71results about How to "Improve retrieval performance" patented technology

The invention discloses a large model knowledge base construction method based on multi-granularity retrieval. The method comprises the steps of obtaining a target text and analyzing a document structure, splitting and reorganizing text data according to three levels of titles, fragments and sentences, and constructing three levels of knowledge of the titles, the fragments and the sentences; using a large language model to construct a generative index as question-level knowledge according to the sentence-level knowledge; using a vector model to generate a first vector representing index keys of fragment-level knowledge and / or sentence-level knowledge, and a second vector representing index keys of title-level knowledge and / or question-level knowledge; and respectively storing the first vector and the second vector into a vector database, and storing the index values and the index types corresponding to the first vector and the second vector into a relational database. According to the method, multi-granularity document knowledge base construction is achieved, the expression modes of knowledge under different granularities are expanded, the accuracy of user query and knowledge matching is improved, and the problem that information is lost or incomplete in an existing method is solved.

The invention relates to the technical field of electrical operation safety management, in particular to an artificial intelligence-based electrical operation misoperation prediction and correction system, which comprises a data acquisition module, a data processing module, a misoperation prediction module, a correction suggestion generation module, a real-time monitoring and feedback module, a self-learning module and a data storage module, operation data and environment data are collected in real time through a sensor network, the data are preprocessed, feature vectors are generated, a misoperation prediction model is constructed based on a long short-term memory network algorithm, the misoperation occurrence probability is predicted, a decision tree algorithm is adopted to generate correction suggestions, and prediction results and the correction suggestions are presented through a human-computer interaction interface. The self-learning module is used for dynamically optimizing the prediction model and the suggestion generation model; the data storage module is used for classifying and storing the operation data, the environment data and the prediction result; according to the invention, the misoperation is accurately predicted in real time, the intelligent correction suggestion is generated, and the safety and efficiency of electrical operation are remarkably improved.

The invention provides a building information retrieval method and device, electronic equipment and a storage medium. The method comprises the following steps: establishing a training data set; wherein the training data set is a training image-training text pair set; the training image is an image in the building field, and the training text is description of multiple different languages; inputting the training image-training text pair into the initial retrieval model, and obtaining a target retrieval model when the loss of a target loss function of the similarity between the first embedded vector and the second embedded vector is minimum; wherein the first embedded vector is obtained by converting a training text, the second embedded vector is obtained by a fused image, and the fused image is obtained by fusing feature information of a training image under different scales; obtaining to-be-retrieved data; and inputting the to-be-retrieved data into the target retrieval model to obtain a retrieval result. According to the method, the detail features of the building image are more concerned, the overall features are not lost, and the retrieval performance and adaptability of the target retrieval model in the building field are improved.

The invention discloses a course intelligent question-answering method and system based on retrieval enhancement generation and a large model, and the method comprises the steps: processing a to-be-answered artificial intelligence course knowledge field question, obtaining a question sequence, and obtaining a text representation vector of the question; obtaining a similarity score based on the text representation vector and the paragraph representation vector, and calculating a BM25 score between the question and the paragraph; based on the similarity score and the BM25 score, calculating to obtain a mixed score corresponding to each paragraph, and sorting the paragraphs from high to low according to the mixed scores to obtain a topN document; paragraphs in the topN document are reordered, and a topK document is obtained; and taking the topK document as the context of the question, taking the topK document and the question as input by the large language model, and outputting the answer of the question. According to the method, intelligent question answering of artificial intelligence course knowledge in the education field is realized, the learning efficiency of students is improved, and the teaching burden of teaching assistance of teachers is reduced.

