43 results about "Preference weight" patented technology

A method of generating recommendations for content items comprises providing a domain ontology where concepts are characterized by a term vector with terms and associated weights. Associated term sets, each of which comprises a set of terms that characterize a content item, are further provided. A concept set is generated for each associated term set by determining the concepts of the domain ontology that match the terms of the associated term set. In addition, a user profile for a user is provided where the user profile comprises at least some of the concepts of the ontology coupled with preference weights. Recommendations for content items are generated based on the plurality of associated concept sets and the user profile. The invention may allow improved and / or facilitated generation of recommendations from text based characterizing data.

The invention discloses a personalized research direction recommending system and method based on themes. Paper topics read by users and preference of the users for related paper topics can be obtained through the recommending system according to all the papers read by the users and according to the themes of the papers obtained when training is conducted through a theme model training module, therefore, the recommending system can recommend a new research direction for the users to widen the vision of the users. The innovation key of the personalized research direction recommending system and method based on the themes is to construct a three- layer graph model according to the relationship between the users and the papers and the relationship between the papers and the themes, to calculate preference values of the users for the themes according to the three-layer graph model, to obtain a user-theme preference weight matrix, and to calculate similar user set between the users and other users based on the weight matrix. The preference degree of the themes which are not touched by the users is predicted according to the similarity value of the similar users in the similar user set and according to the preference values of the similar users for the themes, and the research direction, namely, the research theme, is recommended for the users according to the prediction result.

The invention provides a short text classification method and device. The method can comprise the steps that the probability that a short text relevant to a user belongs to each candidate category is determined; the preference weight of the user on each candidate category is determined according to the short text; and the category which the short text belongs to is determined according to the probability and the preference weight. Through the technical scheme, personalized short text classification operation can be realized, and therefore classification accuracy of the short text is improved.

The invention relates to a path planning method and a path planning device along a hand-drawn route in a navigation system. The path planning method comprises the following steps: firstly planning a navigation path according to the traditional method and in accordance with departure place and destination set by a user; if the user is dissatisfied for the planned path, drawing the hand-drawn tracks according to the movement of the finger of the user on a touch screen, then setting preference weight attribute of each hand-drawn track and storing the preference weight attribute; next, adjusting the passing capability weight of surrounding roads of each hand-drawn track, and planning a navigation path with largest passing capability according to the adjusted passing capability weight of the roads. The path planning method and the path planning device have the advantages that not only can identification errors generated by inaccuracy of gesture commands inputted by handwriting of the user in the existing navigation system be solved, but also the problem of unreasonable path planning caused by strict way and avoidance of certain interesting points in the existing navigation planning can be solved.

The invention discloses a television program content searching and recommending method oriented to integration of three networks. The method comprises the following steps: proposing three video tag self-enriching methods, namely, a synonym tag enriching method, a comment tag enriching method and a video similarity tag enriching method to generate a tag description file which can describe massive television program content in detail; designing an interface to completely collect explicit and implicit behavior information of television program audience, proposing explicit and implicit preference weight calculation methods respectively to construct a tag-based user preference model, and meanwhile considering the change of user preference along with time; and proposing a similarity calculation method based on a tag satisfaction relation to realize personalize search and active recommendation service of the massive television program content in an environment of integration of three networks. Through adoption of the method, the demand of detailed description on the massive television program content in the environment of integration of three networks is met. Meanwhile, a television program personalized service with better experience is provided for a user, so that the browsing time of the user is shortened, and the retrieval efficiency of the user is increased.

The invention discloses a preference program recommending device which comprises a user interface module, a tuner module, a preference analysis module and a recommending module, wherein the user interface module is used for receiving channel indication signals corresponding to television channels; the tuner module is used for receiving digital television signals, and the digital television signals comprise program table information and current program information corresponding to the television channels; the preference analysis module is used for analyzing current program keywords of the current program information so as to determine program types corresponding to the current programs, and accumulating preference weight values corresponding to the determined program types in a program type statistical table, wherein the program type statistical table is used for recording the program types and the preference weight values corresponding to the program types; and the recommending module is used for selecting recommended programs from the program table information according to the program types with higher preference weight values in the program type statistical table, and generating a program detailed list according to the recommended programs.

