Patents
Literature
Hiro is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.
Hiro

30 results about "Cold start recommendation" patented technology

Professional field system cold start recommendation method based on knowledge graph

A professional field system cold start recommendation method based on a knowledge graph comprises the following steps of (1) constructing the professional field knowledge graph in a semi-automatic mode, firstly, the professional field knowledge graph is manually initialized based on a professional field classification directory, and then the professional field knowledge graph is automatically expanded through internet knowledge; (2) respectively extracting tags from registration information of a user and a description text of an article, and training a professional field word vector based on an internet text; (3) firstly, respectively carrying out entity linking on the user label and the article label in the knowledge graph, and then calculating a user / article matching degree value based on the shortest path between link nodes; and (4) recommending a plurality of articles with the highest matching degree values to the user. User registration information and article content informationare considered at the same time, and cold start of the system is achieved; a knowledge graph is constructed based on professional domain knowledge, and accurate and quantitative matching of user registration information and article content information is realized.
Owner:杭州智策略科技有限公司

Cold-start recommendation method based on user preferences and trust

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.
Owner:CHONGQING UNIV OF POSTS & TELECOMM

Social information enhancement-based event social network recommendation algorithm

The invention relates to a social information enhancement-based event social network recommendation algorithm. According to the algorithm, the cold-start recommendation problem of event recommendationis solved by trying to combine information of a conventional social network and an event social network. Firstly, by utilizing social information of an event organizer, an event organizer assessmentalgorithm is introduced, thereby enhancing the effect of the recommendation algorithm. Meanwhile, the call guide effect of event participants is calculated, and the influence effect of the participants on a user is quantized and serves as a factor of a recommendation event. Secondly, the influence effects of the organizer and the participants of the event, and information in various aspects such as time, place, contents, groups and the like of the event are integrated, and selection preferences of the user are found. Finally, in combination with a search sorting algorithm, a recommendation list is calculated for each user, thereby realizing personalized recommendation.
Owner:WUHAN UNIV

Cold start recommendation method and device based on user interest migration and storage equipment

The technical problem to be solved by the invention is to extract interests and preferences of users to complete personalized user cold start recommendation by utilizing historical behavior information of new e-commerce users on a content platform in allusion to defects in the prior art. In order to achieve the purpose, according to the cold start recommendation method and device based on user interest migration and the storage equipment, group attributes of the users are constructed by utilizing behavior footprints of new users on other platforms. Then clustering is carried out based on the theory of human grouping according to the interest characteristics of the users, the users are divided into a plurality of subsets, and then commodity preferences are obtained through calculation according to the historical behaviors of the users of the subsets and serve as candidate sets of the users of the type. By the adoption of the technical scheme, the commodities which the new user may like can be predicted more accurately, the new user can find the commodities which the user may like more quickly and find a social circle suitable for the user himself / herself more quickly, the new user can obtain more friendly use experience on a new platform, and therefore the retention rate of the new user is increased.
Owner:HANGZHOU ZHICONG NETWORK TECH LOMITED CO

A context automatic coding recommendation method and system for complete cold start

The invention discloses a context automatic coding recommendation method and system for complete cold start, and the method comprises the steps: carrying out the coding of all kinds of context information of an old user, enabling the coded coding vectors to be connected in series, enabling the coding vectors to serve as input vectors, and enabling the input vectors to be input into a context perception recommender; allowing a context awareness recommender to set a local matrix for each neighbor of each old user for expressing correlation between the context-feedback information, enabling the feedback information of each old user to be obtained through calculation according to the user file and the local matrix; And finally, according to the file correlation degree between the new user andthe old user, calculating n neighbor old users corresponding to the new user, constraining feedback information of the old user through file correlation of the new user and the old user, calculating feedback information of the new user, and obtaining a cold start recommendation list of the new user. By adopting the embodiment of the invention, the context association relationship between new and old users can be mined, the new user recommendation problem is solved, and the accuracy of complete cold start recommendation is improved.
Owner:SHANTOU UNIV

User cold start recommendation algorithm based on collaborative filtering hybrid filling

