652 results about "Question answer" patented technology

Provided in the present application is an intelligent question answering method based on a collaboration between a large language model and a knowledge graph, relating to the technical fields of artificial intelligence and natural language processing, the method comprising: decomposing a complex question into a plurality of simple questions, and analyzing the degree of association between the simple questions and a basic function so as to form a multi-hop reasoning path; automatically extracting structured information from the simple questions on the basis of a multi-task learning framework of a large model, so as to construct a knowledge graph; and constructing a cumulative reasoning learning framework on the basis of a logic reasoning large model, and performing iterative verification on a process result formed by the knowledge graph on the basis of the multi-hop reasoning path, so as to correct the reasoning path until a correct answer is inferred.

A method of question answering using enhanced retrieval-augmented generation according to an embodiment includes receiving, by a computing system, a user query, pre-processing, by the computing system, the user query to determine whether the user query is associated with malicious intent, retrieving, by the computing system, relevant data from a knowledge base by using a keyword index and a semantic index in response to determining that the user query is not associated with malicious intent, prompting, by the computing system, a large language model to generate an answer to the user query based on only the relevant data retrieved from the knowledge base, and receiving, by the computing system, the answer to the user query from the large language model in response to the prompt.

The invention provides a complex question and answer method and system based on adaptive task deconstruction and multi-modal evidence aggregation, belongs to the technical field of natural language processing, and designs a dynamic Few-shot prompt construction method based on dependency syntax fingerprints to ensure that a prompt template is matched with a question structure; the invention discloses a dynamic problem deconstruction method based on confidence evaluation and auto-reflection. The method comprises the following steps: recursively decomposing a problem tree by using a large language model; converting the problem tree into a standardized linear task execution sequence by a problem tree context dependence specification and task sequence generation method; obtaining a high-correlation evidence set of each task based on an evidence generation method of two-way recall and cross encoder rearrangement; and the task sequence is reasoned and dynamically optimized by a question answer extraction method based on double-strategy aggregation reasoning. According to the method, the accurate complex question and answer result can be provided on the premise of ensuring the question disassembling quality, restraining error propagation and comprehensively recalling evidences.

The invention provides a knowledge question and answer library agent construction method and system. Efficient knowledge management and question and answer are achieved through cooperation of multiple agents. According to the system, firstly, a multi-modal information extraction agent is constructed, and heterogeneous data such as texts, images and tables are converted into structured vectors and stored; meanwhile, the knowledge graph is dynamically constructed and continuously optimized by the self-adaptive knowledge graph construction agent, and a new relationship is derived through combination of symbolic logic and a graph neural network, so that an evolvable knowledge network is formed. In the question and answer stage, a query analysis agent deeply analyzes the intention of a user and generates sub-queries; retrieving the vector library and the knowledge graph in parallel by the retrieval enhancement generation agent; and the reasoning and synthesizing agent integrates multi-source information and generates an accurate answer with a complete source label through a large language model. Dynamic knowledge management, precise semantic analysis and system self-evolution are achieved, and the method is particularly suitable for professional field scenes needing high-reliability questions and answers.

The invention discloses a dynamic feature retrieval generation and management method based on a cross attention mechanism, which belongs to the technical field of natural language processing, and comprises the following steps: step 1, converting knowledge base document fragments into atomic knowledge units, each atomic knowledge unit comprising question and answer pairs, codes as key value pairs, and adding dynamic priority weights; self-attention query is replaced with a double-channel query structure, one channel is used for cross attention, a cross attention query vector is generated through linear transformation, and key value pairs of atomic knowledge units are used; and step 3, training a cross attention adapter, freezing the weight of the language model, optimizing parameters of the adapter, and dynamically adjusting a loss function by using pre-judgment parameters based on input sequence context complexity. By means of the method, deep correlation description of user input and knowledge fragments can be achieved, semantic ambiguity is eliminated, knowledge injection and context information are balanced, distortion is avoided, and the strict requirement for semantic precision in the professional field is met.

