13results about How to "Rich expression" patented technology

The invention relates to the technical field of protein prediction, and discloses a protein interaction prediction method based on cross-modal enhanced representation learning. According to the method, a Convform module and a Graphform module are designed, and a global attention mechanism is introduced into each module, so that the limitation that only local features or global features are concerned in an existing method is broken through. According to the design, the model can capture the local mode and the global dependency relationship of the protein at the same time, and more comprehensive protein representation is provided. The invention provides a cross-modal enhancement method based on comparative learning. The problem of difference between a sequence mode and a structural mode is solved through an explicit mode alignment process. The method not only can maintain unique information of each mode, but also can give full play to complementarity among the modes, so that richer protein expression is obtained. The cross-protein fusion module provided by the invention can more accurately capture essential characteristics of protein interaction, and the accuracy of PPI prediction is remarkably improved.

The invention belongs to the technical field of biological feature recognition, and relates to a Transform-based gene data mining method, which comprises the following steps of: loading gene expression data from a file, and preprocessing; inputting the data into a designed model for training and evaluation; the designed model comprises a CNN (Convolutional Neural Network) and an improved Transform model; the improved Transform model comprises an encoder and a classifier; the encoder adopts a multi-head self-attention mechanism, a convolutional self-attention mechanism and a feedforward neural network; processing the data by using the trained model, and storing the fusion attention weight obtained by the coding layer in the data processing process; and finally, averaging the fusion attention weights of all batches, calculating the comprehensive attention weight of each gene at different time points, and identifying the most important gene by calculating and analyzing the attention score of each gene. According to the method, the problems of scarcity and imbalance of gene data can be solved, the risk of overfitting is reduced, and the precision of gene data analysis is improved.

The invention discloses a knowledge retrieval feedback method, electronic equipment and a computer readable storage medium, and the method comprises the steps: obtaining an initial query statement and at least one transwritten query statement corresponding to the initial query statement, and determining a vector feature corresponding to each query statement and a keyword set corresponding to all query statements; obtaining a retrieval sub-list matched with each vector feature, and de-duplicating and re-sequencing the knowledge fragments in all the retrieval sub-lists based on the sequence of the knowledge fragments in each retrieval sub-list and the number of reappearing times of the knowledge fragments in all the retrieval sub-lists to obtain a vector retrieval feedback list; and obtaining a keyword retrieval feedback list matched with the keyword set, and de-duplicating and re-sequencing the knowledge fragments in the vector retrieval feedback list and the keyword retrieval feedback list on the basis of sequencing and reproduction times of the knowledge fragments in the vector retrieval feedback list and the keyword retrieval feedback list to obtain a target retrieval feedback list. According to the scheme, the knowledge retrieval feedback precision can be improved.

The invention discloses an aspect emotion triple extraction method based on a multi-channel graph convolutional network, and relates to a graph convolutional neural network architecture extraction method, and the network architecture comprises three core modules: a feature output module, a multi-channel graph convolutional network module and a table decoding module. The framework effectively utilizes the advantages of BERT, BiLSTM and a prompt learning mechanism to carry out feature extraction. Through parallel three-channel design, word pair relations under positive, negative and neutral sentiments are modeled respectively, so that sentiment information in a text can be extracted and expressed more comprehensively. The multi-channel design not only improves the expression ability of the model, but also enhances the robustness of the model when processing complex emotion information. The method provides core power for developing a user comment sentiment analysis system, and the new technology has far-reaching influence on timely product and service improvement of merchants and personal optimized consumption decision.

The embodiment of the invention provides a platform merchant recommendation method, electronic equipment, a storage medium and a program product. The method comprises the steps of obtaining first historical behavior data when a current user interacts with a platform, and determining an initial recommendation basis of the current user according to the first historical behavior data; according to the first historical behavior data and a pre-constructed reference basis recommendation pool, determining a reference recommendation basis of the current user; wherein the reference basis recommendation pool is a data pool which is pre-constructed based on a large model according to second historical behavior data when a historical user interacts with the platform and is used for determining a reference recommendation basis of the current user; determining a final recommendation basis of the current user according to the initial recommendation basis and the reference recommendation basis; and according to the final recommendation basis, determining and presenting at least one to-be-recommended merchant. The method is used for achieving the effect of improving the accuracy and efficiency of merchant recommendation of the platform.

The utility model discloses a gift box convenient for multi-style display, which comprises a main body formed by paperboards and a shell plate used for covering the main body, the main body is provided with a containing groove used for containing products, the edge of the containing groove is symmetrically provided with supporting plates used for supporting the products along the unfolding direction of the shell plate, and the supporting plates are used for supporting the products. An adjusting plate used for fixing and adjusting the angle of the supporting plate is arranged on the body in a foldable mode, and a sliding block used for adjusting the angle of a product is arranged on the side away from the adjusting plate in a sliding mode. Through independent action or interaction of the three parts, the product can be displayed in three modes, so that the effect of displaying the product in a diversified manner is improved, and the expression modes of the product are enriched.