The invention discloses a user cold start recommendation algorithm based on collaborative filtering hybrid filling. The algorithm is carried out according to the following steps in sequence: calculating the similarity between articles based on a user-article scoring matrix; for each article, selecting the article with high similarity as the adjacent article of the article, predicting the score ofthe article according to the score information of the target user for the adjacent article, and filling the user-article score matrix with the score to obtain a primary filling score matrix; calculating the similarity between the users according to the primary filling score matrix; for each target user, selecting a user with high similarity with the target user as an adjacent user of the user, predicting scores of articles which are not scored by the target user according to the scoring information of the adjacent user, and filling the filled scoring matrix again; and according to the final scoring matrix, selecting the first N articles with the highest prediction scores to recommend to the target user.
Owner:LIAONING NORMAL UNIVERSITY

A new project collaborative recommendation method based on multi-core fusion

The invention relates to a cold start recommendation algorithm based on commodity attribute information, and aims to solve the problem of data loss in new commodity recommendation by using a multi-core weighted fusion collaborative filtering algorithm. According to the algorithm, the incidence relation of the commodities in the attribute space is determined in a multi-core weighting mode, and therefore new projects are recommended to the user. Wherein the multi-kernel learning algorithm is based on an existing kernel function learning algorithm and is used for carrying out weighted summation on all kernel functions, so that the accuracy of the algorithm in a complex data environment is improved; Wherein the attribute similarity is obtained by calculating the similarity of attributes amongthe commodities, so that the calculated preference score of the user for commodity prediction is more interpretable; Wherein the weight is optimized through a learning method of random gradient descent. According to the method and the device, the item similarity measurement for describing user preferences can be learned according to the attribute information of the commodities, so that the new item recommendation accuracy is effectively improved.
Owner:山西开拓科技股份有限公司

Probability matrix decomposition cold start recommendation method fusing attributes and semantics

ActiveCN110851700ASolve the problem of low recommendation accuracyDigital data information retrievalComplex mathematical operationsA priori probabilityEngineering
The invention discloses a probability matrix decomposition cold start recommendation method fusing attributes and semantics. The method comprises the steps of firstly, extracting user attribute information, project attribute information, project text information and user scoring information from a database; and modeling the attribute information and the semantic information by utilizing linear regression to predict potential characteristics, and taking a predicted value as a prior probability of probability decomposition, thereby realizing fusion of the attribute information and the semantic information into probability decomposition of a scoring matrix. According to the method, attribute information and semantic information can be effectively fused into probability matrix decomposition, the ubiquitous problems of cold start and sparsity in a recommendation system are solved, and the method has higher accuracy and low algorithm complexity and is suitable for processing large-scale data.
Owner:ZHEJIANG UNIV OF TECH

Graph network cold start recommendation method

The invention discloses a graph network cold start recommendation method, which comprises the following steps of: inputting pre-scored user-article data into a trained graph network to obtain a recommendation result, the training comprising the following steps: acquiring a sampling local graph of a node set or a local sub-graph to be trained; performing distance re-marking on the sampling local graph to obtain a re-marked label; distributing initial features to nodes of the sampling local graph; obtaining a prediction label and a prediction score of the initial feature; calculating a node classification error by using the prediction label and the remarking label, and calculating a score prediction error by using the prediction score and the real score of the article-user; and performing calculating by using the node classification error and the score prediction error to obtain an overall error, and updating parameters of the graph network by using the overall error. According to the method, through local graph sampling and double-task learning, inductive node reasoning and connection prediction capabilities are further realized on the basis of a deductive graph reasoning task, and the method has a feature representation capability for out-of-graph nodes.
Owner:UNIV OF ELECTRONICS SCI & TECH OF CHINA

Meta-learning-based heterogeneous information network cold start recommendation method and device

The embodiment of the invention provides a meta-learning-based heterogeneous information network cold start recommendation method and device. The method comprises the steps: obtaining first node information and second node information, and obtaining a first feature vector corresponding to a first node and a first feature vector corresponding to each second node; respectively inputting the first feature vector corresponding to the first node and the first feature vector corresponding to each second node into a pre-trained meta-learning model, and aggregating context semantics to obtain a secondfeature vector corresponding to the first node and a second feature vector corresponding to each second node; calculating the similarity between the second feature vector corresponding to the first node and the second feature vector corresponding to each second node; determining a target second node recommended to the first node based on each similarity. According to the method disclosed in the embodiment of the invention, the accuracy of the cold start recommendation result of the heterogeneous information network can be improved.
Owner:BEIJING UNIV OF POSTS & TELECOMM