The invention discloses an intelligent question answering method and device, electronic equipment and a storage medium, and relates to the technical field of artificial intelligence. The method comprises the steps of obtaining an original input question of a user, and performing matching retrieval on the original input question in a pre-constructed professional word explanation library to obtain a target professional word matched with the original input question and an explanation of the target professional word; according to the target professional word and the explanation of the target professional word, rewriting the original input question to obtain a target input question; and performing knowledge retrieval in a pre-constructed knowledge base according to the target input question, and generating reply content for the user according to a retrieval result. According to the method, the original input question of the user is rewritten based on the professional words and explanations thereof, so that the rewritten input question and the text in the knowledge base have co-occurrence words more easily, the knowledge retrieval effect can be improved, and the accuracy of the reply content generated by the intelligent question and answer is ensured.

The invention discloses a large model semantic coding BIM data and engineering entity matching system, which comprises a knowledge graph construction module used for establishing a knowledge graph according to building project historical data; the BIM data and engineering entity coding module is used for coding the current BIM data or the current engineering entity into a vector by using a Chinese pre-training language model; the BIM data and engineering entity similar semantic retrieval module is used for retrieving a plurality of similar vectors of the current BIM data vector or the current engineering entity vector in the knowledge graph by using a retrieval enhancement generation model; and the BIM data and engineering entity matching module is used for matching the current BIM data vector or the current engineering entity vector with the corresponding similar vector by using cosine similarity. The invention further discloses a BIM data and engineering entity matching method for large model semantic coding. Compared with manual operation, the technical scheme of the invention has the advantages of high accuracy, high efficiency and high flexibility.

The invention belongs to the field of artificial intelligence and financial science and technology, and relates to a question and answer reply generation method which comprises the steps that a query text is input into a reply generation model, and the reply generation model comprises an input analysis network, a style recognition network, a retrieval enhancement generation network and a reply selection network; processing the query text through the input analysis network to obtain a text input vector; inputting the user historical dialogue data and the text input vector into a style recognition network to generate a style recognition result; inputting the text input vector and the style recognition result into a retrieval enhancement generation network, and generating a plurality of candidate replies in combination with related document fragments retrieved by a preset knowledge base; a final reply is selected from the plurality of candidate replies through the reply selection network. The invention further provides a question and answer reply generation device, computer equipment and a storage medium. In addition, the invention also relates to a block chain technology, and the final reply can be stored in the block chain. According to the method, the content meeting the long-term preference and personalized requirements of the user can be generated.

The invention relates to the technical field of artificial intelligence, and provides an information retrieval method and device and a medium, which can segment to-be-processed data according to a configured segmentation strategy to obtain the to-be-processed data, and improve the organization and retrieval efficiency of the data. Performing vectorization processing on the segmented data to obtain vector data, and storing the vector data into an embedded database so as to facilitate subsequent similarity retrieval; the problem optimization is performed on the information retrieval request, so that the retrieval content can be better matched, and the retrieval accuracy is improved; a target cue word is generated by utilizing similar knowledge matched in an embedded database and a pre-configured cue word template, so that the retrieval efficiency and accuracy are further improved; according to the method, the retrieval model is retrieved according to the information retrieval request, the retrieval model is utilized to respond based on the target prompt word, and flexible retrieval of the model not only optimizes the retrieval effect, but also reduces the retrieval cost.

The invention discloses a multi-modal Hash retrieval method, system and equipment based on a multi-view center structure and a medium, belongs to the technical field of artificial intelligence and multi-modal Hash retrieval, and aims at achieving the technical purpose of obtaining balance between intra-class compactness and inter-class separability in the multi-modal Hash retrieval process. According to the technical scheme, the method comprises the following steps: constructing a multi-modal data set; constructing a multi-modal Hash retrieval model based on a multi-view center structure; training the model; wherein the construction of the multi-modal Hash retrieval model based on the multi-view center structure specifically comprises the following steps: modal specific prototype learning: respectively extracting refined features from corresponding modals by using an image modal depth multilayer perceptron and a text modal depth multilayer perceptron; calculating the average value of the refined features of the image modality and the text modality to obtain a modality specific prototype, and further obtaining the unique features of the image modality and the unique features of the text modality; carrying out multi-modal integrated prototype learning; and performing multi-view semantic enhanced hash learning.