The invention belongs to the related technical field of product detection, provides a multi-view zero sample anomaly detection method and system based on cross-modal prompt reasoning, and aims at solving the problem of multi-view zero sample anomaly detection by constructing core technologies such as multi-view pose estimation and alignment, static-dynamic prompt collaboration, vision-language progressive fusion, feature space semantic enhancement and the like. And a set of end-to-end anomaly detection and reasoning system is formed. Particularly, a collaborative mechanism of a dynamic learnable prompt pool and a static attribute prompt library is designed, deep fusion of prompts is realized through cross attention, and multi-view feature compression and semantic decoding are performed by adopting a visual angle self-adaptive hybrid expert model. Zero sample anomaly detection and visual question and answer performance is further improved on multiple industrial public data sets, and the method can be widely applied to industrial precision part quality inspection, intelligent manufacturing and other complex scenes needing high-precision and multi-view perception and semantic reasoning.

The invention relates to the technical field of data analysis, and discloses a feedback generation method and device based on a business knowledge graph, equipment and a medium, and the method comprises the steps: collecting multi-source business data, cleaning and segmenting the multi-source business data to form standardized text fragments, and carrying out the entity recognition and relation extraction of the text fragments to construct the business knowledge graph; receiving a natural language query instruction, extracting query intention features and key entity references, mapping the key entity references into the business knowledge graph, and performing association traversal under the guidance of the query intention features to obtain associated knowledge sub-graphs; and serializing the associated knowledge sub-graph into a knowledge background context, and jointly inputting the knowledge background context and a natural language query instruction into a pre-training language generation model to generate natural language feedback information. The method can be applied to business scenes such as financial science and technology, medical health and the like, semantic structuring is carried out on text knowledge, reasoning is carried out in knowledge association in combination with query semantics, accurate question answering is achieved, and knowledge retrieval efficiency is improved.

The invention discloses a knowledge destruction attack method and device based on an RAG system and a medium, and relates to the technical field of internet security, and the method comprises the steps: inputting a target question and an error answer into the RAG system, and generating an initial confrontation text; performing multiple rounds of iterative optimization processing on the initial adversarial text to obtain a target adversarial text; inputting the target adversarial text into a knowledge base corresponding to the RAG system; the RAG system responds to a question demand input by a user, and retrieves and outputs a question answer corresponding to the question demand from the knowledge base; and inputting the question demand and the question answer into a large language model, so that the large language model outputs a wrong answer corresponding to the target question. The method and the device are used for solving the problems of low output result precision, poor attack effectiveness and low concealment when knowledge destruction attack is carried out based on an RAG system in the prior art, and the precision of the output result is improved under the condition that the knowledge destruction attack is effectively carried out with high concealment.

The invention provides a personal question and answer method based on an observation-recording-decision mechanism, which comprises the following steps: an intelligent agent captures RGB images and depth images from all directions through multi-view perception, is used for constructing a 3D scene graph and mapping the 3D scene graph to a 2D semantic map, and meanwhile, the intelligent agent marks each passing position on the 2D semantic map, so that the 3D scene graph is mapped to the 2D semantic map; and dynamically updating the weight of a non-visited position, reducing the selection probability of a boundary point in a marked passing region, based on a 2D semantic map in an observation stage, distinguishing whether a question can be answered or not by an intelligent agent according to an observed RGB picture, if so, directly generating a response, otherwise, navigating to a new region, and repeating the steps until an available or maximum step number is reached, and completing the question answering. According to the method, through construction of a semantic map, weight regulation and control navigation, and fusion design of special VLM analysis and double-criterion decision making, decision making of agent non-redundancy exploration and accurate question and answer is achieved.