Video cold start recommendation method and system

InactiveCN110769288AImplementing Cold Start RecommendationsDimensionally consistentSelective content distributionComputer graphics (images)Engineering
The invention discloses a video cold start recommendation method and system. The video cold start recommendation method comprises the steps: S1, carrying out the dimension reduction of a new video based on an Inception network, and generating a video vector for the new video; S2, storing the video vector in Faiss; S3, summing corresponding video vectors by adopting five videos recently watched bya user, taking a mean value as a user vector, and indexing Faiss; and S4, returning a video corresponding to the video vector with a small distance from the user vector to the user. According to the video cold start recommendation method, the new video is subjected to frame capture processing to form the plurality of pictures, and the feature vector is generated for each picture to generate the video vector, and the vector index is carried out based on the Faiss to carry out video recommendation, so that the cold start recommendation of the video is realized, and the complexity is low, and therecommendation efficiency is high.
Owner:HANGZHOU QUWEI SCI & TECH

Cold start recommendation method and device, computer equipment and storage medium

The embodiment of the invention belongs to the technical field of big data, and relates to a cold start recommendation method and related equipment, which can be applied to the field of intelligent security and protection, and comprises the following steps: when a first key sent by a first server is received, calculating a first scoring parameter of a local server; when first encrypted data sent by a second server is received, calculating a first encryption similarity according to the first scoring parameter and the first encrypted data, and decrypting the first encryption similarity accordingto the first key to obtain a first decryption similarity; sending the first decryption similarity to a first server; and when the total similarity sent by the first server is received, generating anarticle recommendation table of the target user according to the total similarity. In addition, the invention also relates to a blockchain technology, and the article recommendation table can be stored in a blockchain. According to the invention, protection of user privacy information during cold start recommendation is realized.
Owner:PING AN TECH (SHENZHEN) CO LTD

Video recommendation method and device, electronic equipment and storage medium

The invention relates to a video recommendation method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring account attribute information of a target account; performing conversion processing on the account attribute information to obtain account feature information used for representing the target account; determining video attribute information of each candidate video in the candidate video set, an interaction account having an interaction behavior with the candidate video and interaction account attribute information; determining video feature information for representing the candidate video according to the candidate video, the video attribute information, the interaction account and the interaction account attribute information; and according to the video feature information of each candidate video and the account feature information of the target account, determining a target video recommended to the target account from the candidate video set. Therefore, video recommendation is performed for the new user according to the video interaction data of the existing interaction account, and the cold start recommendation accuracy of the new account is improved.
Owner:BEIJING DAJIA INTERNET INFORMATION TECH CO LTD

Article cold start recommendation algorithm integrating relationship mining and collaborative filtering

The invention discloses an article cold start recommendation algorithm integrating relationship mining and collaborative filtering. The algorithm comprises: first, using an item attribute matrix as abasis, calculating a plurality of binary relationships between every two attributes by adopting a relationship mining method; expanding limited article attributes into more relationship attributes; and then obtaining an attribute relation matrix, calculating the attribute similarity between the articles, meanwhile, fusing article scoring information for similarity weighting calculation, achievingpersonalized recommendation of the new articles. The problem of cold start of the new articles in a recommendation system can be systematically solved, and the recommendation accuracy and the articlediversity are improved.
Owner:LIAONING NORMAL UNIVERSITY

Cold start recommendation method and device and electronic equipment

The invention discloses a cold start recommendation method and device and electronic equipment, and the method comprises the steps of obtaining a user feature value of a new user for a set user feature according to the user data generated by the new user through employing a third-party application; obtaining a classification feature value obtained by classifying each product in a product set according to the user feature; obtaining a matching degree between the new user and each product according to the user characteristic value and the classification characteristic value of each product; andobtaining a product recommendation list of the new user at least according to the matching degree between the new user and each product.
Owner:ALIBABA GRP HLDG LTD

Cold start recommendation model evaluation method and system, computer equipment and storage medium

PendingCN113220557AAchieve the purpose of the evaluationHardware monitoringManufacturing computing systemsRecommendation modelEngineering
The invention relates to a cold start recommendation model evaluation method and system, a computer and a readable storage medium, and the method comprises the steps: an evaluation sample obtaining step of selecting N target users from a target user group as samples; a material alternative set acquisition step of selecting M to-be-recommended materials associated with each target user in the to-be-evaluated recommendation model application scene; a material alternative set screening step of obtaining a preference score of each target user on the to-be-recommended material, normalizing the preference score, and screening the to-be-recommended material according to the preference score to obtain a screened material; a to-be-evaluated model scoring step of establishing positive and negative samples, scoring the positive and negative samples, and splicing the positive and negative samples into a multi-dimensional vector; and an evaluation index obtaining step of calculating the multi-dimensional vector by using the sorting evaluation index to obtain a corresponding evaluation index, and performing weighted statistics to obtain an evaluation index of the recommendation model. According to the method and the system, the performance of the recommendation model is accurately evaluated under the condition that the user feedback data is lacked.
Owner:SHANGHAI MININGLAMP ARTIFICIAL INTELLIGENCE GRP CO LTD

