32 results about "Attribute weight" patented technology

Apparatuses, systems, and techniques to cause weights assigned to properties of chatbot answers to be dynamically adjusted. In at least one embodiment, one or more neural networks are used to characterize one or more chatbot queries and cause weight values assigned to properties of answers to the one or more chatbot queries to be dynamically adjusted, based, at least in part, on the characterization of the one or more chatbot queries.

The invention relates to the technical field of intelligent scheduling, and discloses an intelligent scheduling control method and system based on multi-objective optimization, and the method comprises the steps: obtaining an attribute set of different tasks, obtaining a disturbed task attribute initial value, and carrying out the calculation to obtain an importance score of the tasks during initialization; based on the task attribute initial value and the attribute weight, calculating the conflict degree between the tasks through the task scheduling conflict degree; defining an optimization target according to the importance score of the task and the conflict degree between the tasks, and automatically generating an association rule; and after optimization, calculating the task importance score of the final time step based on the task attribute value and the attribute weight of the final time step, and generating a final scheduling scheme according to the final importance score of each task. According to the method, through dynamic conflict evaluation, adaptive weight adjustment and a closed-loop feedback mechanism, the environmental adaptability, global optimization and stability of a scheduling system are remarkably improved, and resource competition and time delay are effectively reduced.

The invention provides a multi-attribute score fusion recommendation method based on user preferences and attention, and the method comprises a data collection and preparation module which is used for obtaining and cleaning user, project and score data; the attribute coding module is used for carrying out numerical coding on user and project attributes; the multi-attribute feature fusion module is used for obtaining user-project comprehensive representation by splicing attribute vectors and utilizing a deep neural network to extract interaction features; the fusion attention mechanism module is used for dynamically distributing each attribute weight and realizing weighted fusion of multi-attribute features; the comprehensive evaluation score calculation module is used for predicting the score of the user on the project according to the fusion feature and the attention distribution; the global preference modeling module is used for fusing historical behaviors of the user and improving the modeling capability of the overall preference; the recommendation result output module is used for generating a personalized recommendation list; and the method evaluating and updating module is used for evaluating recommendation performance and dynamically optimizing the model.

The invention discloses a big data measurement asset supply and demand matching and inventory optimization method, and relates to the technical field of big data processing and supply chain management. The method comprises the following steps: segmenting supply chain asset data through a hash function according to multi-dimensional attributes to generate a data fragment set, and constructing a global index tree according to the data fragment set; and monitoring the load state of the leaf nodes of the global index tree in real time, and migrating and verifying data during unbalance. And querying the global index tree, positioning fragment data associated with the query request from the uniformly distributed leaf nodes, generating a preliminary matching asset set, calculating the matching degree of each asset and the query request, and generating a resource allocation result. And updating the global index tree, and generating the latest data view representation. And adjusting the attribute weight coefficient of each dimension in the hash function, and generating optimized storage layout configuration. Irrelevant data are filtered through a neighbor algorithm, and a final matching result set is generated. Balanced storage, efficient query and accurate matching of asset data are realized, and the overall response speed and the resource utilization rate are improved.

The invention provides an optical cable fault rapid positioning method, system and equipment, and the method comprises the steps: obtaining multi-dimensional attribute parameters of a target optical cable section, specifically including a static attribute, a dynamic attribute and an environment attribute for evaluating the operation state of an optical cable; inputting the multi-dimensional attribute parameters into the weighted analysis model and the trained machine learning prediction model in parallel; wherein the weighted analysis model calculates and outputs a first candidate fault position set based on an attribute weight data set, and the attribute weight data set comprises weight scores defined for different types of attribute parameters; analyzing the current multi-dimensional attribute parameters by using the trained machine learning prediction model, and outputting a second candidate fault position set; and fusing overlapped data of the first candidate fault position set and the second candidate fault position set to generate comprehensive fault positioning information. Based on the method, the invention also provides an optical cable fault rapid positioning system and equipment. According to the method, the accuracy of fault positioning and the decision robustness are remarkably improved.

