29 results about "Rating matrix" patented technology

The invention discloses a collaborative filtering recommendation method based on user article scores fused with multi-dimensional evaluation indexes, and the method comprises the following steps: a score data preprocessing stage: inputting a score matrix, optimizing a loss function of BiasSVD by using a stochastic gradient descent method, obtaining parameters required by a BiasSVD score calculation formula, and calculating the score of the user article scores according to the parameters required by the BiasSVD score calculation formula; outputting a complete score data table with enhanced robustness; in the multi-dimensional evaluation index establishment stage, three key indexes are calculated, and the indexes are subjected to normalization calculation to form a multi-dimensional novelty evaluation framework; in the recommendation generation stage, according to the algorithm, a post-processing weighted sorting mode is combined with a multi-dimensional novelty index and user article score data, the user similarity is calculated through cosine similarity, scores of the users on the articles are predicted through a weighted summation method, and a comprehensive recommendation list is generated, so that the problem of data sparsity of a traditional collaborative filtering algorithm is effectively solved; the problems of recommendation result homogenization, insufficient novelty and long tail effect are solved.

The deep neural network recommendation system with prediction reliability comprises the following steps: obtaining and cleaning data information from a database to obtain required information; dividing a user-item rating matrix R into corresponding several independent and only containing 0-1 binary sub-matrices according to different rating values set by the system; using a double-tower model in a deep neural network to train the obtained sub-matrices in parallel to predict the probability value of the same blank area in each sub-matrix; normalizing the obtained probability value in each sub-matrix, and taking the maximum normalized probability value as the reliability probability; finding the rating represented by the corresponding sub-matrix according to the obtained reliability probability, and taking the rating as the predicted rating of the blank area; comparing the predicted rating with the reliability threshold value set by the system, filtering the predicted rating with a reliability probability lower than the threshold value, and retaining the rating with a higher reliability probability; sorting the filtered predicted rating set, pushing the first k items with higher predicted ratings to the target user, and forming a personalized recommendation list.

The invention discloses a data poisoning attack method based on a deep learning recommendation algorithm, and the method comprises the steps: precisely screening abnormal scoring users and interaction items thereof through a statistical method in combination with a non-consistency measurement NCM and P-value, and constructing a selection item set; secondly, injecting a forged user, endowing a selected item set and a target item set with the highest score, scoring a filling item set according to normal distribution, and forming a user-item score matrix R'with poisoning data; a double-layer iterative optimization framework is adopted, and an outer layer maximizes a target project hit rate through attack loss CWLoss, enhances loss SFALoss through spectral features and hits attack traces through regularization. And the inner layer uses the user-project scoring matrix R'of the poisoning data to update the model parameters based on the deep learning recommendation algorithm until the attack converges. According to the method, the recommendation result can be remarkably manipulated under the condition of only a small number of poisoning users, high concealment and high attack efficiency are both achieved, and an effective means is provided for evaluating and improving the robustness of a recommendation system.

This invention discloses a PNN-based multi-feature extended autoencoder recommendation algorithm in the field of personalized data recommendation research, comprising the following steps: extending the auxiliary feature information of items through the open knowledge graph DBpedia; representing the extended feature information as an LsiModel vector and using PNN for feature cross-fusion to mine the potential feature relationships between different features and embed them into low-dimensional feature vectors; merging rating information, attribute information, and extended feature information into a semi-autoencoder, extracting robust feature representations through the semi-autoencoder to help the rating matrix better reconstruct the output, comparing it with the original rating matrix, calculating the prediction accuracy, and achieving more accurate recommendations. This invention utilizes knowledge graphs to extend item feature information and effectively fuses feature relationships between different features, learning higher-level feature representations through a semi-autoencoder to achieve the goal of providing more accurate recommendations for users.