The invention relates to a device for recommending a program to an IPTV (Internet protocol television) terminal user, which is used for recommending a basic time period for watching a new program to the IPTV terminal user, wherein the basic time period has periodicity. The device comprises a user watch record database, a preference defining unit and a recommending unit, wherein the user watch record database is used for recording the program information of the programs watched by the user, and the program information includes the basic time periods of the programs watched by the user; the preference defining unit obtains the user watch record from the user watch record database, defines and calculates the preference weights of the programs watched in different basic time periods by the user, and sorts the preference weights according to the magnitude; and the recommending unit recommends the basic time period with high preference weight to the user to watch the new program according to the weight sorting result output by the preference defining unit. Through the invention, the functions of the IPTV service are expanded in a personalized manner, and the experience and satisfaction degree of the user are improved.

One embodiment of the application discloses an information push method and device. The method comprises following steps: determining a first class corresponding to a first operation behavior of a user on a client side; adjusting preference weight values of the user on the classes in a class set according to the first class; the class set comprises at least the first class, the preference weight values represent the preference degree of the user on the classes; recommending information to the client side according to the adjusted preference weight values. The embodiment of the present application improves the accuracy of the determined user personalized information and further improves the accuracy of pushing the recommendation information to the client side.

The invention discloses a picture pushing method, a mobile terminal and a computer readable medium. The picture pushing method comprises the steps of determining the types of preferred pictures according to preference weight values of pictures of each type; according to the types of the preferred pictures, searching for pictures of the corresponding type in a server, and generating a set by the searched pictures; according to the preference values of the pictures, sorting the pictures in the set according to a sequence of the preference values from high to low; and pushing the whole sorted set, or pushing the pictures in the sorted set in sequence. According to the technical scheme, targeted pushing can be performed according to user preferences, so that the quality of the pushed pictures is effectively ensured and the user satisfaction is improved.

The invention discloses a user grouping method. The method is carried out through a calculating device which is connected to a data storing device, wherein the times of each user connecting to each wireless network, and the preference weight of each user to each label are stored in the data storing device. The method comprises the following steps: determining the space similarity of every two users according to the times of each user connecting to each wireless network; determining the property similarity of every two users according to the preference weight of each user to each label; determining the group similarity of every two users based on the space similarity and the property similarity; determining the group feature vector of each user based on the group similarity of every two users; clustering the users based on the group feature vector of each user; and dividing the users into a plurality of groups. The invention also discloses a corresponding calculating device.

The present invention provides a quantitative trading strategy generation method based on multi-objective optimization, a device, a device, and a storage medium. The method includes: selecting target factors generated by the strategy; scoring each target factor on interval samples, and standardizing the scores , so that the score of the interval sample is in the defined interval; form the objective function, the objective function is composed of the objective factor and its corresponding weight, and the weight includes multiple groups; use the primary strategy to backtest the objective function, and generate the backtest result; select the expected Generate multiple groups of multi-objective weights based on the backtest results; select weights and adjust them according to investor preferences, and output optimal strategies. The present invention selects multi-target factors to form a quantitative strategy according to investor preferences, and further optimizes the strategy according to the preference weight among multiple groups of weights, so as to better reflect investors' investment preferences; evaluate the backtest results of the strategy by selecting a high score of a single target factor , reducing the number of strategies and the corresponding backtest running time, reducing the burden on system resources.

The invention relate to a shortest route planning and navigating method. The method comprises the following steps: step one, according to the user preference, adding a user preference weight to a route distance weight, and using a formula: M(n)=P(n)+Q(n)*A to preprocessing the data to obtain a comprehensive evaluation weight M(n); step two, selecting a special node that a user has to pass by, taking the special node as the cutting point, cutting the road network into a first network and a second network by the cutting point, wherein the first network is a node network from the start point to the special node, and the second network is a node network from the special node to the final point; step three, replacing the route distance weight P(n) in Dijkstra algorithm by the comprehensive evaluation weight M(n) obtained in the step one, and then using the improved Dijkstra algorithm to carry out shortest route on the first network and the second network so as to obtain the shortest route from the start point to the special node to the final point.