The invention relates to the technical field of intelligent locks, in particular to an intelligent lock with an intelligent acquisition function and a control method thereof.Importance levels of corresponding acquisition functions are represented through attribute weights, then the attribute weights are corrected through data calling records, and personalized weights better conforming to user habits are obtained; then a fitness function is determined based on the personalized weight, so that an optimal acquisition function calling strategy can be found through an optimization algorithm, and the calling strategy not only can select and reject a plurality of acquisition functions according to the residual electric quantity so as to prolong standby time, but also can match user habits and reserve common and important acquisition functions of a user; the purpose of balancing energy consumption and functions of the intelligent lock is perfectly achieved, it is ensured that safety and convenience are provided, high efficiency and energy efficiency are kept, user experience is greatly improved, and good prospects are achieved.

The invention relates to a large-scale heterogeneous multi-attribute equipment scheme clustering decision-making method and a medium. The method comprises the following steps: uniformly converting heterogeneous attribute data in a digital twin decision-making information system into binary connection numbers; determining an attribute weight by adopting an MEREC method, and constructing an attribute non-compensable value based on the attribute weight and the binary connection number; generating a scheme similar network according to the attribute non-compensable value; performing network clustering operation on the scheme similar network, and outputting a scheme clustering result; and based on the scheme clustering result, calculating the comprehensive target distance of the scheme in the cluster by adopting the grey target theory, and fusing the global value of the scheme and the comprehensive target distance to generate a final scheme score. According to the method, the problems of heterogeneous information fusion distortion, attribute non-compensatory mechanism deficiency and low large-scale scheme clustering efficiency in a digital twinborn decision scene are solved by constructing heterogeneous data unification-attribute non-compensatory mechanism modeling-network clustering-intra-cluster decision.

The application provides a material similarity analysis method based on a semantic library and a knowledge graph, comprising the following steps: S1: judging whether the decisive attributes of two materials are similar, if any of the decisive attributes is not similar, then the two materials are non-redundant materials, otherwise, entering S2; S2: calculating the similarity of the field part according to the non-decisive attribute weight and the distance of each field; S3: combining all the knowledge graph paths and weight values to calculate the similarity of the graph part; S4: adding the similarity of the field part and the similarity of the graph part to obtain the final material similarity, and judging whether the two materials are high similarity through the final material similarity. The material similarity analysis method provided by the application optimizes the accuracy of the similarity algorithm.

The invention discloses a zero sample fault detection method based on weighted semantic consistency embedding, and belongs to the field of industrial fault diagnosis. The method comprises the following steps that firstly, industrial equipment is used for collecting multi-mode sensor data, and standardized data are obtained through normalization, wavelet denoising and time sequence segmentation preprocessing; determining a fault attribute dimension based on domain expert knowledge, assigning each fault class to form a semantic attribute vector, and constructing a semantic attribute weight matrix through mutual information gain between attributes and fault tags; embedding data and semantic attributes into a shared space by using a data encoder and a semantic attribute encoder, decoding and calculating reconstruction loss by combining a data decoder and a semantic attribute decoder, and introducing weight to construct weighted semantic consistency loss; combining the two types of loss to train an encoder and a decoder, and extracting training data features by using the trained encoder and training a multi-attribute classifier; and finally, in a zero sample scene, extracting embedded features without fault data, inputting the embedded features into an attribute classifier, and matching fault categories through a maximum posterior probability. According to the method, the equal-weight hypothesis defect of an existing semantic consistency embedding method is broken through, key semantics are strengthened through mutual information gain weights, consistency loss is weighted to resist noise interference, and the precision, robustness and generalization ability of zero sample fault diagnosis are remarkably improved.

Embodiments of this application disclose a node selection method performed by a first node in a blockchain. In the embodiments of this application, nodes in a blockchain are grouped in an order of attribute weights of the nodes, to obtain a plurality of node groups. A node group whose attribute weight meets a preset condition is selected from the plurality of node groups, to obtain a target node group, and a preset quantity of candidate consensus nodes are selected from the target node group. A non-candidate consensus node in the target node group is stored into an initial node group. The initial node group is used as a new target node group, and the operation of selecting a candidate consensus node from the target node group is performed. Candidate consensus nodes are broadcast to one or more second nodes in the blockchain for verification.