The invention discloses an attention mechanism fused improved YOLOv8 small target detection method and system, and the method comprises the steps: collecting a small target picture and monitoring video data containing a small target, carrying out the preprocessing of the monitoring video data, constructing a small target data set, and carrying out the marking and dividing of the small target data set; the original YOLOv8 network model is optimized to obtain an improved YOLOv8 network model; training the model by using the training set, and adjusting model hyper-parameters by using the verification set in the training process to obtain a small target detection model; evaluating the performance of the small target detection model by using the test set; and deploying the small target detection model in a monitoring camera, and carrying out real-time monitoring and alarming on a small target appearing in a monitoring range of the monitoring camera by using the small target detection model. The technical problems that the detection accuracy is reduced and the actual application requirement is difficult to meet due to the fact that the mouse imaging area is small, the mouse imaging area is shielded and the night contour texture is not clear in the prior art can be solved.

The present disclosure belongs to the field of voice interaction technology, and relates to a painting method and device, storage medium and electronic device based on voice interaction. The method includes: collecting voice information, and performing voice recognition processing on the voice information to obtain painting elements; inputting the painting elements into a pre-trained generative adversarial network so that the generative adversarial network outputs a painting image. The present disclosure can realize the function of outputting a painting image by inputting the painting elements obtained by voice recognition processing into the generative adversarial network. On the one hand, drawing a painting image through voice information provides users with a new way of painting and enriches the user's expression method; on the other hand, it helps users to complete the expression of the painting image without having to learn the operation method of the drawing tool and master the painting knowledge, thereby improving the accuracy and fit of the user's expression of ideas through painting.

This invention relates to the fields of graph neural networks and heterogeneous graph representation learning technology, specifically to a heterogeneous graph node classification method based on relation-aware label propagation. The method includes the following steps: first, based on a given heterogeneous network, then performing type-specific linear transformations on nodes of different types and projecting their features onto a common feature space; second, designing a relation-aware label propagation algorithm for target nodes in the heterogeneous network, generating pseudo-labels through relation subgraphs; and third, employing a two-layer aggregation strategy based on type attention to transmit information and effectively fuse multi-level information of nodes, resulting in richer and more accurate node feature representations. This invention significantly improves classification performance and robustness by designing a relation-aware label propagation method to obtain pseudo-labels, utilizing two-layer aggregation based on type attention for heterogeneous message transmission, and combining it with a multi-objective optimization strategy.

The present invention belongs to the field of power information text data processing, and provides a power information triplet fault priority detection method, system and device. Based on the power information positioning matrix, a power information positioning center matrix is ​​generated, and the distribution positioning and distance positioning between each power information positioning matrix and the center matrix are calculated respectively; the power information dispersion distance of each power information triplet is generated according to the comparison of distribution positioning and distance positioning, and the power information triplet is sorted based on the power information dispersion distance to obtain the power information triplet fault priority sequence, and the fault problem is processed in the order of the priority sequence. The priority detection and processing of fundamental faults in the power system is realized, and the efficiency and accuracy of power grid fault troubleshooting are significantly improved.

The present invention provides a method, device, electronic device and storage medium for sending recommendation information, the method comprising: sending recommendation information of a target reading object to a target display device; wherein the recommendation information comprises search expression information pointing to the target reading object; the target reading object is determined based on the reading needs of the user, the search expression information is constructed based on the structured information of the target reading object, and the structured information comprises the attribute information of the target reading object. By identifying the reading needs to determine the target reading object, and constructing appropriate search expression information through the structured information of the target reading object, it is possible to accurately recommend to the user and enrich the expression of the recommendation information, thereby satisfying the user needs and improving the user experience.

The invention discloses a method for recommending scientific and technological documents by integrating multi-dimensional features of metadata, including: obtaining document interaction record data and document metadata, and performing preprocessing; constructing a knowledge graph based on the preprocessed document metadata, and linking the preprocessed document interaction record data with entities in the knowledge graph to obtain a final knowledge graph; outputting a first embedding vector of each document in the document interaction record data through a knowledge graph embedding model; outputting a second embedding vector of each document in the document interaction record data through a language model; fusing the first embedding vector and the second embedding vector to obtain a final embedding vector of each document in the document interaction record data; obtaining a user research intention representation vector according to the final embedding vector; and obtaining a recommended document set according to the user research intention representation vector and a recommended document list. The invention can improve the accuracy of document recommendation.

The present application discloses a knowledge retrieval feedback method, an electronic device, and a computer-readable storage medium, including: obtaining an initial query statement and at least one corresponding transcribed query statement, determining the vector feature corresponding to each query statement and the keyword set corresponding to all query statements; obtaining a search sublist matched by each vector feature, deduplicating and reordering the knowledge fragments in all search sublists based on the ranking of the knowledge fragments in each search sublist and the number of recurrences of the knowledge fragments in all search sublists, and obtaining a vector search feedback list; obtaining a keyword search feedback list matched by a keyword set, deduplicating and reordering the knowledge fragments in the vector search feedback list and the keyword search feedback list based on the ranking and number of recurrences of the knowledge fragments in the vector search feedback list and the keyword search feedback list, and obtaining a target search feedback list. The above scheme can improve the accuracy of knowledge retrieval feedback.