131results about How to "High degree of personalization" patented technology

The invention provides an information recommendation method and device based on artificial intelligence, electronic equipment and a computer readable storage medium. The method comprises the steps ofobtaining a historical behavior information sequence of a user; according to the vector representation of each piece of information in the historical behavior information sequence, determining a behavior feature vector representing the behavior feature of the user; iteratively encoding the behavior feature vector for multiple times to obtain a user interest vector representing the interest of theuser; and determining a plurality of pieces of information satisfying similar conditions with the user interest vector to execute recommendation operation for the user. According to the invention, therecommendation accuracy and the click rate index of the recommendation information can be improved.

Embodiments of the invention provide a method and a device for human-computer interaction. The method comprises the following steps of determining emotional color of a user according to Interactive information input by the user, and carrying out resource search; determining an answer sentence pattern corresponding to a search result from a preset sentence pattern according to the search result; and generating an answer statement according to the search result, the answer sentence pattern and the emotional color. According to the method and the device for human-computer interaction, which are provided by the embodiments of the invention, the flexible and individual answer is provided for the user in human-computer interaction, and the interactive experience of the user is improved.

The invention relates to the field of gynecological medical apparatus and instruments, in particularly to an implant for uteruses. The implant for the uteruses comprises a membrane (1). The implant for the uteruses has the advantages that problems of irritation influence on uteruses due to existing implants implanted in uterine cavities for a long time, poor uterine mucous membrane repair effects of conventional medicine, influence on the whole bodies of patients due to hormone medicine for oral administration and the like can be solved by the aid of the medicine to a great extent, and effects and individual degrees can be greatly improved during phased utilization.

The invention discloses a network electronic album display method and system. The network electronic album display method comprises A, receiving photos uploaded by a user and obtaining photo parameters, and calling stored album templates according to the photo parameters and album template matching parameters input by the user; B, adding the photos into all the called album templates and generating a preview effect album; C, generating an electronic album according to the results selected by the user in the preview effect album. According to the network electronic album display method and system, parameter matching is firstly performed to screen the album templates, so that the template needed by a user can be obtained quickly; a user can quickly find the satisfactory album template through preview effects, the trouble caused by the previous act of adding photos in templates one by one for selection can be avoided, and the network electronic album which the user is satisfied with can be generated quickly. Besides, the album template of the electronic album can be changed at regular time, so that the personalized degree of the network electronic album can be improved, and the user experience can be further improved.

The invention discloses a recommendation reason generation method, which comprises the steps of obtaining at least one recall result for search data according to the search data of a target user; according to the search data of the target user, the recall results and the portrait of the target user, obtaining recommendation reasons of the recall results through a preset intelligent question and answer model; wherein the intelligent question and answer model is a first machine learning model obtained by training at least one sample question and answer data combination; wherein the sample question and answer data combination comprises a sample user portrait and historical evaluation data of at least one sample user, a recommendation object corresponding to the historical evaluation data, andhistorical search data of the sample user for the recommendation object. The technical problems of high cost and insufficient individuation are solved. The method and the device have the beneficial effects of reducing the recommendation reason generation cost and improving the individuation degree of the recommendation reason.

The invention relates to a case-reasoning-based personalized recommendation method for an E-commerce website commodity. An objective problem is constructed according to a feature and an access behavior of a current access user and an original case is established according to features, historical access behaviors, and purchase records of other users; a fuzzy clustering way is used for carrying out clustering according to a target case feature, cases with high similarity levels of user features and interests and the like are classified into the same class and the class is identified, and a high-similarity-level case base is established; and a resolved case set most similar to the target user case is recommended by using a user case similarity retrieval and an improved fuzzy mean clustering algorithm and then is corrected and reused, and a dynamic case base is also updated. According to the invention, the system real-time performance is enhanced; and a problem that the previous recommendation system relies too much on commodity evaluation by users can be solved.

The invention provides an anti-fake structure and a production method of the same. The anti-fake structure comprises an anti-fake characteristic layer and N colored layers which are stacked vertically, wherein N exceeds or is equal to two; the different colored layers have mutually different colors; the anti-fake characteristic layer comprises one or multiple first areas which at least penetrates the anti-fake characteristic layer; the first areas randomly penetrate the M colored layers, and M is an integer between 0 and N and contains 0 and N; the different first areas penetrate different quantities or the same quantity of the colored layers; and the first areas solely constitute anti-fake information and/or constitute the anti-fake information together with areas besides the first areas on the anti-fake characteristic layer. The anti-fake structure and the production method of the same provided by the invention can enhance randomness of the anti-fake information and anti-fake reliability by designing a physical structure.

