286results about How to "Solve the cold start problem" patented technology

The invention relates to a collaborative filtering algorithm, in particular to a collaborative filtering recommendation algorithm based on article sorting and user sorting, which is characterized by including steps: A, article clustering and sorting; B, user clustering and sorting; C, integration of article clustering and user clustering; and D, sorting recommendation. Compared with the prior art, collaborative filtering recommendation algorithm has the advantages that the data clustering is completed by means of modified KMEANS algorithm, the method is simple, and extensibility is improved while the problems of sparsity and cold boot are solved.

The invention discloses an information recommending method based on a social network. The information recommending method includes the following steps that first, trust degree and similarity between users are calculated, and a user relation matrix is constructed through weighted values; second, the users are clustered through a community discovering algorithm, and then a closest neighbor set of the users is formed; third, scores are predicted, and a recommending list is generated. The information recommending method based on the social network can achieve the following advantages that first, the cold start problem is solved: trust degree is introduced into the method, if enough neighbors cannot be obtained according to the common grading articles in the recommending process, trustable friends can serve as the start point of prediction, and thus the cold start problem can be relieved, and user coverage can be improved; real time performance is improved: community division is performed on the user network through the community discovering algorithm commonly used in social network analysis, in other words, same user interests are clustered, and thus the time for finding the neighbor set of the users is greatly shortened, and the real time performance of the information recommending response is improved.

The invention provides a news recommendation system and method based on an FOLFM model. Based on a content recommendation method, a news content model is expressed abstractly through a latent class model and content characteristics, and an individual latent class preference model is built for each user. Real-time training is carried out on a real-time behavior record of a user to obtain preference, on certain latent class news, of the user, whether the news is recommended to the user is determined through calculation, and a final news recommendation list is obtained after a series of processing processes. The news recommendation system and method based on the FOLFM model deeply excavate the interest of the user, improve recommendation accuracy and satisfaction of the user, avoid a cold starting problem of the news, and guarantee performance under the condition that the recommendation effect is improved as much as possible. The experiment shows that the news recommendation system and method based on the FOLFM model not only guarantee the requirements for high accuracy and high speed, but also realize visual real-time recommendation for the user.

The invention discloses a multilayer quotation recommendation method based on a literature content mapping knowledge domain, and belongs to the field of information recommendation and intelligent information processing. The method comprises the following steps: firstly, obtaining the query requirement of a user, wherein the query requirement consists of the key words of the title and the digest of a thesis which needs to recommend a quotation thesis or quotation literature; then, on the basis of the literature content mapping knowledge domain, expanding and querying a retrieval word, wherein the mapping knowledge domain consists of the research object word and the research behavior word node of the literature, and edges which express various semantic relations including synonymy, synonym, an up and down position, part-whole, juxtaposition and the like; and finally, constructing the inverted index of the literature in a data set, selecting a candidate quotation, calculating the similarity between the candidate quotation and query, and adopting a gradient progressive regression tree to carry out quotation recommendation. The method carries out multilayer quotation recommendation on the basis of the literature content mapping knowledge domain, enlarges the range of the candidate quotation, accurately expresses the research object and contents of the thesis, improves efficiency for users to obtain a relevant literature and has a wide application prospect.

The invention discloses a combined wrong question recommendation method based on a knowledge graph. Wrong questions relevant to weak knowledge points of a learner can be accurately recommended for the learner by adopting the method. The method comprises the steps that knowledges are extracted from large-scale unstructured test question data to establish the knowledge graph; text mining and word segmentation are conducted on the wrong questions of the learner to extract wrong question keywords, and thus knowledge points including in the wrong questions are determined; semantic near neighbors of the knowledge points are obtained by analyzing semantic similarity of the test questions; the wrong question knowledge points are mapped into the knowledge graph to obtain test question entities conforming to their knowledge points. In addition, similarity weight calculation is conducted on a test question library to obtain similarity matrixes of test paper, a collaborative filtering technology is utilized to obtain recommended test questions of the wrong questions. Finally, two recommendation results are further combined in weighing, mixing, superposing and element-level modes, and a final recommendation result is given.