A mobile app cold start recommendation method based on mobile app running list

The invention relates to an App cold start-up recommending method based on a mobile app operation list and mainly relates to the fields of information retrieval and system recommendation. The method comprises the following steps of acquiring large number of app information and arranging keywords, calculating a weighing undirected graph indicating keyword relationships, and calculating and pushing the weighing undirected graph according to to-be-pushed App information. After training data is trained to form an initial model, the model can greatly recommend according to the operation list content and the cold start-up problem can be well solved; the App cold start-up recommending method has great extendibility; and deficiencies of content recommendation can be overcome to some extent.
Owner:HANGZHOU YUANCHENG TECH CO LTD

A Collaboration-Based Recommender System and Its Working Method

A recommendation system based on collaboration, including: an edge system module, a collaboration center system module, and a target system module, and the edge system module, the collaboration center system module, and the target system module are connected in sequence; the edge system module collects and stores user information and sends it to the collaboration center The system module provides the required specific user information; the cooperation center system module responds to the cooperation request of the target system module, sends a cooperation request to the edge system module, and performs inconsistency elimination and reasoning fusion on the user information, and outputs the required information to the target system module. Specific user data; the target system module provides a personalized recommendation service to the user, and initiates a collaboration request to the collaboration center system module during a cold start. The present invention adds a user cooperation mechanism to the recommendation system, and the cooperation center system module completes the cooperation function between multiple recommendation systems, performs fusion processing on the multi-source data of a specific user, and effectively solves the cold start problem in the recommendation system.
Owner:SHANDONG UNIV

Recommendation method and system for cold start based on login operation and computer equipment

The invention relates to a cold start recommendation method and system based on login operation, computer equipment and a computer readable storage medium. The recommendation method comprises steps ofa data obtaining step, obtaining the known user information and the login operation information of a new user; a login information relationship tree establishment step for establishing a login information relationship tree by analyzing the data of the known user information; wherein the login information relationship tree comprises the known user information and a corresponding recommendation information type; and a new user login recommendation step for comparing and searching in the operation branches of the login information relationship tree based on the login operation information of thenew user to obtain an information type matched with the login operation information of the new user and taking the information type as a new user login recommendation information type. The corresponding recommendation information type is automatically matched based on the login operation information of the user, the user does not need to select an interest label, and the user experience is improved while the accurate and personalized information recommendation is achieved.
Owner:BEIJING XUEZHITU NETWORK TECH

Information recommendation method and device, electronic equipment and storage medium

The invention relates to the technical field of information recommendation, and discloses a method for information recommendation. The method comprises the following steps: acquiring personal characteristic information of a user, label information of the user, context information of the user and social network information of the user; obtaining a first alternative recommendation list according to the personal feature information, obtaining a second alternative recommendation list according to the tag information, obtaining a third alternative recommendation list according to the context information, and obtaining a fourth alternative recommendation list according to the social network information; extracting to-be-recommended information from the first alternative recommendation list, the second alternative recommendation list, the third alternative recommendation list and the fourth alternative recommendation list according to a preset extraction rule, and combining the extracted to-be-recommended information to obtain a cold start recommendation list; and recommending the to-be-recommended information in the cold start recommendation list to the user. Therefore, accuracy of recommending information to a new user can be improved. The invention further discloses a device for information recommendation, electronic equipment and a storage medium.
Owner:SHANGHAI MININGLAMP ARTIFICIAL INTELLIGENCE GRP CO LTD

Cold boot recommended method, device, computer equipment and storage medium

The embodiment of the present application belongs to the field of big data technology, and relates to a cold start recommendation method and related equipment, which can be applied to the field of smart security, including: when receiving the first key sent by the first server, calculating the first key of the local server Scoring parameters; when receiving the first encrypted data sent by the second server, calculate the first encrypted similarity according to the first scoring parameter and the first encrypted data, and decrypt the first encrypted similarity according to the first key to obtain Decrypting the first similarity; sending the first decrypted similarity to the first server; when receiving the total similarity sent by the first server, generating an item recommendation table for the target user according to the total similarity. In addition, the present application also relates to blockchain technology, and the item recommendation table can be stored in the blockchain. This application realizes the protection of user privacy information during cold start recommendation.
Owner:PING AN TECH (SHENZHEN) CO LTD