The invention provides an agent-based multi-model collaborative evaluation method and related equipment, and the method comprises the steps: carrying out the semantic analysis of a natural language task request inputted by a user, decomposing an analysis result, and generating a corresponding task graph; calling a worker agent according to a subtask sequence of the task graph, obtaining candidate answers output by different language models, and generating an answer envelope set; projecting the embedded vectors of the answer envelope set to a unified semantic space, and performing weighted fusion on the embedded vectors to generate an aggregation vector; performing quantitative scoring on the answer envelope set and the aggregation vector to generate a multi-dimensional scoring matrix and candidate answer subsets; and carrying out aggregation calculation on the candidate answer subsets according to the multi-dimensional scoring matrix, determining global confidence distribution, and carrying out information supplementation on a low confidence region to generate a final candidate answer. According to the method, the reliability of the question-answering system is effectively improved by performing collaborative evaluation and unified aggregation on multi-model output.

The application relates to a privacy protection recommendation method and system based on matrix decomposition and belongs to the technical field of privacy protection recommendation. The method comprises the following steps: obtaining a historical rating matrix of a user; performing matrix decomposition on the historical rating matrix to obtain a shared rating matrix; sending the shared rating matrix and a scored item set to a recommendation server to generate public parameters, a decomposition matrix and a shared value of the decomposition matrix; generating inner product triplets and multiplication triplets by using a triplet server; calculating errors between real ratings and predicted ratings by using the recommendation server to generate an error shared matrix; optimizing the error shared matrix until the error squares of all the error shared matrices are less than a threshold value; calculating partial derivatives of a loss function on the decomposition matrix to obtain a gradient shared matrix; updating the decomposition matrix to obtain a user decomposition shared matrix; and generating a privacy protection recommendation result of the user by using the user decomposition shared matrix. The application provides strong protection in the aspect of protecting user privacy.

The invention discloses a digital marketing recommendation method and system based on big data, and relates to the technical field of commodity recommendation. The method comprises the steps of obtaining original data of an e-commerce platform, obtaining a structured data set through preprocessing, performing fuzzy clustering on a scoring matrix, generating and iteratively updating a membership matrix, reducing the influence of initial clustering deviation, calculating clustering attribution in combination with user geographic information, and correcting scores according to the clustering attribution to obtain a final scoring matrix. Inputting the matrix and the user and commodity feature vectors into a recommendation model, and outputting a commodity recommendation result; clustering multiple membership degrees of the users by adopting fuzzy clustering, and adapting to complex overlapping characteristics of user preferences under sparse data of the distributors; the initial deviation is reduced by iteratively optimizing the membership degree, attribution is determined in combination with geographic information, and the clustering fitness and distinction degree are improved; the commodity recommendation model effectively solves the problem of low accuracy of the commodity recommendation result under the condition of reducing the calculation complexity.

This invention discloses a graph mining-based questionnaire analysis method, system, electronic device, and medium, relating to the field of data processing. The method includes: acquiring questionnaire data to be analyzed; establishing a questionnaire matrix based on the questionnaire data; cleaning the questionnaire matrix to obtain a scoring matrix; calculating the similarity between different attribute scores based on the scoring matrix to establish a core attribute adjacency matrix; using attributes in the core attribute adjacency matrix as nodes in the network to establish a core attribute network; performing graph mining on the core attribute network to obtain a minimal core attribute network; and analyzing the influence of different core attributes based on the minimal core attribute network to obtain the questionnaire analysis results. This invention can reduce the influence of extreme data in questionnaires and improve the interpretability of the data in questionnaires.

The application discloses a fair personalized recommendation method and device based on mutual information decoupling and a storage medium, and steps of the method comprise the following steps: 1. constructing original data, including a user-product rating matrix and a user sensitive attribute matrix; 2. constructing a sensitive embedding network to learn sensitive embedding of a user and a product, including a biased one-hot encoding layer, a sensitive information encoder and a sensitive attribute prediction layer; 3. constructing a hybrid embedding network to learn hybrid embedding of the user and the product, including a hybrid one-hot encoding layer, a hybrid information encoder and a preference prediction layer; and 4. constructing a non-sensitive embedding network to learn non-sensitive embedding of the user and the product, including an unbiased one-hot encoding layer, a mutual information lower bound optimization layer and a mutual information upper bound optimization layer. Through the double mutual information fairness constraints on the embedding vectors, the application improves the fairness of the recommendation system while ensuring the recommendation accuracy.