The invention discloses a method and device for providing commodity information. The problem that server pressure is large in the prior art is solved. According to the method, with respect to each commodity category, the first moment of user ordering under the commodity category at the last time is determined, the second moment for clicking the commodity information under the commodity category by a user for the first time is determined within the time period from the first moment to the current moment, a valid time interval is determined according to the time interval from the second moment to the current moment, the preference weight number of each commodity category is determined according to the valid time interval and an ordering expected time interval, and the commodity information with the preference weight numbers being larger than a set threshold value under each commodity category is provided. By the adoption of the method, through the time period from the moment for clicking the commodity information under each commodity category by the user to the current moment, the preference degree of the user to each commodity category is quantified, the commodity recommending accuracy is improved, the frequency that the user uses a search function is effectively reduced, and the server pressure is lightened.

In various implementations, routing factors are identified based on a routing request associated with a user, where the routing factors include route preferences of the user. Routes are generated based on the routing request. Preference weights are determined for the route preferences, where the preference weights correspond to machine learning models based on sensor data provided by one or more sensors in association with the user. Route scores are determined for the routes based on the preference weights. A suggested route is provided to a user device associated with the user, where the suggested route corresponds to a selected route of the routes and is provided based on the route score of the selected route.

The invention discloses a social network user dynamic and static interest mining method comprising the following steps that background information profile and generation content of a user are acquired from social media; the static interests SI={SI1, SI2,...,SIm} are extracted from the background information profile of the user, each interest point SIi is a binary group SIi=(kwi, wi), and 1<=i<=m, wherein the kwi is the keyword, and wi is the preference weight of the user for kwi; and dynamic DI={DI1, DI2,...,DIn} is extracted from the generation content of the user, each interest point is a triple group DIi=(topici,wi,T), and 1<=i<=n, wherein topici is formed by multiple keywords, wi is the preference weight of the user for topici, and T={t1,t2,...,ts}, ti (1<=i<=s) is each time point of discussing topici of the user, i.e. ti (1<=i<=s) refers to the distribution situation at different time points. According to the method, the characteristics of the interests of the social media users can be more reasonably described so that the method is more suitable for subsequent deep analysis of the characteristics of the interests of the social media users.

The invention relates to the technical field of big data, and in particular to a user behavior statistical method and device and terminal equipment. The user behavior statistical method comprises thefollowing steps of: carrying out statistic on behavior data of a user; filtering behavior data which does not satisfy a preset condition in the behavior data so as to obtain target behavior data; calculating a plurality of index weights according to the target behavior data; and then calculating an interest preference weight of the user according to the plurality of index weights and determining an interest preference of the user. According to the method, the interest preference of the user is calculated according to the behavior time of the user in a period of time, so that the interest preference change of the user can be better reflected, and personalized recommendation can be carried out on the user according to the current interest preference of the user, thereby improving the user experience.

The invention provides a user profile creating method based on user requirements. The method includes the steps that 1, in the searching process of a user, behavior features for returned pages of the user and page feature vectors of pages clicked by the user are extracted, wherein the page feature vectors comprise features reflecting the layout of the pages, features representing the reading difficulty of the pages and page data volumes; 2, based on the feature vectors of the multiple pages clicked by the user, a user profile of the user is constructed, wherein according to the behavior features for all the pages of the user, user preference weights of the pages are set. The invention further provides a corresponding personalized search ranking method and system. The personalized search ranking method and system can be sufficiently adapted to preference for different features of the user, improve the accuracy of a ranking result and improving the judgment precision of the satisfaction degree of the user.

The invention discloses a desktop icon layout method and apparatus, and a computer readable storage medium. The method comprises the following steps: collecting use frequency and residence times of application programs within a certain time period, and performing analysis to obtain user preference weights of the application programs in a terminal; traversing the application programs in the terminal, and performing object description on related information of the application programs, and constructing a global relationship network among the application programs by using a directed hypergraph; and obtaining a layout relationship of desktop icons of the application programs on the terminal desktop and recommending the layout relationship to the user. According to the desktop icon layout method, the global relationship network GAN model is constructed, meanwhile the preference of the user on each application program is obtained, the global relationship and the personalized relationship among the application programs are imported into the desktop icon layout of the application programs, the desktop icon layout mode better satisfying the operation habits of the user is recommended, and the user satisfaction degree is improved on the layout effect.