The invention discloses a mobile terminal desktop icon adjusting method and a mobile terminal desktop icon adjusting device. The method comprises the following steps that a desktop icon adjusting mode is entered through an pre-concerted interactive mode; the position of a selected icon on a desktop is adjusted and the icon is defined as a novel icon, wherein the position of the icon is adjusted; vectorization processing is conducted to the novel icon; the size of the novel icon which undergoes the vectorization processing is adjusted; the novel icon is displayed on the desktop, wherein the size of the novel icon is adjusted. The mobile terminal desktop icon adjusting method and the device can improve display attractiveness degree and individuation degree of the desktop icon due to the fact that the desktop icon implemented by the vectorization processing and the size and the position of the desktop are adjusted. The desktop icon is closer to life. The experience of users is improved.

The invention discloses a method for generating software service in network, according to the need of user. It is characterized in that: user via user terminal refers service request; the server terminal analyzes the user demand, and extracts needed function and relative software elements, to attain the information of all elements to be sent to the user terminal; user checks the feedback element information, selects needed element, and uses the graphics element provided by user terminal to connect the interfaces of said elements according to demand; the user terminal forms XML document based on the connecting project of user to be sent to server terminal; the server terminal according to said connecting project generates assemble operate script, extracts unite packing element, and expresses and executes the operation script to complete the software server assembly; at last, feeding generated server program back to user. The invention can realize the projection from the user demand to element, to generate relative software and realize the software server according to demand.

The invention relates to a method for manufacturing a oral prostheses stent formed by compounding moulding of additional and subtractive materials, which comprises the following steps: 1) acquiring a three-dimensional model data of a patient's oral maxilla: directly scanning a patient's mouth through a scanning device, and directly obtaining three-dimensional point cloud data of the patient's mouth, and exporting the 3D point cloud data by the scanning software to output a triangular mesh model; 2) obtaining a CAD model of the final printable manufactured stent; 3) using a 3D printing solid stent using an additive manufacturing technology; 4) transferring the prostheses stent to-be-processed from a SLM additive manufacturing machine to a machining tool, and taking out the prostheses stent with a substrate, a support and positioning marks manufactured in step 3) from the SLM machining tool; and 5) obtaining the manufacturing coordinate system of the additive by mechanical or visual positioning method, and performing numerical control programming according to a coordinate conversion relationship between a coordinate system and a machining coordinate system to perform the subtractive material manufacturing process on the prostheses stent. The method has high processing precision and wider application range.

The embodiment of the invention discloses a text label extraction method and device, equipment and a storage medium. The method comprises the steps of obtaining each text of a to-be-extracted label, and vectorizing each text to obtain a text vector corresponding to the corresponding text; clustering the text vectors to obtain at least one text clustering result; carrying out keyword extraction on each text clustering result to obtain each label candidate word corresponding to each text clustering result; determining a text label of each text according to each label candidate word corresponding to each text clustering result. By means of the technical scheme, automatic extraction of the text label is achieved, and the accuracy and comprehensiveness of text label extraction and the expandability of the label extraction method are improved.

The invention discloses a method and a device for acquiring house type transformation information. The method comprises the following steps: after acquiring current capability feature information of auser by utilizing current health capability information of the user, searching first target house type transformation information corresponding to the current capability characteristic information and corresponding transformation prediction accuracy in the first learning model, obtaining second target house type transformation information corresponding to the current health ability information and corresponding transformation prediction accuracy according to a second incidence relation in the second learning model, the transformation prediction accuracy of the corresponding house type transformation information and the current health ability information; when the first target house type transformation information is different from the second target house type transformation information, determining the target house type transformation information corresponding to the maximum transformation prediction accuracy as recommended house type transformation information; otherwise, determiningthe first target house type transformation information or the second target house type transformation information as recommended house type transformation information. According to the method, the personalized degree of the recommendation scheme and the recommendation accuracy are improved.

The invention discloses a resource recommendation method and device, computing equipment and a storage medium. The resource recommendation method includes the steps that a data model for resource recommendation is trained by using resources adopted by a user, a scoring model of the user for to-be-recommended resources is obtained; the scoring of the user for the to-be-recommended resources is calculated based on the scoring model; the to-be-recommended resources are sorted according to scoring orders from high to low; and the predetermined amount of resources sorting highest are recommended tothe user. The resource recommendation method and device based on a relief network is used, the maximum success rate of the recommended resource sets adopted by the user is ensured, and accordingly the individuation degree of the recommendation result is improved.

