129results about How to "Improve recommendation quality" patented technology

The invention discloses a multi-strategy commodity recommendation system basing on context information. The recommendation system acquires the operation information of a user through an information acquisition part which is operated by the user, analyzes the operation action of the user and establishes the interest description model of the user. During the interaction process between the user and an electronic commerce website, a recommendation strategy fitting the present user and the context information of the system is dynamically selected according to a strategy selection rule. The recommendation strategy describes and generates a personalized commodity recommendation list according with the interest and the requirement of the user according to the interest of the user. Through the selection of the recommendation strategy, the multi-strategy commodity recommendation system basing on context information improves the adaptability of the system to various applications and system dynamic changes. And compared with the existing recommendation system, the multi-strategy commodity recommendation system basing on context information is improved in the recommendation quality, the recommendation scale and the recommendation performance.

The invention discloses a television program recommending method and system. The method comprises the following steps of structuring a television program management cloud and a user management cloud, and determining an associated relationship between a television program resource and a user watching behavior way; obtaining a recommended result through adopting a program clustering and user clustering bi-clustering collaborative filtering algorithm based on the television program management cloud and the user management cloud; and performing television program recommendations on a target user according to the recommended result. The method and the system provided by the invention can be used for providing in-depth services for users, realizing the characteristics of university and reliability, and ensuring the feasibility and the reliability of the platform through selecting real data.

The invention discloses a recommendation algorithm based on multi-index grading. The recommendation algorithm comprises the following steps of firstly, recognizing index keywords, secondly, extracting suggestion grading, thirdly, constructing a user and commodity similarity matrix, fourthly, using a two-way clustering algorithm for obtaining a clustering matrix, fifthly, conducting single in-cluster recommendation and sixthly using a comprehensive function algorithm for obtaining a final recommendation result. According to the recommendation algorithm, the problem that a user may need individual recommendations for different index preferences for different commodities can be solved, the high accuracy is achieved, and the recommendation result with the higher quality can be obtained.

The invention discloses collaborative filtering recommending method and system based on client characteristics. The method comprises the following steps of: obtaining user information and carrying out information expression, neighbor formation and recommendation generation. Based on a most adjacent collaborative filtering algorithm, the method comprises algorithm name, algorithm input and algorithm output, and also comprises algorithm analysis. The invention also discloses a collaborative filtering recommending system based on client characteristic forecasting, which comprises data loading, recommendation engine as well as client mail notification and recommendation daemon process.

The invention relates to a news recommendation method and system, and relates to the technical field of information. The method and system aim to push news to different types of users who are interested in the news. The method includes the steps of extracting features of search and query data, according to behaviors of a certain type of users on the search and query data, calculating and recording interest weights of the features of the search and query data on the type of users, extracting features of multiple pieces of news to be pushed, searching for the interest weights of the features of the news to be pushed from the recorded features and the interest weights, calculating interest scores of the news to be pushed for the type of users, and pushing the news to the type of users according to the interest scores of the news to be pushed for the type of users. By means of the method and system, interests of different types of users can be analyzed, and the news according with the interests of the users is pushed to the users.

The invention relates to a collaborative filtering method for personalized recommendation fusion content and behavior. The method comprises the following steps of, (1) characteristic input, includinga project-attribute matrix representing a project content and a user behavior matrix representing user behaviors; (2) content-based project clustering for calculating the similarity of projects and clustering the projects; (3) score prediction and feature filling including carrying out score prediction on the non-scoring projects, and filling a user-project scoring matrix; (4) behavior-based userclustering including clustering users according to a project clustering result and a user-project scoring matrix; (5) score predication and project recommendation including determining the clusteringcluster where the target users are located, finding a nearest neighbor user set, performing score prediction on the non-scoring projects of the target users, and finally recommending the first N projects with the highest prediction scores to the target users. Compared with the prior art, the collaborative filtering method effectively solves the problems of data sparsity and cold start, and ensureshigh recommendation efficiency.

Techniques and systems are described that enable the intelligent recommendation and presentation of demand-responsive transit services. Implementations can analyze current user activities with respect to historically-based user transit behaviors in order to determine the user's imminent transit demands. Prior-scheduled transit options meeting the imminent demand are searched, and filtering parameters including user preferences and demand-responsive transit provider constraints are retrieved. Embodiments further retrieve possible demand-responsive transit options for meeting the demand from a demand-responsive transit system scheduling / information service. Embodiments consolidate and analyze these diverse sources to identify a set of preferred transit alternatives for available transit options that meet the filtering parameters. Implementations are described that integrate with an existing transit application present on a user device (e.g., a mobile device). Certain embodiments may include capabilities for surfacing an enhanced transit options view that shows an identified imminent transit demand and a selection of preferred transit alternatives.