The invention belongs to the field of text-video retrieval, and particularly relates to a subtitle generation-based multi-granularity semantic alignment text-video retrieval method, which comprises the following steps of: establishing an SAEC model, training the SAEC model, and carrying out text-video matching retrieval through the trained SAEC model. According to the method, a new model SAEC is provided, the model significantly improves the retrieval accuracy from the text to the video by matching local details in the video and understanding the global semantics of the video, and meanwhile, the model not only can understand the direct content of the video, but also can grasp a more complex situation by jointly generating the subtitles and the video.

The invention discloses an API (Application Program Interface) text retrieval method and device and electronic equipment. The method comprises the following steps: receiving a query text which is input by a target object and is used for retrieving an API (Application Program Interface); a preset model is adopted to retrieve target interface information corresponding to the query text in an API interface document, the preset model is obtained by using the API interface document as a data set for training, and the API interface document comprises interface metadata and description information of characteristics of a platform where the API interface is located; and returning the target interface information to the target object. According to the method and the device, the technical problems of low retrieval efficiency and low retrieval accuracy of the retrieval of the APIs when the number of the APIs is large in the related technology are solved.

The invention discloses a personal data retrieval method and device based on a deep neural network, and relates to the technical field of data retrieval. The personal data retrieval method and device based on the deep neural network specifically comprise the following steps: S1, data preprocessing and index construction; s2, designing a deep neural network model; s3, performing strategy retrieval and fragmentation processing; s4, displaying a retrieval result; s5, performing feedback and optimization; the semantic relation between the file names is learned through the deep neural network, the intelligent retrieval capacity exceeding traditional keyword matching is achieved, the intention of a user can be accurately understood, and even if the file names are not completely matched, related files can be found; after complete file name retrieval fails, a fragmentation processing strategy is adopted, deep neural network similarity calculation is combined, the retrieval range is effectively expanded, the retrieval recall rate and accuracy are improved, and meanwhile the input burden of a user is reduced.

The invention provides a recognition model and a recognition method for medical conference content, and the recognition model is obtained through the following processes: constructing a medical industry knowledge graph, extracting a needed SPO triple from the medical industry knowledge graph to generate a dialogue corpus, and generating a medical industry corpus data set; pre-training a large language model by using the medical industry corpus data set; a Fine-tune corpus data set is made; constructing a Prompt template, wherein the Prompt template comprises three parts, namely an instruction, question information and auxiliary information; on the basis of the Fine-tune corpus data set and the Prompt template, performing fine tuning on a pre-trained large language model on a downstream task to enable the pre-trained large language model to effectively adapt to a new task; and an increment weight is newly added to optimize the fine-tuned large language model, so that weight fusion of the large language model is realized, and a recognition model of the medical conference content is obtained. According to the invention, the big language model and the RAG fusion technology are utilized to carry out natural language processing on the medical conference content, so as to solve the problems in the prior art.

The invention provides a knowledge graph representation method and device based on similarity diffusion, and the method comprises the steps: constructing a target knowledge graph based on a multi-modal large model and a human-in-the-loop mechanism; constructing an adjacent graph matrix of the target knowledge graph based on the similarity relationship among the entities in the target knowledge graph; based on the adjacent graph matrix, performing similarity diffusion under local clustering constraint to obtain a first similarity matrix; and based on the first similarity matrix, performing geometric transmission on the target knowledge graph in a manifold graph to obtain relation representation of the target knowledge graph. Therefore, the more effective and robust relationship representation of the target knowledge graph is obtained, the construction and updating of the knowledge graph can be guided through the relationship representation, and the mining, fusion, reasoning and retrieval capabilities of the knowledge graph are improved.

The invention relates to the technical field of image retrieval, and discloses a Hash image retrieval method and system based on improved triple loss, and the method comprises the following steps: inputting a query set image into a trained Hash network model based on improved triple loss, and obtaining a Hash code of the query set image; comparing the difference degree between the Hash code of the query set image and each Hash code in the Hash code library; sorting the difference degrees, and outputting a retrieval result according to the sorting; and calculating the average accuracy of the image query set according to the retrieval result. According to the invention, the retrieval precision can be improved.

