43results about How to "Precise personalized recommendation" patented technology

The invention discloses an association-rule recommending method based on self-adaptive multiple minimum supports. The method comprises the following steps of firstly, establishing a commodity-classifying hierarchical tree according to commodity classification, and classifying concrete commodities according to the classifying hierarchical tree; next respectively carrying out minimum-support threshold-value setting on each concrete commodity and the upper-layer class of a concrete-commodity layer, and then mining frequent item sets and generating rules by utilizing a multiple-minimum-support association-rule expanding algorithm on the basis of the support threshold-value setting, wherein the threshold-value setting relates to the influences of time factors, concrete-commodity price factors and concrete-commodity brand factors; finally generating recommendation for each user by adopting a TOP-N recommending method. When personalized recommendation is made for the user by the association-rule recommending method, the characteristics of different objects can be better embodied by considering the influences of many factors on the multiple-minimum-support threshold-value setting for the concrete commodities and the classes; meanwhile, a data-sparsity problem and a cold-starting problem in a recommending system are relieved, so that the personalized recommendation can be more accurately made for the user.

The invention discloses a real-time goods recommendation method based on customer shopping intention exploration. The method comprises the following steps: 1, constructing a tree-shaped customer shopping intention model; 2, detecting user behaviors in real time and identifying user intention; 3, detecting the change of the user intention in real time; 4, tracking the change track of the user intention and predicting the deep intention of the user: predicting the deep intention of the user based on the discrete intention of the user; and 5, determining a goods recommendation strategy: determining the goods recommendation strategy according to the shopping intention of the customer. The method disclosed by the invention automatically identifies the shopping intention of the customer, performs precise individualized recommendation, and improves the customer satisfaction; and moreover, the method automatically detects the change of the shopping intention of the customer and adjusts the goods recommendation strategy. The method adapts to the changing interest of the customer in real time, deeply explores the shopping intention of the customer based on the user intention model and provides professional recommendation service to the user.

The invention discloses a matrix decomposition recommending method based on an expanded label, comprising: constructing an item-label matrix based on the item label data and calculating the label similarity; constructing a first item pair-label vector based on the item-label matrix; and according to the label similarity, extending the first item pair label-vector as a second item pair-label vector; establishing an item similarity matrix based on the second item pair-label vector; solving a user implicit characteristics matrix and an item implicit characteristics matrix based on the item similarity matrix and the pre-constructed item scoring matrixes; predicting the score by the user to the unrated items; and building a list of recommended items for the user. At the same time, the invention also discloses a matrix decomposition recommending system based on the expanded label. The invention can provide the user with more accurate personalized recommendations when the label is sparse, and relieves the problem of cold starting of the item to a certain extent.

The invention relates to a catering resource recommendation method based on a video application and a system thereof and belongs to the information pushing field. A problem that routine video application pushing information does not match with a user demand so that a user experience is poor is solved. The method is characterized by acquiring characteristic information of a user when the user operates the video application, and according to the characteristic information, acquiring a plurality of characteristic labels of the user; matching the plurality of characteristic labels of the user with a plurality of preset characteristic labels of catering resources in a catering resource database one by one so as to acquire a matching degree of the user and the catering resources in the catering resource database; and according to the matching degree of the user and the catering resources in the catering resource database, screening the catering resources suitable for the user and pushing to the user. The method and the system are used for accurately recommending catering resource information to the user in a customized mode, an experience of the user to use the video application is increased and maneuverability of the video application is enhanced.

The invention discloses a collaborative filtering recommendation method of a theme based on viewpoint enhancement. The method includes the steps that first, comment text attribute words are extractedbased on an LDA theme model; then, the similar attribute words are classified as one category by utilizing a word2vec model to calculate relations among the attribute words based on the attribute words extracted by the LDA; the attribute words with the same meanings are collected as an attribute surface; the sentiment polarity of comment texts is obtained through the analysis of attribute surfaceviewpoint enhancement; next, a scoring matrix is constructed according to the polarity of sentiment words and attribute words, and the similarity among users is calculated according to a scoring method and through a collaborative filtering algorithm; finally, goods with high scores in the matrix are recommended to other users according to a k-nearest neighbor algorithm. Experimental results show that the collaborative filtering recommendation method is excellent in F value extracted by attribute words and mean absolute error of recommendation.

The invention provides an intelligent news recommendation system based on emotion protection. A method comprises the following steps: 1, extracting news features and text feature words by utilizing a BERT pre-training model, and constructing a news feature matrix through news feature vectors; 2, performing emotion filtering on the text information to establish an emotion grading model, and performing emotion grading on user comments, news titles and contents to distinguish negative and positive degrees; 3, clustering news labels through a clustering algorithm, distributing weights to news browsed by users according to user comment emotion levels and user behavior time, and constructing a user matrix according to user feature information; 4, predicting the emotion level of the user in the next time period according to the time sequence of the user emotion; and 5, generating a recommendation table by calculating the similarity between the user and the news vector, predicting the emotional state of the user, and recommending the news in proportion by using a Bayesian method to realize dynamic pushing. According to the invention, negative energy and negative public opinions are prevented from hurting the psychology of the user and harming the public safety of the society.