The invention relates to an interest recommendation method and device, a server and a storage medium. The method comprises the following steps that: obtaining the characteristic information of a target user; according to the characteristic information, adopting an interest similarity prediction model to predict an interest similarity between the target user and a user group, wherein the interest similarity prediction model is realized on the basis of a tree model coding hybrid linear algorithm; according to the interest similarity, determining a recommended user which has the similar interestwith the target user in the user group; and obtaining the interest list of the recommended user, and creating a recommendation list for the target user. By use of the method, the problem of the cold boot of a traditional collaborative filtering algorithm can be solved, wherein the problem is a recommendation problem when the user does not have enough watching history. The accuracy of a recommendation result is improved, a click rate is obviously improved, and a purpose of "everyone has their own view" of personalized recommendation is realized.

The invention discloses a university library-oriented books personalized recommendation method, and solves the problems of poor large-scale data storage and query, extendibility and recommendation effect in an existing books recommendation algorithm of a university library. According to the basic thought, the method comprises the following steps of firstly, building a graph model by taking readers, books and the like in the library as nodes; secondly, converting operation log files of the readers into a reader-books category preference matrix, calculating similarity between the readers by the reader-books category preference matrix and a reader personal information matrix, and establishing an associated graph spectrum by taking operations and mined information as edges; thirdly, by combining the associated graph spectrum with spectral clustering, proposing a new books personalized recommendation model, and performing calculation to obtain class cluster distribution about the readers; and finally, when books recommendation needs to be carried out, calculating a recommended books list according to a collaborative filtering algorithm in a class cluster corresponding to a reader.

The invention provides a user portrait based personalized precision marketing method, a server and a storage medium. The method comprises the steps of obtaining label information of a user, building auser portrait according to the label information, obtaining the current Internet surfing behavior of the user, performing prediction and analysis on the behavior, screening out contents which the user may like or be interested in based on the user portrait when the user is predicted and analyzed to have a purchase or consumption intention, and carrying out personalized recommendation according tothe contents which the user may like or be interested in. Each user has a unique user portrait, and a list of contents which the user may like or interested in can be screened out through performingdeep analysis by combining the user portrait and the current Internet surfing behavior, thereby being capable of well performing personalized recommendation according to the condition of each user, improving the recommendation accuracy, and also being capable of well solving a problem of cold start of new users at the same time.

The invention provides an entity recognition method and device and computer equipment. The method comprises the steps of performing recall processing on to-be-identified corpora; adopting a dictionarymatching module to recall possible candidate slot position combinations of the corpus to be identified; marking the to-be-identified corpus according to the to-be-identified corpus; acquiring a corresponding slot position labeling sequence: respectively obtaining a coding vector of a corpus to be identified and a coding vector of each slot marking sequence by utilizing a deep learning network, selecting the coding vector of the slot marking sequence most similar to the coding vector of the corpus to be identified, and taking the slot marking sequence corresponding to the coding vector as an optimal slot marking sequence, thereby obtaining a named entity of the corpus to be identified. Therefore, by utilizing the advantages of the rule template and the deep learning algorithm, the named entities of various types of corpora can be quickly, simply and accurately identified, the cold start problem is solved, the obtained identification result is generally a computer language, and the electronic equipment can directly respond to the identification result.

The invention discloses an information push method based on internet-surfing log mining and user activity recognition.The method includes a data preprocessing stage, a potential pushing user screening stage and a timely push stage.The data preprocessing stage comprises a webpage portrait, webpage level classification and a user portrait.In the potential pushing user screening stage, similarity between new information and historical web pages which users have access to is solved through a method based on matrix decomposition, and potential pushing users who may be interested in the new information are screened out.The timely push stage includes situational information acquisition and activity recognition based on a mobile intelligent device, and information timely pushing adaptive to the activity situation.By mining internet-surfing log data, potential users interested in information are found, when new information is generated, the new information is pushed to the potential users in good time, and therefore information push efficiency is improved.

The invention relates to an individualized recommendation method based on user preferences and commodity properties, belonging to the field of machine learning. The individualized recommendation method comprises the following steps of adding attribute information of a project and doing recommendation by commodity property information when preference information does not exist. Meanwhile, the recall ratio of a recommendation system is improved by the recommendation method. With the adoption of the individualized recommendation method based on the user preferences and the commodity properties, the problem of cold starting based on a new project is solved.

The invention relates to a recommending method integrating user and project rating and characteristic factors, which mainly aims at solving the problems of cold starting and data sparsity in traditional collaborative filtering recommending algorithm. A selected experimental data set is classified into a training data set and a test data set; similarity based on users and user characteristics and similarity based on projects and project characteristics are respectively calculated on the training data set; the similarities based on the users and the user characteristics are combined by selecting appropriate weight; a user nearest neighboring matrix is calculated respectively according to a dynamic selection threshold value method and a K nearest neighbor (KNN) method, an optimum method is selected through the comparison, and user recommend results are calculated by utilizing a recommending formula; the similarities based on the projects and the project characteristics are combined by selecting an appropriate weight; a project nearest neighboring matrix is calculated respectively according to the dynamic selection threshold value method and the KNN method,, an optimum method is selected through the comparison, and project recommend results are calculated by utilizing a recommending formula; and the user recommend results are combined with the project recommend results by selecting an appropriate weight.