The application discloses a kind of based on multiple confidence medical data missing value batch repair method, the method includes: using attribute weight updates sample confidence, missing sample set is introduced into model training;Using sample confidence optimization loss function, fill the missing value of data set;The sample confidence is calculated according to the attribute relationship between samples, according to the data attribute to be predicted and the number of missing values in sample, different confidence is given to sample;The influence degree of sample in model training process is adjusted by dynamic selection of confidence.Optimize model architecture, so that the network can fill the missing value of multidimensional data in batches at a time, can also eliminate the identity mapping problem of network transfer function, and enhance the intercorrelation between nodes.The application improves the utilization of data, improves the filling accuracy and filling efficiency of data.

The invention provides a PLC isomer dynamic access control method based on attribute identification, which solves the problem that the access authority of an industrial PLC system is difficult to dynamically control in real time in a complex heterogeneous environment, and remarkably improves the safety protection capability and the control efficiency of the system. Firstly, attribute data such as user identities and device types are collected for attribute importance evaluation and weight preprocessing, and attribute weight distribution is achieved; secondly, constructing a deep trust model based on a full-connection neural network, training the model according to the preprocessed data, outputting a dynamic trust level, and executing a hierarchical access decision; and finally, designing a neural network self-organizing mechanism, pruning and optimizing redundant nodes of the network by calculating comprehensive importance scores of neurons, and realizing dynamic simplification of a model structure. According to the method, the security access level of the industrial-grade PLC can be calculated through different attribute data, so that a necessary basis is provided for the security operation control of the PLC.

A price analysis method and system based on a multi-attribute template, comprising: constructing a multi-attribute template, generating attribute weights and coding rules through association rule mining, and fusing commodity descriptions and business fields into attribute vectors; calculating similarity based on weighted Euclidean distance, normalizing attention weight through Softmax, and weighting and summing historical prices to obtain reference price; calculating deviation rate and comparing with threshold value to perform hierarchical disposal; constructing a knowledge graph to detect associated abnormalities and price avoidance paths; and implementing batch backtracking based on a version-identified feature cache table. The present application realizes fine-grained price analysis and hidden abnormality mining, and supports dynamic template updating, coding rule adaptive calibration and multi-level rollback mechanisms, significantly improving system robustness and cold start capability.

The invention discloses an intelligent automatic distribution method, system and device for college paper marking experts and a medium. The method comprises the steps of calculating a first attribute weight, a second attribute weight and a third attribute weight; inputting the historical data corresponding to each fourth attribute into a linear regression equation to obtain a fourth attribute weight of each fourth attribute; selecting a preset number of target experts; based on the multiple first attributes, the first attribute weights, the multiple second attributes, the second attribute weights, the multiple third attributes, the third attribute weights, the multiple fourth attributes and the fourth attribute weights, the test questions and the target experts are matched, and a matching success record set is obtained; and performing matching effect evaluation on the successful matching record set to obtain an evaluation result, and if the evaluation result is greater than or equal to a preset threshold value, taking the successful matching record set as a target matching result, and allocating paper marking experts according to the target matching result. The accuracy of matching between the expert and the test paper to be read is improved, and the matching efficiency is improved.

The application discloses an industry chain knowledge question and answer interaction method and system based on a large model, groups each process node contained in a target industry chain based on attribute weight values and continuity between process nodes and upstream and downstream relationships, so that the obtained node set contains process nodes that not only reflect the relationship between the upstream and downstream, but also can realize the solidification of the relationship between the process nodes based on the process sequence through the arrangement order between the node sets, and then realize the solidification of the knowledge structure of the target industry chain, reduce the phenomenon that the knowledge is too loose and it is difficult to find the logic and rules contained therein due to the weak knowledge structure, and provide conditions for the model to generate targeted and clear answers. The application is applicable to the application of electronic digital data processing technology in the field of question and answer interaction related technology.