The invention discloses an asynchronous vector index construction and vector retrieval method, and belongs to the technical field of databases, the asynchronous vector index construction is that a vector database takes a time sequence database as a storage engine of a bottom layer, and the time sequence database comprises a memory buffer area, a background working thread and a disk; a data storage unit of a time sequence database is used as a unit, when the size of a memory buffer area reaches a buffer threshold value, a background working thread converts vector data of the memory buffer area into a storage format of the data storage unit, and at the moment, if the scale of the vector data reaches a scale threshold value, a vector index is constructed for the vector data and written into a disk; index construction parameters are dynamically adjusted according to the vector data scale, the time sequence database asynchronously combines the data storage units in the disk, and vector indexes are reconstructed during combination; and the asynchronous vector retrieval is to perform partition query based on an asynchronous vector index constructed by an asynchronous vector index construction method.

The invention discloses a computer data management system and method, and relates to the technical field of data management, and the method comprises the steps: reading a heterogeneous data set, dividing the heterogeneous data set into text, time sequence and binary data groups according to data types, extracting a core field of each data group, and generating a deduplication feature and an index feature, setting a sliding window based on the data type and generating a window index feature abstract; calculating a comprehensive similarity score between the data blocks according to the sliding window, screening a similar data block set, determining a reference data block and a redundant data block, clustering the reference data block to generate an initial cluster, and if the global deduplication rate and the clustering preprocessing validity do not meet preset conditions, re-initializing the sliding window; according to the method, the cluster structure is optimized and the cluster feature abstract is generated based on the retrieval popularity predicted value of the initial clustering cluster, the three-layer index structure comprising the global index, the intra-cluster index and the feature index is constructed, the retrieval efficiency is tested, the index hierarchy is optimized, and the storage efficiency and the retrieval performance of heterogeneous data management are remarkably improved.

The application discloses a cross-natural language code retrieval model training method, a cross-natural language code retrieval method, a device, equipment and a medium, and relates to the technical field of artificial intelligence and software engineering. The cross-natural language code retrieval model training method comprises the following steps: obtaining an original corpus database, and constructing training data according to the original corpus database; performing confusion and inversion on main language codes to obtain main language code samples, wherein the main language code samples comprise main language code positive samples and main language code negative samples; and training an initial model through a gradient inversion layer according to the training data and the main language code samples to obtain a target model. According to the application, the natural language-specific "fingerprint" features in the codes can be removed, the embedding space alignment direction can be unified, the sampling distribution deviation in the training process can be reduced, and the consistency and generalization capability of cross-language code retrieval can be improved.

A host device, a storage device, and systems and methods thereof are provided. The B-tree-based retrieval method executed by a host device includes: acquiring query information from a memory of a storage device based on a B-tree query request for a table; sending query information to the storage device so as to store the query information in a memory of the storage device; sending a B tree query request to a storage device to start a query for the table in the storage device; a result of the B-tree query request is received from the storage device.

The application relates to the technical field of storage computing, and discloses a system, a method and an electronic device for directly searching data in a storage medium, the system comprising a large-scale vector database, a nearest neighbor search operation unit and a host computer, the large-scale vector database is arranged in the storage medium and is used for storing a vector set; the nearest neighbor search operation unit is arranged in the storage medium; the host computer is provided with a CPU and a cache unit; when the nearest neighbor search operation unit receives an NNS_CSD instruction sent by the CPU, the vector of the database is read, similarity measurement is carried out, and the search result meeting the condition is sent to the cache unit, the application directly searches data in the storage medium, the NNS_CSD instruction is used to start the calculation in the storage medium, the nearest neighbor search operation unit in the storage medium is used for vector reading and similarity measurement, compared with a traditional architecture, the migration of data is greatly reduced, and the overall performance is improved.