The invention discloses a medical visual question and answer model construction method based on a feedback loop. The method comprises the following steps: step 1, generating initial candidates; step 2, performing multi-dimensional candidate evaluation; 3, iterative optimization of feedback driving is carried out; step 4; and 5, efficient fine adjustment of model parameters: performing fine adjustment on a pre-trained multi-modal large language model by using the enhanced training data set constructed in the step 4 to obtain a final medical visual question and answer model. According to the method, a feedback loop mechanism is introduced, and cooperative and iterative self-correction and optimization are carried out on generated text interpretation (Ratio) and a visual localization region of interest (RoI), so that the fact reliability, the logic consistency and the visual strong relevance of a final output result are systematically ensured.

The invention provides a cue word optimization method and system based on a large language model, and the method comprises the steps: generating a theme corresponding to a question in a training set, the training set comprising a question-answer pair, and the question-answer pair comprising the question and a standard answer of the question; searching a first cue word corresponding to the theme in a knowledge base, wherein the knowledge base comprises a knowledge base theme and a cue word corresponding to the knowledge base theme; inputting the first cue word corresponding to the theme and the question into a large language model to obtain a response answer output by the large language model; and if the response answer is wrong compared with the standard answer, generating a new cue word based on the question, the theme, the response answer and the standard answer, and updating the first cue word corresponding to the theme into the new cue word in the knowledge base. The cue word optimization mode of the large language model is remarkably improved, and finally the overall accuracy and generalization ability of the model for answering based on the optimized cue words are improved.

The invention discloses a video anomaly detection method based on a traffic scene, and relates to the related field of computer vision, and the method comprises the steps: obtaining an input video clip which comprises a plurality of frames of images, traversing the plurality of frames of images, carrying out the selection through combining with a text prompt, determining a plurality of key frames, and enabling the text prompt to be input by a user; performing context generation based on the plurality of key frames to obtain position and time context information; the key frame, the position, the time context information and the text prompt are synchronized to a large visual language model for visual questions and answers of diversified traffic scenes, abnormal events are extracted, and the abnormal events comprise abnormal scores; and carrying out abnormal event detection analysis according to the abnormal score, and carrying out abnormal detection on the diversified traffic scenes. The technical problem that abnormal events in diversified traffic scenes are difficult to comprehensively and accurately detect in existing traffic scene-based video anomaly detection is solved, and the technical effect of improving the accuracy and generalization of traffic scene video anomaly detection is achieved.

The invention discloses an education question answering model routing method based on deep learning and a knowledge graph, and aims to solve the problems that the existing multi-model selection depends on text similarity, lacks knowledge points and first repair constraints, is easy to route by mistake and generates illusion. The method comprises the following steps: generating problem sub-graphs on an educational knowledge graph, carrying out relation perception graph neural network coding, constructing a model capability graph containing expression distribution and service attributes, forming masked cost under the constraints of types, privacy, time delay, cost and first repair coverage, and obtaining soft assignment by adopting unbalanced optimal transmission with entropy regularization; and interpretable routing is completed in combination with the meta-path attention subjected to interpretation consistency regularization training, so that the technical effects of reducing wrong routing and illusion rate, improving preferred correct routing rate and outputting meta-path interpretation are achieved.

The invention belongs to the field of crossing of remote sensing visual questioning and answering and artificial intelligence, and particularly relates to a remote sensing visual questioning and answering method based on a large language model and a multi-level attention mechanism. Comprising the following steps: 1, constructing a MogaNet-based semantic segmentation network; 2, text coding is carried out through a large language model; 3, constructing a mixed attention guiding module; 4, constructing a bidirectional gating cross attention module; and 5, constructing and cooperatively training a remote sensing visual question-answer network. According to the MLaVQA framework provided by the invention, efficient and accurate visual question-answering processing of remote sensing images can be realized, when large-scale multi-modal remote sensing data is processed, complex semantic questions can be quickly answered, fine monitoring and intelligent management of earth resources are realized, the data utilization efficiency and decision support capability are remarkably improved, and the method is suitable for large-scale multi-modal remote sensing data processing. And technical support is provided for natural resource management, ecological protection and sustainable development.