The invention provides a normative checking method and tool for software requirement cases and a computer program product. The checking method comprises the following steps: extracting attributes from a software demand case, and converting the attributes into feature vectors by adopting one-hot coding according to a demand case related vocabulary set and a preset weight of each attribute; vector dimensions are equal to the size of the vocabulary set, vector positions are in one-to-one correspondence with words in the vocabulary set, and word frequencies are weighted and counted according to attribute weights and then are filled into the corresponding vector positions; on the basis of the feature vectors, classifying the software demand cases by utilizing a classification model; and matching the corresponding writing specification according to the category of the software demand case, and carrying out natural language structured analysis on the description of the software demand case by adopting an improved shift-protocol dependency analysis method to judge whether the description conforms to the writing specification or not. According to the method, the problem of low manual inspection efficiency is solved, the inspection efficiency is greatly improved through multi-dimensional technical optimization, and standardization and intelligentization of software demand review are promoted.

The invention relates to the technical field of fine adjustment of large mine models, in particular to a multi-field data weighted mixing method, device and equipment for fine adjustment of a large mine model and a computer storage medium. According to the method, the domain target weight and the sample multi-homing probability are accurately combined in the batch construction stage, controllable adjustment of the contribution proportion of each domain is achieved, and it is ensured that low-resource or key domains get full attention in training; in the loss calculation stage, multi-field loss is subjected to weighted fusion by using the attribution weight, so that the model learning process better fits the real multi-field distribution of the sample. Finally, according to the method, the performance balance and the overall generalization ability of the fine-tuned large mine model in each sub-field task are remarkably improved, and the adaptability to a complex multi-label scene is enhanced.

The invention discloses a sextuple digital twin modeling method and system, and the method comprises the steps: building a physical architecture model based on a system engineering method, obtaining and preprocessing the operation data of a simulation system, and defining a sextuple attribute target; mathematical models are constructed for the six properties respectively, and an interaction relation model among the six properties is constructed according to the coupling effect among the six properties; carrying out multi-objective optimization, dynamically adjusting six property weight parameters, and generating a digital twinborn model; a multidisciplinary subsystem model is integrated, data compatibility among different disciplinary tool chains is achieved, and a multidisciplinary collaborative simulation platform is formed; and operating the digital twinborn model on the multidisciplinary co-simulation platform, feeding back and correcting parameters of the digital twinborn model in real time, and outputting a multi-dimensional evaluation report. According to the method, models among different subjects are effectively compatible, powerful support is provided for multidisciplinary co-simulation, it is ensured that data are kept consistent in the conversion process, model functions and performance are not damaged, and a reliable basis is provided for optimization and decision making of a complex system.

The invention discloses an alarm data aggregation method and device, equipment and a medium, and belongs to the technical field of computers. According to the method, the importance of condition attributes relative to each equivalence class corresponding to a preset decision attribute set is introduced, and according to the difference between the importance of each condition attribute in the preset condition attribute set relative to the equivalence class and the importance of the whole preset condition attribute set relative to the equivalence class, the equivalence class is determined. And determining an attribute weight of each condition attribute corresponding to the equivalence class, thereby identifying important attributes in respective associated condition attributes under different equivalence classes, and further judging whether the related alarm data is associated with the same attack classification according to the important attributes under each equivalence class. And the alarm data associated with the same attack classification is aggregated.

Devices, systems, and methods are disclosed for dynamically adjusting weights assigned to properties of chatbot responses. In at least one embodiment, one or more neural networks are used to characterize one or more chatbot queries and to dynamically adjust weight values, which are assigned to properties of responses to the one or more chatbot queries, at least partially based on the characterization of the one or more chatbot queries.

The invention provides a personnel dynamic equilibrium distribution method and system based on multi-constraint optimization, and relates to the technical field of personnel distribution, and the method comprises the steps: inputting a to-be-distributed personnel set, a target number proportion of each group, a personnel multi-dimensional attribute set, and an attribute weight set; initializing each group of current personnel set to be empty, and calculating global distribution of each attribute; and carrying out person-by-person distribution on the to-be-distributed person set by taking the target person number proportion of each group as a constraint, calculating a plurality of comprehensive expected values distributed to the plurality of candidate groups one by one according to the person multi-dimensional attribute set and the attribute weight set in combination with the global distribution of each attribute, and generating an optimal distribution decision by taking the maximization of the comprehensive expected values as a target. According to the method, the technical problem that the rationality of a team structure is affected due to unbalanced distribution of attributes of personnel among groups caused by the fact that adjustment cannot be made according to the real-time progress of distribution or the imbalance condition of the current attributes by generally adopting a static weight polling distribution method in the prior art is solved.