The invention discloses a traditional Chinese medicine composition for treating arthrolithiasis. The composition is prepared from, by weight, 25-40 parts of the extract of sunflower alkali, 15-25 parts of the extract of chicory, 15-30 parts of the extract of the seed of Job's tears, 10-15 parts of the extract of poria cocos, 3-5 parts of the extract of Chinese angelica, 1-5 parts of the extract ofpallas pit viper, 5-12 parts of hawthorn fruit powder and 5-12 parts of pawpaw powder; the preference weight parts of raw materials are 30 parts of the extract of the sunflower alkali, 20 parts of the extract of the chicory, 20 parts of the extract of the seed of Job's tears, 11 parts of the extract of the poria cocos, 4 parts of the extract of the Chinese angelica, 3 parts of the extract of thepallas pit viper, 7 parts of the hawthorn fruit powder and 7 parts of the pawpaw powder. The composition has the advantages of inhibiting the synthesis of purine, reducing purine trione, dissolving calculus uric acid, repairing damaged liver and kidney cells, increasing human body immunity and has significant meaning on the health of a arthrolithiasis patient.

A system and method of routing transactions within a computer network, by at least one processor may include: receiving a transaction request to route a transaction between a source node and destination node of the computer network; identifying a plurality of available routing paths for propagating the transaction between the source node and destination node; obtaining transaction parameters for each available routing path; receiving a set of preference weights including one or more preference weights, each corresponding to a transaction parameter; selecting one or more routing paths from the plurality of available routing paths, based on the one or more transaction parameters and respective preference weights; producing a routing scheme, comprising an ordered list of the one or more selected routing paths, according to the received set of preference weights; and routing the requested transaction through nodes of the computer network according to the routing scheme.

The invention discloses a method for personalized cloud recommendation in a multi-cloud platform environment. Firstly, property attention priorities of a user are constructed into preference paired-comparison matrixes, the property priorities of cloud platforms are constructed into property paired-comparison matrixes, further inconsistent elements in inconsistent columns are amended by identifying inconsistent columns and contradicted columns in all the paired-comparison matrixes, the contradicted elements in the contradicted columns are corrected, the paired-comparison matrixes meeting the consistency requirements are obtained, finally preference weight vectors of the user to the cloud platforms are calculated by utilizing an analytic hierarchy process, the cloud platforms are recommended to the user based on the weight vectors, and accordingly a cloud recommendation scheme meeting the personalized requirement of the user in the multi-cloud platform environment is achieved. The method effectively improves the efficiency and accuracy of a cloud recommendation system and makes the system have better practicability.

The invention relates to a correlation evaluation method for Internet of Things object information searching and sorting. The method comprises the following steps: 1) establishing a user personalizedinterest model based on user historical query information; 2) establishing a quality scoring mechanism of the article based on the historical query information of the user; and 3) constructing a tripartite graph based on the two contents, and obtaining a correlation score through a random walk algorithm. The method has the beneficial effects that the personalized interest model of the user is mainly established, the quality scoring mechanism of the article attributes is established according to the historical query information of the user, and the correlation evaluation model of the candidatearticles is established. After the model is established, a user-article-article attribute tripartite graph is constructed based on the preference weight of the user and the article attribute quality score; correlation scores of the target user and different candidate items are obtained through a random walk propagation algorithm according to the item attribute tripartite graph, and correlation evaluation of the candidate articles is achieved.

The invention discloses a sequential smooth community partitioning method based on label diffusion. The method comprises the following steps: 1) obtaining label values of each node in the network to be partitioned at the preceding time of the current time; 2) judging whether that oscillation numb is greater than a preset threshold value; 3), if not, acquiring the first preference weight value of the node, taking the first preference value as the current preference value of the node; 4) if yes, acquiring a second preference weight value of the node, taking the second preference value as the current preference value of the node; 5) updating that label value of the node at the current time and returning to execute the step 2) until each node in the to-be-divided network is grouped into a community composed of nodes with identical label values. The embodiment of the invention also provides a timing smoothing community dividing device based on label diffusion. By applying the embodiment ofthe invention, the accuracy of community dividing can be improved.