The invention discloses a big language model enhanced data consanguinity analysis method and device and a medium, and relates to the technical field of data consanguinity, and the method comprises the following steps: obtaining metadata and a source code, and carrying out data preprocessing on the metadata and the source code to obtain a graph database; based on the graph database, semantic task nodes are determined through data enhancement of LLM interaction; obtaining user query data, and performing query conversion processing on the user query data according to the semantic task nodes to obtain question answers corresponding to the user query data; obtaining abnormal warning data, and performing root cause tracing on the abnormal warning data to determine diagnosis context data; and performing multiple rounds of LLM diagnosis on the diagnosis context data to obtain a repair data stream. Through the method, the technical problem that an existing data blood relationship technology highly depends on manpower is solved.

The invention provides an agricultural production knowledge intelligent question answering method and device based on natural language processing, and the method comprises the steps: carrying out the analysis of a natural language query sequence inputted by a user based on a field adaptive semantic understanding model, and outputting a structured semantic representation and a question vector; based on the structured semantic representation and the question vector, performing retrieval matching in the dynamic agricultural knowledge graph; reasoning based on a retrieval matching result, and generating and outputting a final answer and a matching degree score of the final answer; the field adaptive semantic understanding model is obtained by pre-training a mask language model based on a term set in the agricultural field; the dynamic agricultural knowledge graph is constructed and dynamically updated based on a multi-source heterogeneous data flow in the agricultural field. Therefore, the understanding accuracy of the agricultural terminology, the retrieval timeliness and the answer accuracy are improved, and meanwhile, the reliability of the answer is enhanced through the matching degree score.

The invention discloses an intelligent test paper composition and scoring method and system based on a knowledge graph. The method comprises the following steps: S1, constructing a student knowledge graph state sub-graph; s2, constructing a question bank knowledge graph; s3, on the basis of a node embedding modeling method of an improved DeepWalk algorithm, performing joint training and low-dimensional representation learning on knowledge point nodes by utilizing a continuous bag-of-words model of multi-head context aggregation, and constructing a topic adaptation degree scoring index; s4, outputting an optimal test paper scheme set; s5, collecting a current answer record when the student completes the test paper; and S6, extracting a semantic feature vector and a behavior feature vector, constructing a fusion score feature vector, and generating scores of test paper subjective question answers. According to the invention, accurate test paper composition and multi-dimensional subjective question intelligent scoring for individual knowledge states of students are realized, the method is suitable for personalized learning evaluation and teaching feedback scenes in an education evaluation platform, and the method has the advantages of high personalization, fine feedback and automatic scoring.

The invention discloses a large video model training method and a related device, and relates to the technical field of video recognition, and the method comprises the steps: collecting a training video data frame to obtain an image frame, inputting a preset prompt word, a user question and the image frame into a large image model, and obtaining a thinking chain and a question answer. Performing cold start on the video large model based on the thinking chain and the question answer to enable the video large model to have thinking chain output capability; and combining training video data and questions to generate space and time disordered data and thinking chain data. And inputting the three types of data into the model to obtain corresponding outputs, calculating the accuracy of each output, obtaining space and time accuracy reward values, and training the model through a group relative strategy optimization algorithm in combination with a thinking chain consistency reward value to obtain an inference video large model. According to the method, the trained video large model can have thinking reasoning capability based on thinking chain implementation.

The invention discloses a retrieval enhancement generation method and system based on multi-dimensional reordering, and relates to the technical field of information retrieval, and the method comprises the following steps: S1, constructing a tensor index, a keyword index and a compression abstract; s2, driving a Qwen large model to execute query expansion and hypothetical answer generation through a dual-task query processing template; s3, mixed retrieval is executed through a semantic and keyword double-path index, and two-stage reordering is executed by using an improved DistilBERT model and a Qwen large model; s4, performing multi-dimensional evaluation through sub-problem decomposition; s5, iterating according to the sequence of the sub-questions, and gradually generating answers to the sub-questions; and S6, executing multi-dimensional evaluation correction, and generating a final answer. According to the method, the limitation of a traditional retrieval enhancement generation technology on information matching precision, context correlation evaluation and answer generation quality is overcome, and an efficient and accurate solution is provided.

