to the technical field of data processing, and discloses a modular new energy automobile motor dismounting and detection teaching method and system. The method comprises the steps that motor parameters are recognized through RFID, student ability is tested to generate an initial scoring matrix, ability portrait vectors are constructed, a personalized training sequence is generated through collaborative filtering and a Bayesian algorithm, a visual algorithm is adopted to recognize component superposition guide animation, and operation data are collected; and a rack test injection fault is controlled, a detection data update capability vector is collected, a weak knowledge point backtracking dependency relationship is identified, training sequence matching resources are adjusted, and a file is generated through cyclic execution. According to the method, self-adaptive training sequence generation based on student ability portraits, accurate operation guidance assisted by augmented reality, intelligent analysis and evaluation of multi-dimensional learning data and targeted intensified training driven by a knowledge graph are realized.

The invention relates to the technical field of logistics, and discloses a damaged express processing method, device and equipment and a storage medium, and the method comprises the steps: obtaining the damage link feature and warehousing time feature of a target problem express; inputting the damaged link features and the warehousing time features into a pre-trained logistics responsibility allocation prediction model to obtain a target responsibility allocation suggestion; distributing the target problem express item to a corresponding responsibility distribution center; displaying the target problem express based on the user role permission to obtain a to-be-processed problem express list corresponding to the user role permission; performing priority ranking on the target to-be-processed problem express items based on an intelligent ranking algorithm of collaborative filtering; obtaining core attribute features of the core to-be-processed problem express; inputting the core attribute features into a pre-trained deep learning fusion model, and outputting a target processing suggestion; according to the scheme, responsibility determination, intelligent sorting and intelligent output are carried out, so that the timeliness of problem express processing and the overall processing efficiency are improved.

The invention discloses a spatial crowdsourcing fuzzy task recommendation method and system based on differential privacy, and relates to the technical field of spatial crowdsourcing. Performing fuzzification processing by using a K-anonymity technology according to a task request of a task requester to obtain a fuzzy request and submitting the fuzzy request to a fog node; the fog node receives the fuzzy requests, aggregates the fuzzy requests by using a user-based collaborative filtering technology to obtain a group request, and sends the group request to the cloud server; the cloud server performs task recommendation on the group request by using a BERT2FTR model, and sends the group request to the fog node; and the fog node distributes the group result to the corresponding task requester. According to the method, a BERT2FTR model subjected to differential privacy training is introduced to design a fuzzy recommendation and privacy protection mechanism, fuzzy processing on task requests is realized by using a K-anonymity technology, and the task requests are aggregated by using a user-based collaborative filtering technology. And processing fuzzy task recommendation by using a BERT2FTR model to realize privacy protection.

The invention discloses a smart tourism recommendation method based on granular computing and a type-2 fuzzy set. The method comprises the following steps: acquiring multi-source heterogeneous tourism data; performing uncertainty modeling on the multi-source heterogeneous tourism data by using a Bayesian neural network, outputting posterior distribution, and mapping the posterior distribution into membership function parameters of a generalized type-2 fuzzy set to obtain multi-modal type-2 fuzzy information particles; based on a particle calculation framework, with coverage rate-specificity collaborative maximization as a target, performing optimal particle size distribution on the multi-mode type-2 fuzzy information particles to generate optimal particle size information particles; inputting the optimal granularity information particles into a two-channel deductive learning framework, realizing knowledge-data dynamic fusion by the two channels through a shared attention layer, and outputting a fused tourist preference feature vector; and based on the preference feature vectors, constructing a multi-granularity graph neural collaborative filtering model, respectively generating a personalized recommendation list, a group recommendation list and a socialized recommendation list, and outputting interpretable rules.