The application discloses a retrieval method and related device. According to a to-be-retrieved input string, prefix matching is performed in a target dictionary tree to obtain a target leaf node. According to a remaining string of the to-be-retrieved input string, each full string in a data file corresponding to the target leaf node is checked to obtain a checking result of each full string, the checking result being used to indicate whether the corresponding full string matches the to-be-retrieved input string. If the checking result indicates that a target full string in the data file matches the to-be-retrieved input string, a word corresponding to the target full string is determined as a candidate of the to-be-retrieved input string. The application combines the target dictionary tree and the data file to perform retrieval, the target dictionary tree has the characteristic of high retrieval efficiency, and meanwhile, since the remaining string is a partial string of the to-be-retrieved input string, the time for performing retrieval in the data file is also reduced. Therefore, the retrieval efficiency and retrieval performance are improved, even the response performance of the entire input method is improved, and the input experience is improved.

The invention provides a big data-based image-text retrieval method and system. The method comprises the following steps of: obtaining to-be-retrieved image-text modal data; performing feature extraction on the image modal data and the text modal data to obtain corresponding image feature representation and text feature representation, and determining an image-text modal fusion entropy through the image feature representation and the text feature representation; determining a text description array and an image description array according to the image-text modal fusion entropy, and performing attention conversion according to the text description array and the image description array so as to obtain local image-text fusion features and global image-text fusion features in a two-dimensional modal space; and splicing the local image-text fusion features and the global image-text fusion features to obtain a fusion modal representation between the image modal information and the text modal information, and performing image-text retrieval based on the fusion modal representation. By adopting the scheme, modal fusion can be performed on the image modal information and the text modal information, so that the retrieval performance during image-text retrieval is improved.

A cross-view image retrieval method for unmanned aerial vehicle navigation, comprising the following steps: step one, first enhancing the image, step two, based on the ConvNeXt network for cross-view image retrieval for unmanned aerial vehicle navigation, multi-scale feature extraction is performed on the enhanced image; step three, multi-dimensional feature fusion is performed based on an attention mechanism; step four, after multi-dimensional feature fusion, the loss value is calculated, and a detection model is obtained based on the determined weight to perform image retrieval; through the multi-scale feature convolution and multi-dimensional feature fusion technology, the technical problems of insufficient matching precision and poor retrieval robustness caused by the differences in view, scale and structure between satellite images and unmanned aerial vehicle images are solved; the application also includes a system, a device and a medium for implementing the above method.

This application discloses a trademark retrieval database generation, retrieval method, and apparatus that enhances local features. The method involves constructing a multi-classification training set for trademark graphic elements and training it in an image multi-classification model to obtain a trademark element multi-classification model; constructing a target detection training set for important graphic elements based on trademark data and training it in an image target detection model to obtain a target detection model; inputting all trademark data into the trademark element multi-classification model for prediction and selecting trademark elements as field information; inputting all trademark data into the target detection model to detect important graphic element targets and extracting them into sub-images; inputting all sub-images and the original trademark data image into a trademark encoding model for encoding to obtain encoding vectors; and inputting the field information and encoding vectors into a vector database to obtain the trademark retrieval database. Using the trademark retrieval database provided in this application can significantly improve trademark retrieval results.

The invention belongs to the field of image retrieval, and particularly relates to an interactive image retrieval method based on multi-modal fusion. Comprising the following steps: encoding a text, a sketch and an image by adopting a deep neural network model to obtain a text embedding vector, a sketch embedding vector and an image embedding vector; fusing the text embedding vector and the sketch embedding vector to obtain a fused embedding vector; constructing a reward function according to the image embedding vector and the fusion embedding vector; constructing and training a reinforcement learning model containing a reward function, and finishing model training when a preset number of training times is reached; the sketch of the user and the text description of the retrieval object by the user are obtained in real time, the sketch and the text are input into the trained reinforcement learning model for real-time retrieval, and the model feeds back the retrieved image to the user; the user further optimizes the sketch according to the feedback image until a target image satisfying the user is retrieved; according to the method, the stable performance of retrieval feedback can be ensured, and better retrieval experience is realized.