The invention provides a personalized recommendation method based on a user portrait. The method comprises the following steps: acquiring label information of a user, and according to the label information, establishing the user portrait, and acquiring the initial state of the user; according to the user portrait and the initial state, determining a one-step transition matrix of the user, and calculating an interest list of the user according to the one-step transition matrix; performing recommendation according to the interest list. The user portrait and the initial state are used as basis at the same time, so that personalized recommendation is performed for each user, the recommendation accuracy rate is increased, and besides, the problem of cold start by a new user is solved. In addition, the invention further provides a personalized recommendation device based on the user portrait.

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 discloses a recommending system and method combining an SNS and a search engine technology. The recommending system and method combining the SNS and the search engine technology introduces user search behaviors and SNS friend information so that the user search behaviors and the SNS friend information are used as references to obtain recommending results from an angle of a user himself and an angle of a friend of the user. The recommending system comprises a friend information obtaining module, a recommending result generation module based on the SNS friend intimacy, a search behavior information obtaining module, a recommending result generation module based on the search behavior information and a result integrating module. The recommending method comprises a first step of obtaining friend information, a second step of obtaining the recommending results based on the SNS friend intimacy, a third step of obtaining the search behavior information, a fourth step of obtaining the recommending results based on the search behaviors, and a fifth step of integrating the recommending results. According to the recommending system and method combining the SNS and the search engine technology, the characteristics of sharing and exchanging of the SNS are applied, on the one hand, the transmission efficiency of products in an e-commerce platform is improved, on the other hand, complementary to the technology which is based on the search behaviors is formed, and therefore the accuracy, the reliability and the comprehensiveness of the recommending results are improved.

The invention discloses a method for constructing industrial knowledge map based on industrial chain, which relates to the technical field of knowledge map in artificial intelligence. Firstly, the industrial chain is modeled, and then the industrial knowledge map is constructed according to the constructed industrial chain model. The industry knowledge map based on the industry chain established by the technical proposal of the embodiment of the invention can clearly reflect the entity-relation-entity and ' entity-eroperties-attribute value ' between the industry chains and within the industrychains, can facilitate financial researchers to further use knowledge mapping to study industrial chain conduction and event driving to find important events, to analyze information emotion and so on; The construction of industry knowledge map based on industry chain can effectively reduce the problem of excessive noise in the process of industry knowledge extraction, and the cold start problem of knowledge extraction can be avoided by using feature thesaurus to construct entity relationship.The embodiment scheme can realize incremental knowledge learning and effectively reduce the dependenceon professional researchers.

The invention provides a personalized digital library resource recommending method and system based on mixed recommendation. The recommending method includes the following five steps of 1 user data managing and maintaining; 2 book resource data extracting and managing; book resource content data analyzing; 4 user behavior data analyzing; 5 personalized book resource recommending. Existing digitized information in a digital library is used for recommending book resource in a personalized user-oriented mode, a new thought is provided for effectively using the book resource, and the personalized digital library resource recommending system based on mixed recommendation is provided based on the method. The method is used for data analysis and resource recommendation, a user can be effectively helped to find interested book materials, the circulation use ratio of the book materials is effectively increased, user satisfaction is effectively improved, and effective technical support is provided for knowledge innovation in China.

The invention discloses a recommending method of a knowledge map based on the location service field, which includes: extracting a location entity and obtaining an entity set as a seed set of the knowledge map, corresponding the seed set with the entities in the knowledge map, forming an entity correspondence table, in a knowledge map triple in that knowledge map, using Word2Vec model to embed vocabulary into n-dimensional space, generating corresponding vectors, obtaining a position or domain entity vector set E and a relation vector set R, translating the entity vector set E and the relationvector set R by using a TransE algorithm, and obtaining a triple vector set capable of quickly calculating semantic similarity between entities; according to the location or domain entity vector setE, calculating respectively the semantic similarity simA, B (A, B) between the searching locations or domains to generate the semantic similarity matrix of the tourism location, using Semantic Similarity Matrix for Top-K Recommendation List, clustering the recommendation list according to machine learning clustering algorithm, and recommending the clustering result to the user. The method has highprecision and can solve the problems of cold start and sparsity.