The invention relates to a query expansion method based on a user query association degree. The method comprises the following steps of: step 1) making statistics on the total number of web pages browsed by a user; step 2) dividing interest types; step 3) making statistics on the number of the interest types contained in the web pages browsed by the user and the number of the web pages; step 4) calculating interest preference weights of the user to the interest types; step 5) constructing a user preference model of the user; step 6) constructing a plurality of user preference models with methods in the steps 1) to 5); step 7) forming multiple user groups; and step 8) forming a search space by document click information of all users in the user group. According to the method, users closely related to user interests are taken as background knowledge and bases of query expansion, and based on this, an associative relationship among query words is analyzed by utilizing associative query, so that the problem of low primary retrieval quality is solved and the retrieval efficiency is improved.

The invention provides a family member attribute prediction method based on a subject model, a device and an intelligent terminal. The method includes: obtaining a probability distribution of each video in the media asset library on each topic; according to the user's preference weights for each video and the probability distribution of each video on each topic, the user's topic preference probability distribution is generated. Using the probability distribution of user's topic preference as the input feature of the logistic regression model, the probability distribution of various user's attributes output from the logistic regression model is obtained. According to the probability distribution of various user attributes, the number of attributes conforming to the salient attribute characteristics is counted, the number of attributes conforming to the salient attribute characteristics is confirmed as the number of family members, and the user attributes corresponding to each family member are recorded. The family member attribute prediction method provided by the present application can greatly improve the family member attribute characteristic coverage, and effectively enrich thebasic attribute dimension of a large screen user portrait crowd.

The invention provides a desinsection composition which comprises nitenpyram and carbamate pesticide, wherein the weight ratio of the nitenpyram to the carbamate pesticide is between 1 to 1 and 1 to 100; the preference weight ratio of the nitenpyram to the carbamate pesticide is between 1 to 10 and 1 to 50, and the deinsection composition is mainly used for preventing and controlling homopteran pests. The desinsection composition has the advantages that the components are reasonable so that the preventing and controlling effects are obvious, the quick action is good and the persistent period is long; and the cost is saved, the insecticide resistance of pests is delayed, and the pollution to the environment is reduced.

A system and method of routing transactions within a computer network, by at least one processor, including: receiving a transaction request to route a transaction between one of a plurality of source nodes and a destination node of the computer network; extracting from the transaction request one or more transaction parameters pertaining to the destination node; receiving a set of preference weights wherein each preference weight corresponds to a transaction parameter; selecting a source node from the plurality of source nodes based on at least one received preference weight and at least one corresponding transaction parameter; and routing the requested transaction through nodes of the computer network between the selected source node and the destination node.

The invention discloses a method and a device for route recommendation, and relates to the technical field of computers. A specific implementation of the method includes: analyzing the geographic locations of the starting point and the ending point, and determining matching road sections corresponding to the starting point and the ending point respectively; determining a comprehensive preference weight according to a preference learning component; Preference weights to generate recommended paths. In the present application, the technical means of recommending the route by first determining the section of the starting point and the end point corresponding to the recommended route and the corresponding preference, avoids the technical defects of unavailability, high cost, and disregarding the user's preference of the route recommended by the prior art, thereby making the recommended route possible. The route is more in line with the personalized needs of the user, and greatly saves the technical effect of the cost of route recommendation.

The invention provides a guided symbol execution method based on static path analysis, which can quickly reach the position of the specified code, and provide the test input to reach the target code. Step 1: First, determine the detection target through static analysis, and find the code or instructions related to memory allocation in the target program; Step 2, obtain the path-related functions and basic blocks from the program starting point to the specified code point, and determine the path accessibility according to the path. Calculate the weights of these functions and basic blocks with distances and distances; step 3, preferentially select the static analysis reachable paths with large weights to traverse; step 4, static path analysis; step 5, correlate the paths obtained by static path analysis Function and basic block information are placed in Paths[funcs, bbs], all target instructions to be analyzed are stored in the Vector of allocCallInst, and all current path states are stored in states. When the symbolic execution engine executes the analysis on the target program, through the loop Interpret and execute the instructions in the program one by one.