The invention discloses a personalized diet recommendation method and system based on a knowledge graph. The method comprises the steps: acquiring menu data of a knowledge base, extracting attribute information of a menu, and constructing a menu entity node; acquiring existing user historical behavior data, calculating similarity between menus, and generating a menu knowledge graph according to the similarity; according to the menu knowledge graph, generating a first recommendation candidate set with high similarity with the historical behavior data of the user; calculating the tendency degree of the user to the menu attributes, and generating a relation graph of the user and the menu attributes; screening out an optimized second recommendation candidate set according to a relation graph of the user and menu attributes; calculating and generating a relation graph between the users according to the relation graph between the users and the menu attributes; and generating a final recommendation candidate list according to the relation graph between the users and the collection behaviors of the target similar users. According to the method, layer-by-layer recommendation list screening is carried out based on the multiple knowledge graphs, and accurate personalized recommendation is realized.

The invention relates to a pushing system for academic conference papers based on data mining, which comprises a user login registration module for user registration and login, and a system for modifying and saving user information through a database. The data capture module is used for crawling relevant paper information from the designated conference website by using Webmagic open source tool; paper browsing and recording module, used for browsing, searching, collection, note-taking and message functions; paper update detection and new paper detection module, which is used to regularly monitor the update of papers on the conference website; the paper pushing module is used for triggering the mail pushing task every day when a new paper is detected or a recommended paper is needed for production. The personalized recommendation module is used for inferring the papers that the user may like according to the historical data of the user. The invention frees researchers from complicated and time-consuming activities of searching and tracking papers and improves scientific research efficiency.

The invention provides a single-classification collaborative filtering method based on a user personality hierarchical structure, and relates to the field of computer data processing. The method for applying the user personality hierarchical structure to the single-classification collaborative filtering algorithm provided by the invention is suitable for processing implicit feedback data. A traditional single-classification collaborative filtering method does not have a hierarchical structure fused with a user; according to the method provided by the invention, based on weighted matrix decomposition, users are layered by introducing personality trait information, a layered structure of the users is constructed, existing information is fully utilized, the problem of data sparseness is relieved, and the unfriendliness of a recommendation system to new users is reduced, so that personalized recommendation of commodities to the users is more accurate.

The embodiment of the invention relates to a recommendation content determination method and device and a storage medium. The method comprises the steps of determining predicted click probabilities corresponding to multiple to-be-recommended contents corresponding to a user according to a predicted neural network model, and sorting the multiple to-be-recommended contents according to the predicted click probabilities to determine target recommended contents. According to the invention, the technical problem that accurate personalized recommendation cannot be carried out for new contents and new users due to the fact that a recommendation system in the related technology cannot carry out modeling on the new contents and the new users is solved.

The invention relates to a user browsing behavior reporting method and device, a readable storage medium and electronic equipment. The method comprises the steps: traversing an information base to determine a target display area meeting reporting conditions at least according to display state records, recorded in the information base, of all display areas, wherein the reporting conditions at leastcomprise that part of the display areas are displayed; and in response to the determined target display area, reporting user browsing behavior information, the user browsing behavior information at least comprising target display data associated with the target display area. According to the method, the defect that the display data which is not truly seen by the user is reported to the server canbe effectively avoided, the accuracy of analyzing user preferences by the server is improved, the server can accurately perform personalized recommendation for the user, and the browsing experience of the user is improved. Moreover, the process of determining whether the display area meets the reporting condition by the terminal is independent of the browsing operation of the user on the page, sothat the process of determining the target display area meeting the reporting condition is simplified.

The invention relates to a fragmented knowledge intelligent aggregation method. The method comprises the steps that step 1, knowledge element ontologies are defined; step 2, fragmented knowledge ontology associated aggregation is defined; step 3, associated aggregation rules based on ontology implication are established; step 4, the determination of aggregation associated rules is conducted; step5, the determination of fragmented knowledge associated rules based on the knowledge element ontologies is conducted; step 6, fragmented knowledge aggregation associated discovery is conducted; step 7, fragmented knowledge aggregation is achieved. According to the method, the associated relationship between two or more knowledge element ontologies is determined through supporting and confidence level determination, and fragmented knowledge aggregation is achieved through a strong associated method; fragmented knowledge characteristics are analyzed by the method, facing the demands of online study, original solid knowledge structures are repartitioned and dynamically aggregated into a knowledge cluster with a self-organizing capability to finally complete fragmented knowledge aggregation, and learners are guided to fully utilize fragmented time to accurately acquire meaningful knowledge contents.

The embodiment of the invention discloses a display method and device for inputting association words, a storage medium and electronic equipment. The method comprises the steps that historical behavior data of a predetermined user are acquired; determining interest search words of a predetermined user according to the historical behavior data; searching interest search words matched with the received input words in all interest search words; and displaying the matched interest search word in the input association word list. According to the embodiment of the invention, the predetermined historical behavior data is acquired, the interest search term of the predetermined user is determined by combining the historical behavior data, and the interest search term matched with the input term isdisplayed in the input association word list; each user has the interest point of the user, through the embodiment of the invention, accurate personalized recommendation of thousands of people and thousands of faces can be realized, and the user experience is better.

The invention discloses an information recommendation method and device, is used for realizing unified personalized recommendation for users between an application program interface and a Web page, and relates to the technical field of the Internet. One specific embodiment of the method comprises the steps that an application program end locally creates Web service for communicating with a Web page in a browser, the Web page in the browser requests the local Web service through communication, obtains a user ID pre-stored in an application program and communicates with a recommendation servicemodule, and the recommendation service module is used for recommending the user ID to the application program end; a recommendation service module generates personalized information corresponding to the user according to the received user ID and returns the personalized information to the Web page for display. According to the embodiment of the invention, the consistency of personalized recommendation of the same equipment and different media can be realized in a user non-inductive manner.