The invention discloses a personalized POI recommendation method based on multi-influence embedment. By conducting associative embedded learning on seven bipartite graphs (a user-user graph, a user-time frame graph, a POI-time frame graph, a POI-region hierarchy graph, a POI-category hierarchy graph, a user-gender graph and a user-POI graph) and a check-in sequence, influences on the aspects of social contact, time, geography, semantics, user gender, user preference and sequence are integrated; meanwhile, certain expansibility is achieved, influences on other aspects can be conveniently integrated, so that the problems of data sparsity and cold start are effectively solved, and high-quality POI recommendation is provided for users.

The embodiment of the invention discloses a knowledge graph processing method and device for commodity recommendation, relates to the technical field of Internet, and can construct a knowledge graph between a commodity and a user. The method includes: according to the initial user vector and the initial commodity vector, carrying out training and learning on a first neural network model, and obtaining a vector result of a hidden layer in a calculation result of the first neural network model and taking the vector result as vector representation of a user and vector representation of a commodity; Calculating the distance between the vector representation of the user and the vector representation of the commodity, and taking the calculated distance as the similarity; Constructing a second neural network model by using the initial user vector representation and the initial commodity vector representation, and taking a vector result of a hidden layer in a calculation result of the second neural network model as a neighborhood vector representation of a commodity; Calculating the relevancy between the commodities according to the obtained neighborhood vector representation; And constructing a knowledge graph between the user and the commodity.

The invention discloses a fine-grained radio station audio content personalized organization recommendation method, which includes the steps of automatically segmenting and marking audio programs of a radio station according to semantics, mining user preferences based on the Internet big data, automatically arranging a personalized program list and real-timely pushing programs, and relates to the fields of audio processing, machine learning, big data analysis, recommendation systems, data mining and the like. According to the method, an algorithm process of automatically segmenting and marking traditional broadcast audio programs based on the semantics is provided, a technical scheme for personalized content recommendation based on the Internet big data is also provided, and a fine-grained audio content personalized organization recommendation method is achieved. According to the method disclosed by the invention, the cold start problem is integrated into account, the program list organization and generation, real-time program switching, real-time program push and other factors during the listening time of the user are combined, a simple mode that the current radio station transplants the FM live broadcast to the Internet streaming media for playback is changed, and from the view point of audience users, the needs that the users listen to the contents of interested programs at right time can be met.

The invention discloses a method, user modeling equipment and a corresponding system for carrying out personalized recommendation for users. The method comprises the following steps: extracting information models of famous persons concerned by a user from an information model database, wherein the famous persons are persons meeting predetermined conditions in a social network, and information models established for the famous persons on the basis of a first modeling strategy are stored in the information model database; establishing a user model for the user according to the information models of the famous persons concerned by the user and a second modeling strategy; acquiring information which is extracted according to the user model and is suitable for the user and recommending the information to the user. According to the method, the user modeling equipment and the corresponding system disclosed by the invention, the user model established according to the information models of the famous persons concerned by the user can accurately represent interests of the user, and the information recommended for the user according to the user model can be accurately matched with the interests of the user, thus the information recommendation accuracy is high, and the cold start problem caused by shortage or absence of historical behaviors of users is solved.

The invention discloses a dialogue intention recognition method and system based on a multi-dimensional semantic interaction representation model, and belongs to the field of natural language processing dialogue systems. The method comprises the steps that (1) establishing a dialogue knowledge base wherein the knowledge base comprises universal common dialogue data, statements of a user in a business scene and intentions to which the corresponding statements belong; (2) performing feature extraction based on a pre-trained language model on dialogue information in the dialogue knowledge base toobtain a semantic vector; (3) obtaining a semantic vector of the current dialogue information; (4) constructing an interactive attention mechanism and a convolutional neural network in combination with semantic vectors of the dialogue statement and the current dialogue statement in the knowledge base, and calculating to obtain a confidence coefficient; and (5) screening the confidence coefficients to obtain an intention recognition result or judge that the intention in the knowledge base is missed. According to the method, the problems that a traditional pre-trained language model does not focus on the semantic information level, so that the distinction degree is insufficient, and sensitive information is neglected are solved, and the recognition accuracy is higher.