This invention discloses a supervised multimodal hash retrieval method and system based on label enhancement, belonging to the field of artificial intelligence and multimedia retrieval technology. The technical problem this invention aims to solve is how to better capture the similarity information between multimodal data points and achieve better performance and accuracy in multimodal retrieval tasks. The technical solution includes: data preprocessing: acquiring and organizing public datasets of image and text modalities, and dividing each public dataset into training, testing, and retrieval datasets; extracting deep features: using a pre-trained network model to extract features from the raw data of the public datasets of image and text modalities respectively, obtaining deep features of the image modality and the text modality; offline training; variable update and optimization; and online query. The system includes a data preprocessing unit, a feature extraction unit, an offline training unit, a variable update and optimization unit, and an online query unit.

The invention provides a method and system for constructing and identifying a coal mine water disaster prevention and control intelligent knowledge base, and the method comprises the steps: 1, creating a general knowledge base, configuring the general knowledge base, and importing a coal mine water disaster prevention and control sample data set into the general knowledge base after the configuration succeeds; 2, converting the coal mine water disaster prevention and control sample data set into vector samples in a configured universal knowledge base, and pre-training the vector samples to obtain an intelligent recognition model; 3, packaging the intelligent identification model and a configured universal knowledge base to obtain a coal mine water disaster prevention and control intelligent knowledge base; and 4, inputting a target question, analyzing the target question based on the coal mine water disaster prevention and control intelligent knowledge base, and outputting a target answer corresponding to the target question.

A multi-modal transfer learning framework for video corpus segment retrieval, the learning framework is trained by a segmented pre-training transformer using image-text pairs, video-text pairs and a VCMR dataset to transfer the knowledge of ITR and VR to VCMR, so that better performance can be achieved on a relatively small VCMR dataset, comprising: a retrieval module, the retrieval module comprises a text encoder, a visual encoder and a video retrieval module; the text encoder adopts a segmented pre-training transformer to generate a query representation, and encodes the query representation to generate a text representation; the visual encoder is used to generate a video representation by adopting a visual transformer; the video retrieval module is used to retrieve video segments; a segment positioning module, the segment positioning module introduces L1 loss and IoU loss to directly predict the time span in a fine-grained manner.

The embodiment of the invention relates to a training processing method of a video search model and a video search processing method and device, and the method comprises the steps: obtaining a sample video retrieval text, and extracting a first sample retrieval text feature of the sample video retrieval text through an initial video search model; distributing a plurality of first sample feature weights in one-to-one correspondence with preset multi-mode video information; determining a first sample video matched with the first sample retrieval text feature and the plurality of first sample feature weights through an initial video search model; and calculating a first matching degree between the first sample video and the sample video retrieval text, and when the first matching degree is smaller than a preset matching degree threshold value, adjusting model parameters of the initial video search model until the adjusted matching degree is smaller than the preset matching degree threshold value, thereby obtaining a trained video search model. According to the technical scheme, the matching degree of the searched video clip and the video retrieval text is ensured, and the video retrieval effect is improved.

The invention relates to a continuous Hash learning retrieval method and device for stream data, and the method comprises the steps: constructing a Hash retrieval model, introducing a continuous learning framework into the training process of the Hash retrieval model, firstly extracting a first feature from unlabeled training stream data based on self-supervised learning through a residual network, and then extracting a second feature from the unlabeled training stream data based on self-supervised learning; the convolutional neural network can quickly learn labeled sample data by using the first feature, and feeds back the learned second feature to the residual network to consolidate the residual network; the training stream data comprises first real-time stream data collected in real time and pre-stored stream data in a database, so that the model can avoid forgetting old pre-stored stream data knowledge while continuously learning new real-time stream data, and the continuous learning framework is also effectively connected with a hash function; the residual network and the convolutional neural network are used for continuous learning, and then extracted features guide learning of a hash function, so that the quality of generated hash codes is continuously improved, and the retrieval performance is further improved.