The application provides a multi-field content recommendation method and system with novelty fusion, and the application optimizes a user-item rating matrix based on a BiasSVD model, adds a novelty factor to predict the user rating matrix, and alleviates problems such as unsatisfactory recommendation effect caused by data sparseness in a traditional model; similarity is calculated based on user feature information, cold start problems caused by excessive dependence of a traditional recommendation algorithm on historical data are solved, the recommendation result is closer to the real rating, and the recommendation accuracy is improved; the novelty relationship between users is introduced, and the novelty of the recommendation result of a traditional collaborative filtering recommendation algorithm is improved. The method disclosed in the application can alleviate the problem of sparse rating data, and can improve the novelty of the recommendation result, so that the recommendation result is not too single, thereby avoiding user aesthetic fatigue.

This invention proposes a rating-based reliable recommendation system, which includes: acquiring and cleaning data from a database to obtain the required information; classifying users and items with different preference levels according to a noise identification rule based on user preference consistency, and identifying noise in the original rating matrix; generating an initial rating reliability matrix according to a designed rating reliability matrix generation module; filling non-interactive areas in the rating reliability matrix according to a designed positive sample filling rule to balance the data; using a generative adversarial network to train the acquired rating matrix and reliability matrix to predict the rating value and reliability probability of each blank area; filtering predicted ratings with reliability probabilities below the system threshold by comparing them with a reliability threshold, and retaining ratings with higher reliability probabilities; sorting the filtered predicted rating set, and pushing the top k items with higher predicted ratings to the target user to form their personalized recommendation list.

The application discloses an online course recommendation method and device based on class concentration detection, which first collects basic information data of students, vectorizes the basic data of the students, and generates an initial recommendation list by using a user-based collaborative filtering method; then, concentration detection is performed on video records of the students during class, the proportion of the concentration state is calculated, and the proportion is taken as a score of the students on the course; finally, the existing score is converted into a matrix form, a recommendation result is obtained by using a matrix decomposition-based collaborative filtering method, and the recommendation result is fused with the initial recommendation result to obtain a final course recommendation result. The application combines the concentration of the students during class, can better feedback the interest degree of the students on the course, and obtains a more accurate student-course score; and the application can improve the sparsity of the student-course score matrix, and has a better effect when the matrix decomposition-based collaborative filtering method is applied.

The picture fuzzy set-based collaborative filtering recommendation model comprises the following steps: obtaining and cleaning data information from a database to obtain an original rating matrix; based on picture fuzzy set theory, converting the rating matrix into four matrices representing different preference degrees of users, namely, a membership matrix, a neutral matrix, a non-membership matrix and a rejection matrix; fitting the preference matrix by using a BeMF model to learn feature vector information of the users and the items, and obtaining picture fuzzy numbers of the users for the rated items through inner product of the feature vectors; scoring the picture fuzzy numbers by using a scoring function; learning potential global user preferences in historical ratings by using a multilayer perceptron; combining the scoring of the picture fuzzy numbers (fine-grained preference) and the output value of the multilayer perceptron (global preference) to form a comprehensive prediction value and sort the comprehensive prediction value to generate a recommended item list.

The application discloses a time-aware adaptive interest point recommendation method based on K-means clustering, which comprises the following steps: first, collecting and sorting check-in data sets, and converting to generate a user-time-location three-dimensional score matrix; second, extracting a two-dimensional check-in score matrix in each time slot, and generating a one-dimensional score vector of each time slot by using a data compression technology; based on the one-dimensional score vector, the K-means method is used to cluster the time slot; third, calculating the dynamic similarity of users in each time slot; based on the time clustering, the score method of the traditional user-based collaborative filtering algorithm is improved, so that the interest point prediction score can be adaptively generated according to the current recommendation time; a plurality of unvisited addresses ranking at the front at the current time are recommended to the user; fourth, the recommendation quality is evaluated by using a recommendation precision index, and the accuracy and effectiveness of the proposed technology are evaluated by comparing the recommendation precision of the technology proposed by the application with that of other classical recommendation systems.

