58 results about "Inference tree" patented technology

First order predicate logics are provided, extended with a bilattice based uncertainty handling formalism, as a means of formally encoding pattern grmmars, to parse a set of image features, and detect the presence of different patterns of interest implemented on a processor. Information from different sources and uncertainties from detections, are integrated within the bilattice framework. Automated logical rule weight learning in the computer vision domain applies a rule weight optimization method which casts the instantiated inference tree as a knowledge-based neural network, to converge upon a set of rule weights that give optimal performance within the bilattice framework. Applications are in (a) detecting the presence of humans under partial occlusions and (b) detecting large complex man made structures in satellite imagery (c) detection of spatio-temporal human and vehicular activities in video and (c) parsing of Graphical User Interfaces.

The invention provides a case judgment result determination method, device and equipment and a computer readable storage medium, and the method comprises the steps: obtaining the case judgment data ofa target case, and carrying out the information extraction of the case judgment data, so as to determine the case knowledge information of a to-be-constructed legal knowledge graph; constructing a legal knowledge graph of the target case according to the case knowledge information and a preset legal provision knowledge base; and determining a case judgment inference tree of the target case according to the legal knowledge graph, and determining a case judgment result of the target case according to the legal knowledge graph and the case judgment inference tree. The invention relates to data analysis and a knowledge graph, and the accuracy of a case judgment result is effectively improved.

The invention relates to the technical field of natural language processing, and discloses a structure prompt-based reasoning collaborative SQL generation method. According to the method, a small model carries out structured analysis on natural language query and a target database mode, SQL substructures and combination sequence prompts are generated, a large model constructs a multi-path inference tree based on the prompts, and path nodes sequentially generate SQL substructure fragments. Top-k paths are screened through path node scores and path cumulative scores, comprehensive sorting is performed by applying a heuristic scoring function, and candidate SQL with the highest score is selected. And verifying and correcting the candidate SQL according to execution feedback, and outputting a final SQL result after ensuring that semantics and structures are correct. According to the method, the problems of insufficient reasoning ability of a small model, poor reasoning stability and illusion of a large model and the problems of Tree-of-Though reasoning path redundancy and frequent focus switching are solved, the reasoning reasonability and execution efficiency in a Text2SQL task are improved, and the method is suitable for an application scene from a complex natural language to structured query.

The invention provides a code large model self-evolution method based on Monte Carlo tree search, which comprises the following steps: S1, inputting a code generation task into a pre-training reasoning model, and generating a reasoning step according to a reasoning task; s2, constructing an inference tree, and obtaining an optimal inference path and a reward score according to an improved Monte Carlo tree search algorithm; s3, according to the optimal reasoning path, based on the automatically generated reasoning and reward data, performing pre-training model self-evolution to improve the reasoning model; and S4, enabling the pre-training model to generate a code of the reasoning scheme, and training the pre-training model to obtain a code large model. According to the method, under the condition that a large amount of fine adjustment data is not needed, the capacity and accuracy of large model code generation can be effectively improved, and the adaptability of the model in a complex reasoning task is further enhanced.

The invention provides a method and a system for a positioning fault. The method comprises the following steps: receiving alarm information; determining an inference tree to which the received alarm information belong, wherein a rule set which can be identified by the system and used for inferring the fault is stored in the inference tree; positioning the fault by the determined inference tree. By virtue of the method and the system for positioning the fault, the alarm can be quickly restrained, so that the efficiency of positioning the fault is effectively improved.

The invention provides an intelligent conversation recommendation method and device based on a knowledge graph. The method comprises the following steps: collecting voice information of a user side; the voice information is input into a trained language model, a natural language reply corresponding to the voice information is obtained, and the trained language model is obtained by training a language model through sample conversation historical content and a sample intention inference tree; wherein the sample intention reasoning tree is obtained by reasoning based on a knowledge graph according to a preset conversation intention, a sample conversation historical representation and a sample user portrait; the sample conversation history representation is obtained by encoding the sample conversation history content; the sample user portrait is obtained based on a self-attention mechanism by encoding nodes of the knowledge graph mentioned in the historical content of the sample conversation. According to the method, modeling is carried out on entity reasoning processes under different conversation intentions, so that the accuracy of intelligent conversation recommendation and the smoothness of language generation are effectively improved.