The application discloses a kind of fusion user sentiment and knowledge graph enhanced scenic spot recommendation method, belong to scenic spot personalized recommendation field, including according to scenic spot attribute information, constructs scenic spot knowledge graph AKG;Using translation model TransR learns the low-dimensional representation vector of entity and relation in AKG as the initial input of model;Introduce multi-head attention mechanism to learn scenic spot attribute weight, adopt graph neural network to learn scenic spot embedding;According to historical interaction information, user visit scenic spot sequence is constructed in time sequence;From comment, the sentiment of user is analyzed, user sentiment change function is designed, and user interest embedding is learned;Loss function is designed to train model, user and scenic spot embedding are learned according to the model after training, similarity calculation is carried out, and scenic spot is recommended to user according to the result.The application makes full use of scenic spot attribute information, alleviates the sparsity of data, considers the change of real emotion of user, and improves the recommendation precision of personalized scenic spot.

The invention discloses a high-dimensional data stream-oriented adaptive kernel density anomaly detection method, which comprises the following steps of: calculating each attribute weight based on information entropy, and clustering input data streams in combination with an improved DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm; extracting a representative data subset from each cluster through a KL divergence optimization model; calculating an upper bound filtering normal point of a KOF value based on the reachable distance; calculating an accurate KOF value of the remaining data points and judging that the KOF value is abnormal; the method has excellent performance in the aspects of high efficiency, universality, real-time performance, accuracy and the like, has remarkable advantages compared with the prior art, and provides a new thought and method for solving the problem of anomaly detection in the multi-dimensional data stream.

The invention relates to the technical field of user portrait construction, and provides a user portrait construction method and system based on big data, and the method comprises the steps: collecting behavior data, recognizing a keyword group, and obtaining the initial weight, the attribute tag, the browsing frequency and the browsing time of the keyword group; dividing the interest attribute label, and obtaining an attribute initial weight, attribute browsing time and a first update attribute weight; setting an attribute tag of a pre-pushed advertisement and an advertisement tag weight, determining an interest index of an advertisement pushing party to the same user, and screening high-value users; and determining a second update attribute weight according to the first update attribute weight and the advertisement label weight, and constructing a user portrait in combination with the interest attribute label of the high-value user. According to the invention, the efficiency and accuracy of user portrait construction based on big data can be improved.

The invention relates to a fairness repairing method for constructing a table data classification model of multiple loss items, and belongs to the field of artificial intelligence and machine learning. The method comprises the following steps: firstly, counting the number proportion of various different sensitive attributes in a training set, adjusting the weight of a fair regular term of a loss function according to the proportion to retrain a model, and calculating a change ratio of a fairness index to an accuracy index by utilizing a test set; secondly, randomly replacing each attribute of the table with a reconstructed sample, and calculating the similarity between an original sample and the reconstructed sample as a model comparison loss item; and finally, adjusting the weight of a fair regular term and comparing the weight of a loss term by using the change ratio, and retraining the model to generate a fair repair model. Aiming at the problems that prediction fairness imbalance of a model to other sensitive attributes is easily caused by only processing a single sensitive attribute and the accuracy is influenced by deleting the sensitive attributes in the existing method, fairness repair under the condition that the model accuracy is kept is realized through training sample reconstruction and sensitive attribute weight distribution.