The invention discloses a commodity collaborative filtering recommendation method and system based on user behavior analysis, and relates to the field of commodity recommendation, and the method comprises the steps: classifying commodities, and building life cycle files with differentiated attenuation parameters for different categories; when recommendation calculation is carried out, a dynamic score with timeliness is calculated for each user-article interaction according to an interval between behavior occurrence time and current time in combination with a category life cycle file, and a dynamic score matrix is formed. And carrying out subsequent article similarity calculation and recommendation list generation based on the matrix. In this way, attenuation speeds and modes of different commodity interests can be effectively distinguished, the recommendation result can more accurately capture the current core demand of the user, and the problem that recommendation is inaccurate due to the fact that user interest dynamics, periodicity and category differences cannot be processed in the background technology is solved.

The application discloses a recommendation method based on an interpretable generalized logistic transformation matrix decomposition, and comprises the following steps: converting an original score matrix into normally distributed data through a generalized logistic transformation function; constructing a similarity-based index and a ranking-based index; calculating a probability distribution of a similar user's score on a recommended item and an expected score, and combining the similarity index to generate an interpretability index; integrating the interpretability index into a matrix decomposition target function for optimization; solving a user feature matrix and an item feature matrix through an alternating least squares method or a stochastic gradient descent; calculating a predicted score and mapping back to an original score interval through a generalized logistic inverse transformation; and generating a recommendation list; the application has wider applicability and higher performance in practical application.

The application provides a popularity debiased recommendation model based on item similarity, which comprises the following steps: extracting user item rating data from a database and preprocessing the data to obtain an original rating matrix; calculating the similarity between items based on KL divergence, and calculating the popularity based on item similarity (ISP) based on the similarity; based on the causal theory, decoupling the interaction score of the user and the item into two parts of "real interest" and "herd behavior", and assigning corresponding embedding vectors to the user and the item; by introducing a weight attention mechanism, adaptively learning the weight of the user's interest and herd behavior; based on the obtained weight, weighted sum of the two parts of the score, calculate the final personalized recommendation score; by using the negative sampling method based on ISP, the training data set is divided, and the loss function is constructed to supervise the learning of the decoupled embedding vector; predicting and ranking the score of the item that the user has not rated, and selecting the top k items to form a personalized recommendation list.

The application particularly relates to a post-fusion personalized recommendation model and method based on explicit and implicit feedback features, the model comprising an explicit feature extraction module, an implicit feature extraction module and an overall feature extraction module, the explicit feature extraction module and the implicit feature extraction module being connected with the overall feature extraction module respectively; the method fuses an IBPR model and a BiasSVD model, performs weighted summation on a predicted score matrix obtained by the BiasSVD model and a predicted ranking score matrix obtained by the IBPR model, obtains a final predicted ranking score matrix, ranks all predicted ranking scores of users from high to low, and recommends the first items with higher ranking to the users. The application extracts explicit feedback features by using the BiasSVD model, extracts implicit feedback features by using the IBPR model, and fully utilizes historical score data and implicit feedback data in a data set, thereby relieving a cold start problem of a recommendation system and improving the performance of the recommendation system.

The application relates to a recommendation system attack detection sample data generation method and device. The method comprises the following steps: obtaining a real user score matrix in a data set of a recommendation system; generating a false user score matrix corresponding to the real user score matrix through a generator according to the real user score matrix; mixing the false user score matrix and the real user score matrix to obtain a mixed user score matrix; simulating attacks on a plurality of recommended items through a simulated recommendation system attack discriminator according to the mixed user score matrix to obtain an attack loss; and generating an attack detection sample based on the attack loss and the mixed user score matrix. The method can support accurate detection of the attack resistance of a recommendation system.