Cold start recommendation method and device, electronic equipment and storage medium

The embodiment of the invention provides a cold start recommendation method, which comprises the following steps of: obtaining a text vector of each article under a first article category label, and classifying the text vectors through a pre-trained classification neural network to obtain first article category information of each article; obtaining a preset first mapping relation between the user group and the second article category label, and constructing a second mapping relation between the second article category label and the user group according to the first mapping relation; according to a second mapping relation, mapping the first article category information of each article into a user group, and counting the occurrence frequency of each user group corresponding to the first article category information of each article; and marking a first article category label on the corresponding user group based on the occurrence frequency, and recommending articles to the corresponding user group according to the first article category label and the second article category label. According to the invention, the cold start of the recommendation system can be realized, and the cold start recommendation accuracy can be improved.
Owner:SHENZHEN INTELLIFUSION TECHNOLOGIES CO LTD

Cross-domain commodity cold start recommendation method and device and electronic equipment

The embodiment of the invention provides a Cross-domain commodity cold start recommendation method and device and electronic equipment. The method comprises the following steps: acquiring visual information of a visiting customer; extracting a source label in the visual information, wherein the source label has a source domain attribute; according to the source tag, obtaining a target tag associated with the source tag, the target tag having a target domain attribute, and the source domain and the target domain belonging to different domains; and recommending a target commodity corresponding to the target tag to the visiting customer. On the premise of not depending on user behavior data, the label relation can be efficiently mined, and therefore the cross-domain commodity recommendation effect is improved.
Owner:SHENZHEN INTELLIFUSION TECHNOLOGIES CO LTD

Cold start recommendation method and device

The embodiment of the invention provides a cold start recommendation method and device, relates to the technical field of terminals, and can solve the cold start problem of a recommendation system. The method is applied to a server and specifically comprises the steps that a recommendation list obtaining request from terminal equipment is received, and the recommendation list obtaining request comprises the model of the terminal equipment used by a new user; a first commodity access record and a first section access record corresponding to the first model are determined, the first application is a shopping application, and the second application is a community application; obtaining a first recommendation list according to the first commodity access record; obtaining a second recommendation list according to the first section access record; determining a similar model of the first model and a second commodity access record corresponding to the similar model, obtaining a third recommendation list according to the second commodity access record, and performing duplicate removal and sorting on commodity information in the first recommendation list, the second recommendation list and the third recommendation list to obtain a final recommendation list; and sending the final recommendation list to the terminal equipment.
Owner:HONOR DEVICE CO LTD

Cold start recommendation algorithm based on implicit factor prediction

PendingCN113987363AAddressing missing historical scoring dataSolve problems that cannot be recommendedDigital data information retrievalComplex mathematical operationsFeature vectorData set
The invention discloses a cold start recommendation method based on implicit factor prediction, and belongs to the field of machine learning and recommendation systems. According to score prediction, scores of the users on the items are calculated by predicting user preference factors, and the problem that new users cannot be recommended due to lack of historical scores of the users in a collaborative filtering algorithm is solved. The invention provides a cold start recommendation method based on implicit factor prediction by combining a matrix decomposition algorithm and a collaborative filtering algorithm aiming at the cold start problem of a user. According to the method provided by the invention, implicit factor feature vectors of a user and a project are obtained from a scoring data set through a matrix decomposition algorithm; if the user is a common user, a prediction score is calculated in combination with the implicit factor vector and the bias value of the user item. And if the user is a new user, obtaining an implicit factor feature vector (preference factor) of the user through a collaborative filtering algorithm, and finally calculating a prediction score. Compared with a traditional recommendation algorithm, the algorithm provided by the invention is higher than the traditional recommendation algorithm in both accuracy and expandability, and has certain practical significance.
Owner:NANJING UNIV OF AERONAUTICS & ASTRONAUTICS

A Witkey Task Recommendation Method Based on Hidden Factor Model with Correction Vector