The invention discloses a partial model weight fusion Top-N film recommending method based on user clustering. The method comprises the steps of 1, preprocessing data, wherein inactive users and filmswith very low popularity are subjected to data cleaning, user film label documents are constructed, explicit scoring information is converted into implicit feedback information, and a user-film implicit feedback matrix A is constructed; 2, conducting user clustering, wherein film label information is utilized, user feature vectors are obtained by training an LDA topic model, and user clustering is achieved through a spectral clustering algorithm; 3, determining a local recommending model and training a global recommending model; 4, conducting model weight fusion recommending; 5, proving the effectiveness of the models through a leave-one-out method.

The invention discloses a Top-N movie recommendation method for performing weighted fusion on selected local models based on random anchor points. The method comprises the steps of obtaining eigenvectors of users and movies at the semantic level through an LDA topic model and a GBDT by utilizing movie text data; based on the eigenvectors, calculating Gaussian kernel similarity between the users and the movies; randomly selecting a plurality of (the users and the movies) anchor point pairs, and reconstructing a local training matrix for each anchor point pair in combination with the Gaussian kernel similarity between the users and the movies; performing training for each local training matrix by utilizing an SLIM as a basic recommendation model to obtain a corresponding local recommendationmodel; and finally, generating a final fusion recommendation model through weighted fusion among the local recommendation models. According to the recommendation method, the stability of the models is also maintained in a data sparseness scene, and the problem that a traditional single recommendation model is very easy to over-fit in the data sparseness scene can be effectively solved.

The invention belongs to the technical field of personalized recommendation, and particularly relates to a method and system for solving a cold start problem in a collaborative filtering technology. The method for solving the cold start problem in the collaborative filtering technology comprises the steps that a data set is selected; an initial user or project clustering model is built through an optimized genetic algorithm; clustering is conducted on the basis of the initial user or project clustering model, and a user or project clustering model is obtained; entropy values of new users or new projects to all kinds of clusters in the clustering model are calculated, and the new users or the new projects are subjected to class cluster dividing; the new users or the new projects are recommended. The invention further provides a system for solving the cold start problem in the collaborative filtering technology. The system comprises a selection module, an initial model building module, a clustering module, a class cluster dividing module and a recommendation generation module. Accordingly, an improved genetic algorithm is utilized for conducting K-Means clustering, the initial user or project clustering model is generated, and recommendation is generated for the new users or the new projects.

The invention provides a project recommendation method and device. According to recorded grading objects and scores thereof to projects, the score mean value of the grading objects to the projects and the types of all projects are determined firstly; then the score mean value of the grading objects to the project types is counted; first related objects with similar type preferences are determined from target objects; then second related objects with similar project preferences are found out from the first related objects; the eligible projects are extracted from the projects which are not graded by the target objects and recommended to the target objects. According to the embodiment of the invention, users with related preferred projects and types are selected out and invalid information not interested by the users are removed by using the interest degrees of the users to different types of projects; all effective information is used when the similarity among the users is calculated, so that the similarity calculation is more accurate, the accuracy of calculating the similarity among the users can be effectively enhanced, and the recommendation quality of a recommendation system is effectively improved.

The invention belongs to the field of personalized intelligent recommendation, a collaborative filtering recommendation method and system for learning resources based on knowledge association are disclosed. By combining with the relationship between knowledge and the relationship between resources and knowledge points, the information of knowledge points associated with learners is introduced intothe methods of user similarity calculation and interest calculation, and then the nearest neighbor user group of target learners is obtained, and learner-Learning resource interest score matrix is constructed. Then, the interest score of the resources that the current users are not yet learning and may be interested in is predicted by the knowledge points preference of the similar user groups. Finally, N results with higher interest score are recommended to the current learners. The invention designs a collaborative filtering recommendation algorithm of learning resources based on knowledge association according to the relationship among learners, knowledge points and learning resources, so that the recommendation result is more in line with the actual learning needs of learners.