Embodiments of the present application provide a list data filtering method and device, electronic equipment and readable storage medium, and relate to the technical field of computer. The method comprises: determining at least one candidate list data from the to-be-filtered list according to a target screening condition, and then determining invalid attributes that do not match at least one target attribute from a plurality of preset attributes based on the at least one target attribute corresponding to the target screening condition. The attribute weight of each target attribute can be improved according to the initial weight of the invalid attribute to obtain the first attribute weight corresponding to each target attribute, and then the list data weight of each candidate list data is obtained according to the first attribute weight corresponding to each target attribute. The final valid list data is determined by judging whether the list data weight is greater than a preset list data weight threshold, thereby improving the accuracy of list data screening and making the screened list data more application valuable.

The application relates to a large-scale heterogeneous multi-attribute equipment scheme clustering decision method and medium, which comprises the following steps: converting heterogeneous attribute data in a digital twin decision information system into binary connection numbers; determining attribute weights by adopting a MEREC method; constructing attribute irreplaceable values based on the attribute weights and the binary connection numbers; generating a scheme similar network according to the attribute irreplaceable values; performing network clustering operation on the scheme similar network to output a scheme clustering result; and calculating the comprehensive target heart distance of the schemes in the cluster based on the scheme clustering result by adopting a grey target theory, and generating a final scheme score by fusing the scheme global value and the comprehensive target heart distance. The method solves the problems of heterogeneous information fusion distortion, attribute irreplaceable mechanism loss and low large-scale scheme clustering efficiency in a digital twin decision scene by constructing a 'heterogeneous data unification-attribute irreplaceable mechanism modeling-network clustering-clustered decision' mechanism.

The embodiment of the invention provides a video communication method and device, equipment, a medium and a program product. The method comprises the following steps: acquiring a first attribute set of a first communication terminal in a video communication group, wherein the first attribute set comprises attribute data of the first communication terminal associated with target video data; determining a second attribute set according to the attribute data and candidate attributes of the target video data, wherein the second attribute set comprises video indexes; determining a third attribute set according to the attribute data, the candidate attributes and the frame size of the target video data, wherein the third attribute set comprises performance indexes; and determining an attribute weight according to a service scene associated with the video communication group, and determining a target attribute associated with the target video data according to the second attribute set, the third attribute set and the attribute weight. According to the embodiment of the invention, the target attribute is determined by comprehensively considering the video index, the performance index and the service scene, and the video stream is sent according to the target attribute, so that the user experience is optimized.

The invention relates to the field of data cleaning, and discloses a multi-source heterogeneous document deduplication method and system based on attribute weighted fusion, equipment and a medium, the method is characterized in that repeatability judgment is carried out on two documents based on a key attribute set and a mapping table, so that the repeatability judgment process focuses on and weights the core attributes of the documents, and the data cleaning efficiency is improved. According to the method, the duplicate removal problem is converted from text similarity comparison to discriminant force weighted fusion decision of key attributes, non-key differences among documents are ignored, recognition of repeated contents of the documents is more accurate, accuracy and robustness of repeatability judgment are improved, and a weight mechanism allows that the repeated contents of the documents can be more accurately recognized when part of key attributes are missing or errors exist. Reliable judgment can still be made based on existing key attributes, harsh dependence on information integrity is reduced, and high-precision repeated content recognition can still be achieved in an information missing scene. Besides, by configuring different key attribute sets and mapping tables, the method can be quickly and flexibly adapted to various vertical fields, and the universality and generalization ability are improved.

This invention provides a method, apparatus, device, and medium for personalized configuration of delivery confirmation methods, including: constructing a multi-dimensional attribute library for packages; extracting package attributes corresponding to a target package based on the constructed multi-dimensional attribute library; constructing a feature vector for the target package based on the package attributes of the target package and the logical relationships between the attributes; and completing intelligent matching and decision-making of delivery confirmation rules based on the feature vector of the target package, combined with the global attribute weights corresponding to each package attribute of the target package and a digital delivery confirmation rule library. This invention addresses industry pain points such as rigid processes, insufficient intelligence, significant system performance shortcomings, and poor adaptability to complex scenarios in traditional logistics delivery confirmation models. Through a complete technical system encompassing multi-dimensional attribute modeling, weighted intelligent matching, LLM for complex scenarios, high-frequency rule caching, and high-concurrency traffic control, it achieves precise personalized configuration of delivery confirmation methods and efficient and stable system operation.