The application discloses a recommendation algorithm based on serial auto-encoders, comprising 1) merging item-based rating information and user and item interaction auxiliary information into an auto-encoder for reconstruction output, obtaining a feature representation of the reconstruction output through traditional auto-encoder learning, using auxiliary information to help the original rating matrix to reconstruct, and reducing the loss of effective information; 2) designing a serial connection method of auto-encoders, obtaining the reconstruction output generated by the first auto-encoder, inputting the reconstruction part of the output to the original rating matrix into the second auto-encoder, comparing the output of the second auto-encoder, i.e., a predicted rating matrix, with the original rating matrix, and calculating the prediction accuracy. The application can utilize attribute information of items, process the attribute information through auto-encoders, use the attribute information as extended features of recommendation, and achieve the purpose of more accurate recommendation for users.

The invention relates to a cross-domain recommendation method and device, equipment, a medium and a product, and the method comprises the steps: obtaining first scoring data of a user in a source domain and second scoring data of the user in a target domain, and constructing a joint scoring matrix based on the first scoring data and the second scoring data; complementing missing scores in the joint scoring matrix through an encoder to obtain a complemented scoring matrix; based on the completed scoring matrix and the joint scoring matrix, generating predicted scores of the user for the plurality of to-be-recommended items in the target domain; and determining a target item based on the prediction score, and recommending the target item to the user. According to the method, the missing score is automatically complemented and the cross-domain information is fused, so that the common problems of data sparsity and new user cold start in cross-domain recommendation can be effectively relieved, and the accuracy, robustness and scene adaptability of a recommendation system are remarkably improved.

The invention belongs to the technical field of recommendation systems, and is used for solving the problems of data sparsity, cold start and insufficient recommendation precision in a traditional recommendation system. According to the system, a multi-dimensional weighted scoring model is constructed by fusing user dominant scores, comment sentiment analysis and time dynamic factors, and the accuracy of personalized recommendation is improved. According to the scheme, a BERT model is used for performing emotion classification on user comments, extracting recessive preferences and converting the recessive preferences into numerical emotion scores, then a time decay mechanism is introduced, weights are dynamically adjusted according to time intervals of user behaviors, short-term interests and long-term interests are distinguished, and finally, dominant scores, emotion scores and dynamic time weights are synthesized, so that the user comments are classified. Weighted fusion is carried out through experiment optimized weight coefficients, a comprehensive scoring matrix is generated, unscored item preferences are predicted according to the comprehensive scoring matrix, and a personalized recommendation list is generated. According to the method, through multi-source data fusion and dynamic interest modeling, the timeliness and individuation level of recommendation are remarkably improved, the cold start problem is effectively relieved, and the user satisfaction is enhanced.

The invention provides an auto-encoder and diffusion model collaborative optimization recommendation system based on input influence recognition. The system comprises the following steps: acquiring and cleaning user-project interaction data from a database to generate an original score matrix; calculating a user score mean value and a project score mean value, and performing linear reconstruction on the score matrix based on the weighting parameters to generate a reconstruction score matrix composed of a personalized matrix and a global offset matrix; performing feature learning on the personalized matrix by using an auto-encoder model to generate a personalized prediction matrix representing personalized score differences; performing noise disturbance learning and reverse denoising reconstruction on the global bias matrix by using a diffusion model to generate a global bias prediction matrix representing an overall scoring trend; fusing the personalized prediction matrix with the global offset prediction matrix to obtain a user-project score prediction matrix; and ranking the predicted values of the items which are not scored by the user, and pushing the first k items with relatively high predicted scores to the target user to form a personalized recommendation list of the target user.

The invention relates to a recommendation system fusing fuzzy neighborhood and mixed negative sampling technology, and belongs to the technical field of recommendation. The method comprises the steps of obtaining and cleaning data information from a database; designing a similarity measurement method based on fuzzy preference consistency, and describing a preference relationship between users or items from multiple dimensions of consistency, inconsistency and uncertainty; obtaining an interaction matrix C of the users and the items by using the original scoring matrix R of the users and the items, and performing similarity calculation between the users and the items to obtain nearest reciprocal neighbors of the users and farthest reciprocal neighbors of the items; constructing a neighborhood information guided double-tower generation method, enhancing generation process supervision by aggregating neighborhood semantics, and realizing one-step denoising of a diffusion method by using a generative adversarial network; designing a mixed negative sampling strategy combining strong and weak negative samples, and introducing negative samples with rich information to assist denoising training; calculating a predicted value of the target user u on an unscored item by using a score prediction method; and sorting the prediction sets, and selecting the first k items with the highest prediction values to recommend to the target user u.