The invention discloses a personalized film recommendation method and system based on feature augmentation. The necessity and feasibility that the personalized recommendation system is transferred to a cloud calculation platform are discoursed by analyzing the problems brought by mass data to the existing personalized recommendation system, and by combining a mixed recommendation model adopting feature augmentation, the method comprises the steps that an original extremely-sparse user item rating matrix is filled for the first time through a collaborative filtering algorithm based on items to generate a pseudo two-dimensional table, then the pseudo two-dimensional table is further filled through a collaborative filtering algorithm based on the user, and lastly the defects of strict object attribute matching are avoided through conversion of quantitative knowledge and qualitative knowledge of a cloud module. According to the method, all rating data of the user is fully utilized, the calculation efficiency is greatly improved and calculation speed is greatly increased through parallel calculation of cloud calculation, therefore, the better experience is provided for the user, and the advantages of an enterprise in competition are achieved.

The invention discloses a live broadcast room recommendation method and system for a live broadcast website, and relates to the technical field of Internet video live broadcast. The method comprises the steps: generating the recommended room data of a user according to the identity of the user and historical watching information when the user enters a live broadcast room; carrying out traversal in the generated recommended room data, and finding live rooms in the recommended room data; judging whether the number of found live rooms reaches a recommended number or not: enabling the found live rooms to serve as the recommended rooms if the number of found live rooms reaches the recommended number, and displaying the recommended rooms for the user; or else, compensating for the live rooms according to a compensation rule, enabling all compensated rooms to serve as the recommended rooms, and displaying the recommended rooms for the user, wherein the number of the live rooms does not reach the commended number. The method can carry out targeted recommendation according to the watching habit of the user and the personal preferences, is higher in correlation with the user, is higher in individuation, and is good in user experience.

The invention discloses a coauthor recommending method under a scientific and technical literature heterogeneous network. Probability that a pair of authors establish a cooperation relationship in the future is in direct proportion to willingness that two nodes establish the cooperation relationship with each other, so that the coauthor recommending based on cooperator willingness is provided, and cooperating authors are recommended to authors by calculating probability that two authors cooperate in the future. The coauthor recommending method includes defining attention degree of the nodes according to future influence increase degree and rate of the nodes and ages of the nodes; designing cooperation willingness of two authors due to different relationships on the basis of the attention degree; designing topological feature attributes under different relationships on the basis of the willingness of different relationships; taking the topological feature attributes as an independent variable of a logic regression model, utilizing parameters of a real data training model, and using an acquired function model to calculate probability that the two authors cooperate in the future. The coauthor recommending method takes willingness of two authors which cooperate into consideration, and the willingness is based on node influence and age, so that prediction results of cooperation relationships are improved, and recommending quality of coauthors of authors is improved.

The invention discloses a remote sensing image recommendation method based on multi-attribute fusion, and belongs to the technical field of intelligent recommendation and distribution of remote sensing images. The method comprises the following steps: firstly, carrying out multi-dimensional knowledge graph modeling on responsibility information and ordering information of a remote sensing user andattribute information of a remote sensing image, and respectively carrying out embedded representation on the knowledge graph to obtain a user responsibility embedding vector, a user ordering embedding vector, an image order embedding vector and an image attribute embedding vector; performing semantic information extraction such as target detection and surface feature classification on the remotesensing image to obtain an image surface feature proportion vector to solve a problem of an information representation method of a user and an image in a recommendation process in the remote sensingfield effectively; and then, with a collaborative filtering recommendation network based on multi-attribute fusion, realizing fusion of multi-dimensional input by means of high-dimensional space modeling capability of a neural network, and embedding various input vectors in different spaces into a unified vector space to realize a better pairing recommendation effect.

The present invention relates to the field of interactive network television. Provided are a method and device for service recommendation. The method comprises: acquiring an earlier program that a client recently played in a current period; selecting a recommended program from a program cluster where the earlier program is located, and pushing information of the recommended program to the client, where the program clusters are acquired by division of existing programs by utilizing static characteristics of the programs; and / or pushing an advertisement to the client, where the advertisement is relevant to at least one program in the program cluster where the earlier program is located. The advertisement and program recommended per the solution of the present invention are provided with relevance and similarity to the program that the client played, and thus are of increased rationality; furthermore, the configuration of the program clusters allows a search for to-be-recommended advertisements and programs to be of increased purposiveness, thus increasing recommendation speed.

The invention discloses a mobile Web service recommendation method and a collaborative recommendation system based on user behavior analysis. The method comprises the steps of first constructing a Web service prediction model, and searching two optimal associated users who are the most similar to the a target user in long-term habit and short-term mood through intelligent terminal side user browsing data and wearable device side user physiological data. Therefore, data of the two optimal associated users are used to enrich a sample library of the target user prediction model, so as to achieve the introduction of a sample enrichment mechanism with the minimized noise. A Web resource storage mechanism based on a feature vector is designed, and according to a model prediction result, accurate recommendation of the mobile Web service is realized. According to the scheme of the invention, through the interactive cooperation between an intelligent terminal and a wearable device, application reconfiguration is performed on the wearable device side physiological data, and a user behavior is analyzed from multiple angles of view and multiple dimensions, so that accurate prediction and recommendation of services are realized, thereby improving user experience.