The invention discloses a personalization recommendation method aiming at teaching resources, which at least comprises the following modules: 1) a content-based recommendation module: recommending through taking the content and the attributes of teaching resources as the basis to solve the cold start problem of a new project; 2) a user-based collaborative filtering module: downloading and browsing the similarity of matrix calculation users according to the user rating, and then recommending the users; 3) a project-based collaborative filtering module: downloading and browsing the similarity of teaching resources according to the user rating, and then recommending the users; 4) a neural network module: having strong dynamic nonlinear mapping capability and high precision and satisfaction on teaching resource recommending. The recommending effect is superior to the linear interpolation singly adopting the recommendation module or the recommendation results of the modules.

The invention relates to an item information recommending method based on user network data. The method is characterized by including the following steps of firstly, obtaining item demand information of a user through the user network data, and setting up a user demand feature library; secondly, setting up an item information feature library according to item network data; thirdly, matching the user demand features with the item information features through a text and semantic similarity algorithm, calculating the similarity between item information and user demands, conducting sequencing and filtering, and finally generating the recommending result; fourthly, feeding the generated recommending result back to the user demand feature library and the item information feature library, and conducting training and renewing. Compared with the prior art, the method has the advantages of being comprehensive in information, wide in application range and the like.

The invention belongs to the technical field of the internet and discloses a cross user-based cross-domain recommendation method. The cross user-based cross-domain recommendation method comprises the steps of processing data of a source domain and a target domain, screening out cross users, and fusing the data, in the two domains, of the cross users to form a new scoring matrix; and then performing normalization processing on scores, and performing calculation by using a user similarity calculation formula. According to the method, user similarity can be calculated by completely depending on data of an original domain, and finally recommendation is generated, so that the cold-start problem of the users is solved to a certain extent; an improved user similarity calculation formula of the cross users is proposed, so that a new cross-domain recommendation algorithm is formed; and according to the cross user-based cross-domain recommendation algorithm, the data, in the source domain, of the cross users is migrated to the target domain, and data richness of the target domain is enhanced, thereby obtaining a more accurate personalized recommendation result.

The invention discloses a personalized recommendation method based on user data of on-line courses. The method comprises following steps of: 1), establishing label mapping relations among courses, labels and titles; 2), calculating an error rate list ET of a user to courses by a course recommendation system according to learning records of the user; 3), generating error lists of all courses according to the list ET and calculating the similarity between two courses according to error lists of courses; and 4), as for each user, calculating prediction weight of error rate of each course of the user according to similarity; calculating recommendation weight of each label according to mapping relations and prediction weight of error rate; calculating recommendation weight of each title according to mapping relations and label recommendation weight and then generating a recommendation list for the user according to recommendation weight of titles. The personalized recommendation method based on user data of on-line courses not only helps to solve a cold start problem but also actively attracts attention of the user in order to improve learning motivation.

The invention discloses an RPA robot intelligent element positioning and picking method and system. According to the method, firstly a problem that a remote desktop, a virtual system and other elements cannot be positioned and operated is solved through employing a computer vision technology combining template matching, OCR and image retrieval, and a proposed new scheme can more accurately and stably locate and pick up the positions and contents of the elements on the picture; secondly, an element positioning scheme based on template matching is adopted, a large number of labeling samples arenot needed, and the scheme is more stable and controllable; thirdly, pixel point features, corner point features and convolution features of a deep neural network are fused, so that the template matching effect is more accurate and stable, and meanwhile, the OCR technology can further assist in positioning elements with text information; and finally, a template library retrieval engine is constructed by adopting an image retrieval technology, and which application the page picture to be processed belongs to can be accurately judged, so that the problem of low retrieval and matching speed caused by template library growth in multiple scenes is solved.

The invention relates to a blockchain consensus method. The method comprises the steps that a production node in a blockchain network generates a to-be-confirmed block according to transaction data; the production node broadcasts the to-be-confirmed block to a signature node; the signature node carries out digital signature on the to-be-confirmed block by using a private key based on a primary voting principle, and the primary voting principle is that when the signature node receives a plurality of to-be-confirmed blocks at the same block height, only the received first to-be-confirmed block is digitally signed; the signature node sends the digital signature to the production node; and the production node generates a to-be-valid block according to the digital signature, adds the to-be-valid block to a self-stored block chain, and broadcasts a chain above the to-be-valid block. The blockchain consensus method uses one node and one ticket, meaningless hash is not needed, monopoly type overall control of large computing power and large stock rights is also eradicated, the cold start problem of new nodes is solved, meanwhile, the block production frequency of each node is controlled, and therefore part of nodes are prevented from being cut alone for a long time.