An explainable recommendation method based on sentiment analysis, comprising: 1) constructing an actual user-item rating matrix; 2) obtaining the sentiment score of the user to the item from the comment by using a sentiment analysis method; 3) constructing a user-feature attention matrix and an item-feature quality matrix according to the feature word set extracted in step 2); 4) constructing a matrix decomposition model; 5) training the model in step 4); and 6) personalized recommendation. The present application improves the efficiency, further solves the problems of data sparsity and the possible unexplainability of the recommendation results, and realizes explainable recommendation.

The application discloses a commodity collaborative filtering recommendation method and system based on user behavior analysis, and relates to the field of commodity recommendation. First, commodities are classified, and life cycle archives with differentiated decay parameters are established for different categories. When performing recommendation calculation, the life cycle archives of the categories are combined, a dynamic and time-sensitive score is calculated for each user-item interaction according to the interval between the time of the behavior and the current time, and a dynamic score matrix is formed. Based on the matrix, subsequent item similarity calculation and recommendation list generation are performed. In this way, the decay speed and mode of different commodity interests can be effectively distinguished, the recommendation result can more accurately capture the current core needs of the user, and the problem of inaccurate recommendation caused by the inability to handle the dynamicity, periodicity and category difference of user interest in the background technology is solved.

The application relates to a multi-attribute neural collaborative recommendation system fusing project uncertainty, and belongs to the technical field of recommendation. Data information is acquired and processed from a database to obtain an initial rating matrix; based on information entropy theory, an uncertainty matrix is acquired according to the rating distribution of different users to the same project; user ID and un-rated project ID are obtained from the database and used as the input of a neural network to obtain multi-attribute predicted ratings of the user to the project; the uncertainty is introduced into the neural network framework and used as a weight factor to correct the multi-attribute predicted ratings; the corrected multi-attribute predicted ratings are used as the input of the neural network to perform result prediction and obtain the comprehensive ratings of the user to the project; the projects are sorted in descending order according to the predicted ratings; the number of recommended projects is set, the projects meeting the recommendation conditions are pushed to the target user, and a personalized recommendation list is formed.

The application discloses a kind of multimedia-oriented generative generalization cold start recommendation method, comprising:1. the interactive record of user and product is used to construct score matrix;2. input layer is constructed by one-hot encoding mode, and user, product is mapped to different embedding space by combining the multimedia features of product;3. the embedding of user, product is optimized by bayesian ranking loss function;4. construct generative neural network, including: prior neural network, encoder, decoder;5. construct uniformity enhanced conditional variational autoencoder, to make the latent space of original variational autoencoder more uniform and distinguishable;6. new product embedding generation method based on clustering obtains new product generative embedding;7. the degree of love of user to new product is predicted by the way of vector dot product.The application can make recommendations for new products using multimedia information of products, thereby alleviating the cold start problem in the recommendation system.

The application discloses a kind of large-scale user recommendation methods based on evolution multitask, comprising: step 1, obtaining user and article interaction dataset, step 2, by constructing neural network algorithm, the score of the preference of user to article is excavated, and the rating matrix of user to article is obtained;Step 3: by clustering, group users, and similar user of preference interest is as the same category;Step 4, initialization generates multi-task population;Step 5, information migration between individuals in the same user group, information migration between populations in different user groups, and the optimal user solution is selected by environmental selection iteration, and finally the optimal solution is selected as the recommendation list of the user.The application can reduce the time and space consumed in large-scale user recommendation optimization problem, and improve the accuracy of user recommendation result prediction by clustering technology.