ActiveCN109711651BAccurate description of interest relationshipImprove accuracyResourcesFeature vectorFeature set
The invention provides a hidden factor model weapon task recommendation method with feature vectors. According to the technical scheme, the method is used for solving the problem of information overload appearing in a Weiguan platform and recommending suitable tasks to users, and mainly comprises the following steps: firstly, carrying out user interest degree quantification and feature set construction, quantifying original behavior data, reading the original behavior data into a feature set, and introducing negative sampling to enrich the original behavior set; then establishing a hidden factor model with a correction vector, carrying out training, and generating a recommendation result; and finally, for users and tasks which newly enter and do not have behavior information, providing cold start recommendation based on the correction vector group. According to the characteristics of the Weiguan platform data, the user characteristics and the task characteristics correspond to the correction vectors and are introduced into the hidden factor model, more accurate modeling is carried out on the interest of the user, meanwhile, the cold starting problem when a new user and a new task enter is solved by utilizing the user characteristic vectors and the task characteristic vectors in the model, and the practicability is high.
Owner:厦门一品威客网络科技股份有限公司

Information-enhanced meta-learning method for relieving cold start problem of recommended user

The invention discloses an information-enhanced meta-learning method for relieving a recommended user cold start problem. The method comprises the following steps: constructing a user-article bipartite graph according to a comment relationship of user articles; using the user neighborhood information and the article description text as the source of node information, performing random sampling on the constructed bipartite graph, and constructing meta-learning tasks, wherein each task represents a new user to carry out cold start recommendation to simulate a scene for realizing recommendation for the new user; using a bert method to perform feature extraction work of text data for each meta-learning task to obtain preference information; and using the global parameters of the preference information guide element for generating local parameters of an embedded generation function of each user, inputting element learning tasks into a recommendation model to obtain predicted scores of the users on articles, updating element learning parameters, and directly applying the trained parameters to untrained new users. According to the method, the problem that preference recommendation inaccuracy cannot be evaluated due to less new user interaction is relieved.
Owner:ANHUI AGRICULTURAL UNIVERSITY

Cold start recommendation method and system of television applet

PendingCN114268836ASolve the problem that the amount of data is small and it is difficult to make recommendationsKernel methodsCharacter and pattern recognitionCluster algorithmEngineering
The invention provides a cold start recommendation method and a cold start recommendation system for a television applet. The cold start recommendation method comprises the following steps: step 1, collecting behavior data of using video on demand and behavior data of using the television applet by a user at a television end in a data dotting manner; 2, preprocessing the collected behavior data; 3, performing machine learning training by using a factorization machine FM algorithm to obtain user data, and judging the similarity of the users according to the distance; 4, clustering the user data by using a K-means clustering algorithm; 5, if the most favorite applet of the user exists in N users with the closest User Embedding distance of the user, recommending the applet to the user; otherwise, hot applets in the cluster where the user is located are calculated and recommended, and therefore television applet cold start of the user is completed. The method solves the problem that the television applet is small in user data volume and difficult to recommend at the present stage.
Owner:SHANGHAI SHIJIU INFORMATION TECH CO LTD

Meta-learning method and device of cold start recommendation system, equipment and storage medium

The invention discloses a meta-learning and device of a cold start recommendation system, equipment and a storage medium. The cold start problem is solved through a meta-learning method of a model level and a heterogeneous information network of a data level, a meta-learning model is prevented from entering local optimum to a certain extent by adopting a memory enhanced meta-optimization method, and a recommendation result of a cold start recommendation system is obtained according to a recommendation model. According to the method, on the data level, related semantic information on data is enriched by constructing meta-paths; and in the model level, a memory capable of storing specific semantics is adopted for guiding a model with semantic parameter initialization, and a meta-optimization method is adopted for optimizing the method so as to achieve rapid adaptation.
Owner:HUNAN POLICE ACAD

Cloud exhibition content recommendation method, system and equipment based on generative adversarial network

The invention provides a cloud exhibition content recommendation method based on a generative adversarial network. The method comprises the steps of 1, constructing a generator which is used for generating a score sequence of user browsing behaviors, 2, constructing a discriminator, and carrying out authenticity judgment on the score sequence to obtain effective user browsing behavior data, and 3, recommending unique effective user browsing behavior data as a target result for determining output. According to the method, the user characteristics and the content characteristics are modeled, the cold start recommendation is provided according to the registration characteristics of the new users, the recommendation based on the user characteristics is provided by synchronously utilizing the characteristic similarity of different users, and the recommendation is performed by utilizing the content characteristics to reflect the user preferences from multiple angles; therefore, more latitude combination display is brought to the user, and interaction is carried out to obtain the feedback of the user.
Owner:TURING AI INST NANJING CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products