The invention discloses a method of recommending hit songs and singers in a music on-demand network. Considered are the fact that future heat of each node of the music network needs to meet the change law of historical heat of the node and the fact that the heats of different types of nodes are kept in certain relations; future heats of songs and singers are acquired through a function model prediction phase and a coordinated adjusting phase, and on such basis, the top songs and singers are recommended; in the function model prediction phase, a function model of historical heat time sequences of the nodes is designed and used for primarily predicting the future heats of the nodes; in the coordinated adjusting phase, average edge betweenness of different relational edges of the music on-demand network is counted, and adjusting factors are calculated to adjust the primary future heats obtained in the function model prediction phase. The method has the advantages that considered are both the historical heat information of the nodes and the heat relation of the different types of nodes, ranking prediction results of the future heats of the nodes is improved, and singers and songs are recommended with better quality.

The invention provides an information processing method and a device thereof. The method comprises the following steps: acquiring a project score dataset of users, wherein the project score dataset comprises scores on the projects of users; acquiring a first relevance weight Wi between each node and a user node to be recommended in a user-project score database, wherein the first relevance weight serves as a relevance weight in the same type of projects; calculating a second relevance weight We between each node and the user node to be recommended in the user-project score database, wherein the second relevance weight serves as a relevance weight between different types of projects; acquiring the recommendation degree of each node according to We and Wi; and determining projects to be recommended according to the recommendation degree. By virtue of the information processing method and the device thereof, Wi is firstly calculated by using a random walk mode based on the characteristics of a bipartite graph, and the users are implicitly classified; secondly, the relevance between the users to be recommended and the neighborhood users is directly calculated and serves as We in order to avoid the repeated calculation of the relevance between the nodes, so that the potential preferable project nodes of the users are fully exploited, and the recommendation quality is improved.

The invention discloses a collaborative filtering processor. A method includes the following steps: adopting an improved K-means algorithm to perform clustering on users; selecting a cluster formed by the users similar to a target user in attribute; acquiring a nearest neighbour set of the target user; generating a recommendation set. The main idea includes that the users identical in feature are clustered into a class through clustering to lower dimensionality of a matrix and reduce spatial complexity, sparse matrices are filled through matrix decomposition, and then collaborative filtering is performed on the filled matrices.

A similarity measuring method improved through a collaborative filtering recommendation algorithm includes the following steps of (S1) building a rating matrix R(n*m) of n users in a user set U={U1, U2,..., Un} to m items in an item set I={I1, I2,..., Im}, taking Ra,i as representation of rating of an item Ii, wherein Ua belongs to U and Ii belongs to I, (S2) calculating the similarity sim(Ua, Ub) between a user Ua and a user Ub and the similarity sim(Ii, Ij) between an item Ii and an item Ij, defining a similarity influence divisor epsilon, so that sim'(Ua, Ub) equals to epsilon* sim(Ua, Ub) and sim'(Ii, Ij) equals to epsilon* sim'(Ii, Ij), (S3) taking a parameter lambada in an interval between 0 and 1, and predicting rating of the users to the items according to lambada, epsilon, an average rating value of the users to the items, similarity between the users and similarity between the items.

The application relates to a method and system for recommending contents based on a social network, and a method and system for recommending news. The method for recommending contents based on a social network includes: extracting features of social network data; calculating and recording interest weights of the features of the social network data for a type of user according to a behavior of the type of the user on the social network data; extracting features of a plurality of contents to be pushed; finding interest weights of the features of the plurality of contents to be pushed from the recorded features and the interest weights, and calculating interest scores of the plurality of contents to be pushed for the type of the user; and pushing contents to the type of the user according to the interest scores of the plurality of contents to be pushed for the type of the user. According to the application, interests of users of different types can be analyzed, and the contents matching an interest of a user are pushed to the user.