To enable a manager to accurately recognize information effective for improving an organization without spending as much effort as possible in the field of improving the organization. [Solution] The information processing system has the following functions: using the functions of language model 4, to conduct specific interviews with members of an organization to elicit the organization's challenges and solutions to those challenges; a function to extract problem-solution pairs, which are combinations of problem answers that mention the challenges and solution answers that mention the solutions, from the results of the specific interviews; a function to classify each of the problem-solution pairs based on the specific interviews conducted with multiple members into one of multiple problem categories; and a function to provide information based on the problem-solution pairs belonging to each problem category to the administrator for each problem category.

The invention relates to the technical field of artificial intelligence and natural language processing, and provides a multi-level retrieval and interference document fusion-based RAG-LoRA fine tuning training method, which comprises the following steps of: constructing a domain knowledge base; performing multi-stage retrieval on the vector database to obtain a candidate document set with layered relevancy; generating a training sample based on the candidate document set, respectively selecting a target document and an interference document from the strong correlation document and the weak correlation document, and constructing the training sample containing a question, an answer and an inference chain; performing quality control on the training sample to obtain a reliable training data set for fine tuning of the target model; performing fine tuning training on the target model, and optimizing to obtain parameters of the trained LoRA adapter; the trained LoRA adapter parameters are loaded to the corresponding adaptation interface, the target model completing field adaptation is determined, the strong correlation documents and the weak correlation documents can be utilized at the same time to construct richer and diversified training contexts, and closer simulation of a real retrieval scene is achieved.

A computer-implemented method is provided to preforming bagging adversarial training for question-answer ranking models using neural networks. The method includes generating first question and answer (QA) pairs for a given question as a first training dataset to train a QA ranking model to build a pre-trained QA ranking model. A generative adversarial network (GAN) includes a generator and a discriminator configured to produce adversarial inputs to provide an updated training dataset. The pre-trained QA ranking model is retrained with the updated training dataset with the bagging adversarial training process. A plurality of trained models is sampled to generate a bagged model ensemble as a final trained QA ranking model for QA ranking tasks.

The invention relates to the technical field of natural language processing and knowledge graph crossing, in particular to an LLM thinking chain driven knowledge graph question and answer method and system, and the method comprises the steps: obtaining a knowledge graph question and answer data set and knowledge subgraphs; constructing a knowledge graph question and answer network model driven by the LLM thinking chain; applying the trained LLM thinking chain driven knowledge graph question and answer network model to a knowledge graph question and answer task; according to the method, a more stable, robust and accurate knowledge graph question and answer result can be obtained.

A large-model question answering method and apparatus for the transportation industry, and a device and a storage medium, relating to the interdisciplinary field of artificial intelligence and information security. The method comprises: on the basis of a pre-constructed security knowledge base, retrieving a transportation-related question to be processed; when the obtained retrieval result indicates that there is no reply information corresponding to the transportation-related question in the knowledge base, determining, on the basis of a preset question security interception strategy, whether to currently trigger a question interception operation; when an obtained first determination result indicates that the question interception operation is not to be triggered, performing, on the basis of a preset prompt, question enhancement on the transportation-related question; on the basis of a pre-fine-tuned target large transportation model, processing the obtained enhanced question to obtain target reply information; and on the basis of a preset answer security interception strategy, determining whether to currently trigger a reply interception operation corresponding to the target reply information, so as to complete a question answering operation on the basis of an obtained second determination result. The security and accuracy of question answering can be effectively improved.