The invention discloses a movable-robot-type liquid metal printing device and method. The device comprises a control unit, a driving unit, a moving unit, an attaching unit and a printing unit. The working method comprises the following steps: 1, the control unit analyzes the liquid metal route required to be printed and sends a control instruction to the driving unit; 2, the driving unit controls the moving unit and the attaching unit, so that the whole device moves to a designated position of space and stably attaches to a working surface; 3, the printing unit prints the liquid metal or a packaging material on the surface of the working surface by automatically adjusting a spraying head; and 4, the device can also move according to instructions to realize multi-area continuous printing. The device can automatically move to the working surface at different positions of the space to directly carry out printing and packaging of the liquid metal route, so that various limitations of a fixed device on printing size, position and the like are broken, and the rapid manufacturing efficiency and the personalized degree of the liquid metal electronic route are further improved.

The invention provides an indoor environment intellectualization dynamic control system and method thereof. The indoor environment intellectualization dynamic control system comprises a control subsystem, a data collection subsystem, a weather subsystem, an adjusting subsystem and a human-computer interaction interface, wherein the data collection subsystem, the weather subsystem, the adjusting subsystem and the human-computer interaction interface are correspondingly in communication connection with the control subsystem. According to the technical scheme, the indoor environment intellectualization dynamic control system is simple and convenient to implement, has the good control effect, and has the significant energy saving effect. Micro climate data are collected by the weather subsystem and are compared with indoor parameters collected by the data collection subsystem. The control subsystem obtains various comfort parameters of the environment through analysis, the adjusting subsystem adjusts the parameters of indoor temperature, humidity, ventilation and the like, and provides the most comfortable environment for indoor residents; real-time control and adjustment can be achieved, the comfort degree is guaranteed, and unnecessary energy consumption is avoided; each household and each room are adjusted separately, space is small, and adjusting efficiency is high.

The invention discloses a resource scheduling method and system and a computer readable storage medium. The method comprises the steps of obtaining a control instruction input by a user; determining intention equipment based on the control instruction; under the condition that the intention equipment cannot be started, for each function of the intention equipment, inquiring alternative equipment associated with the function from a constructed knowledge graph, and storing a plurality of functions and equipment associated with each function in the knowledge graph; and determining and starting the to-be-started equipment from all the alternative equipment. By applying the resource scheduling method provided by the invention, under the condition that the intention equipment specified by the user cannot be started, the alternative equipment is reasonably determined according to the association relationship between the functions and the equipment in the constructed knowledge graph, and thenthe equipment to be started is determined and started from the alternative equipment. The problem that resources cannot be scheduled in a personalized mode in the prior art is solved, the personalizeddegree of the system is improved, the scheduling requirements of different resources are met, and the user experience is improved.

The invention discloses a personalized federal learning method based on a multi-head attention mechanism. The method is characterized in that the method comprises the following steps: 1, building a federated learning local model multi-head attention mechanism model: enabling the multi-head attention mechanism to be used in a most classical convolutional neural network, retaining key information through the multi-head attention mechanism, and carrying out feature extraction and selection better to improve the recognition accuracy; and 2, building a federal learning global model multi-head attention mechanism model: considering the influence of the personalized problem of each local model on the global model, and making corresponding changes according to the personalized characteristics of the model when global model parameters are sent to the local models. The correlation of extracted features and the data difference brought by reasonable individuation of each client can be considered, so the individuation degree of the data can be increased on the basis of ensuring the improvement of the accuracy.

The invention relates to an image processing method and device, electronic equipment and a computer readable storage medium. The image processing method comprises the following steps: acquiring enhanced preference information of a user; obtaining a to-be-processed image, and dividing the to-be-processed image into a plurality of to-be-enhanced areas according to the hue distribution of the to-be-processed image; and performing corresponding color enhancement processing on each to-be-enhanced area according to the enhancement preference information to adjust the color attribute of the to-be-processed image. The to-be-processed image is divided into the plurality of to-be-enhanced areas based on the hue distribution, and corresponding color enhancement processing can be respectively executed for different hues, so that the problem of color distortion is effectively avoided in the process of realizing color enhancement. Moreover, in combination with the acquired enhancement preference information of the user, a personalized and accurate color enhancement effect can be provided for the preference of the user, so that the image processing method with high personalized degree and high color rendition degree is realized.