The invention discloses a cold-start recommendation method based on user preferences and trust. The method comprises the steps of S1, measuring comprehensive trust values between users according to social information of the users, and constructing a trust relation matrix; S2, calculating preference similarity degrees of the users according to user scoring data, and constructing a preference relation matrix; S3, utilizing a calculation method of comprehensive similarity degrees to fuse preference relations and trust relations, and using a bee colony algorithm to iteratively update weights in the comprehensive similarity degrees, carrying out multi-objective optimization to enable the weights to become optimal in a self-adaptive manner, and constructing a preference trust relation matrix; S4, selecting a most-trusted neighbour set of the target user to predict scoring values of corresponding items for the target user on the basis of the preference trust relation matrix; and S5, recommending the items with high prediction scores to the target user. According to the method, the precision of user trust measuring is improved, the user behavior preferences are more accurately constructed, and the quality of recommendation for the cold-start user is improved.

The invention discloses a deep learning method fusing scoring data and comment texts. The influence of scoring data sparsity is relieved by introducing auxiliary information. The preference information and the item characteristics of the user can be obtained by utilizing the comment text, and the score data also contains potential association between the user and the item. Most of existing fusionmodels adopt a matrix decomposition method to process score data, in order to better utilize effective information in the score data, a convolutional neural network is utilized to process comment texts, an attention mechanism is introduced to extract representative comments in comment information, and therefore, the score data fusion accuracy is improved. Therefore, user preferences and project characteristics can be better represented. By processing the score data by using a deep neural network to extract deep features in the score data and fusing the features, the method can predict the score of the project by the user.

The invention discloses a tourist destination recommendation method based on user emotion and time dynamic. The method includes the following steps: firstly, obtaining a quantitative value of emotional inclination of a user by using opinion mining techniques, integrating emotion degree into a matrix decomposition model, and expressing user preferences and tourist destination popularity with the change of time by using a time dynamic mechanism; and then constructing a synthetic recommendation model by combining two elements of user emotion and time influence with an SVD++ method. The user preferences are automatically learned by analyzing explicit or implicit feedback, and so features of tourism leisure resources can match user needs. Compared with a plurality of famous recommendation method, the method of the invention is improved in recommendation accuracy and quality, and a series of experimental evaluations are carried out on publicly available datasets to show that a provided recommendation system is better than a conventional recommendation system.

The invention discloses a personalized community recommendation method based on user behaviors, and relates to social networks. Social network micro-blogs are used as a platform to analyze multi-attribute information of static attributes and dynamic attributes of users. Firstly, two aspects of bloggers followed by the micro-blog users and communities which the micro-blog users participate in are considered in a process of calculating user similarity degrees, and a traditional Jaccard similarity degree calculation method is extended to obtain a user similarity set; and then the similarity set is further screened from a perspective of user influences. The influences of the micro-blog users in the communities are related to numbers of fans thereof, and also to numbers of comments and forwarding on the micro-blogs thereof. On the basis thereof, a traditional PageRank algorithm is improved to calculate the user influences; and finally, Top-N is utilized to sort influence sizes to obtain final recommendation object sets. Experiment proves that an algorithm of the invention effectively solves the problem of inaccuracy of results obtained by traditional personalized recommendation algorithms, and greatly improves a surprise degree and novelty of recommendation.

The invention discloses an asymmetrical weighing similarity based collaborative filtering recommendation method and system. The method includes: determining a user similarity asymmetrical weighing factor according to the proportion of user common scoring items; calculating the similarity among users having the common scoring items through the cosine similarity measurement method and the mean square error measurement method according to the user similarity asymmetrical weighing factor; performing fitting calculation on an original similarity matrix of a user through the matrix decomposition gradient descent method to obtain users who do not have the common scoring items; selecting K users having the maximum similarity as neighbors of a target user according to the similarity among the users having the common scoring items or not having the common scoring items, and predicting scores of user for items which have not being scored according to data of the k neighbors; and generating recommendation items of the target user. The method is accurate in calculation of user similarity and is high in recommendation quality, and can be widely applied to the technical field of recommendation.

The invention discloses a multi-relationship collaborative filtering recommendation method based on a dynamic graph attention network. The method comprises the following steps: S1, performing data acquisition and processing; S2, dividing a data set; S3, constructing a fusion model; and S4, model training and project recommendation. According to the invention, a recurrent neural network (RNN) is used for modeling behaviors of a user in a session, the current interest of the user is captured through RNN potential representation, the influence of friends related to the user is captured through agraph attention network, the influence of each friend is weighed by measuring the characteristics of movement along each side according to an attention mechanism, and the current user representation and the social friend representation are combined; the project relationship is obtained from the interaction data of the user and the project, the project relationship and the user dynamic social relationship are fused into the learning process of the user and the project interaction, the influence of multiple relationships on the user and the project interaction is learned, and the recommendationaccuracy is improved, so that the model can better model the user preference.