Systems and methods are provided for a natural language question answering service to provide answers to natural language questions with respect to network-based services or computing domains. A natural language question answering service may receive a natural language question from a client computing device. An aggregator of a natural language question answer service may retrieve paragraphs from a search system and generate prompts based on questions. A Large Language Model (LLM) of a natural language question answer service may receive cues and provide answers. The answer may be verified by a verifier of the natural language question answering service. Attributes may be applied to answers and retrieved paragraphs to generate references, embedded references, and similar questions. A watermarking module of the natural language question answering service can watermark the answer under the condition that the answer is verified.

Embodiments described herein provide systems and methods for retrieval augmented generation. A neural network based language model may be provided a question as a user input. Based on the user input, semantically diverse queries may be generated for retrieval from diverse data sources. For example, a structured data source (e.g., database or knowledge base) and unstructured data (e.g., text articles) may be used to retrieve information relevant to the user input. The retrieve information may be ranked so that the most relevant information is used by the language model in generating an answer to the question in the user input. A non-retrieval based answer generated by the language model may be utilized in some embodiments in generating the final answer.

A domain-focused natural language question and answer system. The system includes one or more electronic processors. The one or more electronic processors are configured to create, from a plurality of files, a plurality of fine-tuning data examples, each respective fine-tuning data example of the plurality of fine-tuning data examples including a question, an answer, and a supporting citation. The one or more electronic processors are further configured to fine-tune a pre-trained foundational machine learning model using the plurality of fine-tuning data examples to generate a fine-tuned machine learning model, input a natural language question to the fine-tuned machine learning model, and retrieve, from the fine-tuned machine learning model, a natural language answer to the natural language question and an indication of one or more files, one or more relevant sections, or both that the natural language answer is based on.

The invention relates to the cross technical field of artificial intelligence and knowledge engineering, in particular to a local knowledge base question-answering system based on a new energy automobile and a construction method thereof, and the method comprises the following steps: integrating a knowledge graph, a ChatGLM2-6B language model and a LangChain framework to construct a local knowledge base question-answering system; obtaining a ternary collaborative architecture for structured management, retrieval and semantic question and answer of heterogeneous data in the new energy field; and based on the ternary collaborative architecture, correspondingly designing a plurality of functional modules for providing an intelligent question-answering solution for the field of new energy vehicles, and forming the local knowledge base question-answering system for the new energy vehicles. According to the'knowledge graph-pre-trained large model-LangChain 'ternary collaborative architecture provided by the invention, structured management, efficient retrieval and high-precision semantic question and answer of heterogeneous data in the field can be realized, and structural management, efficient retrieval and high-precision semantic question and answer of multi-source heterogeneous data in the new energy automobile field can be realized; and local deployment, large model generation and interpretable retrieval can be combined.

A method for embodied agent exploration is described. The method includes building a semantic map of a surrounding scene based on depth information and via visual prompting of a vision language model (VLM). The method also includes utilizing conformal prediction to calibrate a question answering confidence of the VLM. The method further includes performing, by an embodied agent, scene exploration utilizing knowledge of relevant regions of the scene. The method also includes determining, by the embodied agent, when to terminate the scene exploration utilizing a calibrated question answering confidence of the VLM.

The invention provides a financial question and answer method and device based on intelligent retrieval and a large language model. The method comprises the steps that firstly, a multi-dimensional financial knowledge base is built based on permission dimensions and knowledge dimensions; then knowledge in the knowledge base is converted into semantic vectors, and a vector index database is constructed to support accurate retrieval; when a user query is responded, identifying a user permission level and a knowledge category related to the query, decomposing the query into at least one sub-task, distributing the sub-task to the sub-Agents corresponding to the permission and the knowledge category for retrieval, and obtaining a sub-task retrieval result; and finally, the result is input into a large language model to generate an answer, and the large language model can maintain dialogue context and support multi-round questions to generate a new answer. Accurate retrieval of financial knowledge, efficient processing of complex questions and multi-round interactive question answering are achieved, the question answering accuracy, the intelligent level and the user experience are improved, and the method is suitable for policies and regulations query, business rule interpretation and other scenes in the financial field.