A computerized method has the steps of: generating one or more queries from an input question; generating one or more outputs; and outputting an answer based on the one or more outputs. Said generating the one or more outputs has the steps of: for each query, forming a reasoning tree with the query being a root node and a current node, generating one or more candidates as leaf nodes of the current node, by inputting a reasoning path from the root node to the current node into a foundation model, searching the reasoning tree using an artificial intelligence model to select a leaf node as the current node, and repeating said generating the one or more candidate nodes and said searching the reasoning tree until a termination condition is met, and using an updated reasoning path from the root node to the current node as the output for the query.

The invention discloses a model reasoning method and device based on a visual language model, and relates to the technical field of artificial intelligence. The method comprises the following steps: encoding a to-be-reasoned text and a to-be-reasoned image to generate an initial reasoning tree; based on the feasibility parameters of the reasoning action in the to-be-reasoned image and the state parameters of the current node, performing expansion updating processing on the initial reasoning tree to generate a target reasoning tree; based on the rationality parameters of all the end nodes in the target reasoning tree, determining candidate reasoning paths; and determining a target reasoning path according to the consistency parameter of the intermediate node and the rationality parameter of the end node in the candidate reasoning path, and finally outputting target response data through the target reasoning path. The reasoning efficiency and accuracy of the visual language model in a complex reasoning task can be improved.

The invention provides a big language model training method, a big language model reasoning method and device, electronic equipment and a storage medium, and relates to the technical field of artificial intelligence, in particular to the technical field of deep learning, big language models, natural language processing and content auditing. According to the specific implementation scheme, for input content, target nodes corresponding to the input content in an inference tree are determined, and the inference tree comprises multiple levels of nodes; according to the input content and the target node, thinking content, a thinking result and a reasoning path for reasoning the input content are generated, the thinking result is a target child node determined from multiple candidate child nodes of the target node, and the reasoning path is a path from the target node to the target child node; determining the correctness of the reasoning path according to the thinking content and the thinking result; and in response to determining that the reasoning path is incorrect, generating updated thinking content, thinking result and reasoning path according to the input content and the target node.

The invention provides a post-earthquake function evaluation method for a railway station. The method comprises the following steps: determining a relationship between all components forming a transportation distribution function in a railway station and a functional system formed by the components; establishing a railway station function reasoning tree based on the confidence rule base; based on the railway station function inference tree, establishing a confidence rule base to quantify the relationship between each component and the railway station function; calculating the seismic response of the railway station in the target seismic scene for the railway station needing function evaluation, and judging the damage condition of the component according to the vulnerability of the component, or investigating the damage condition of the post-earthquake component in the field; and according to the actual passenger flow of the target railway station, weighting the components and the function system, substituting the weighted component damage condition into the confidence rule base, and performing function evaluation on the post-earthquake railway station. According to the method, the function of the railway station can be evaluated immediately after the earthquake, then the available condition of the station after the earthquake and the capacity of collecting and distributing passengers are rapidly determined, and potential operation risks and function waste are avoided.

The invention discloses a generating method of a scheduling operating command ticket, which comprises the steps of: growing and trimming an intelligent inference tree, wherein the growth comprises the steps of: 1, starting, 2, selecting an operating device and 3, ending, and the step 2 comprises the steps of: 2.1, generating a target node; 2.2, adding one group of attribute nodes for the target node; 2.3, traversing to obtain an attribute node; 2.4, judging whether the attribute node exists or not; 2.5, executing the attribute node and carrying out drop-out analysis and load flow calculation on relevant operations; 2.6, checking whether the drop-out, the error prevention and the load flow are safe or not, if not meeting the requirements, executing the step 2.7, if meeting the requirements, executing the step 2.3; 2.7, adding one group of trigger nodes for the attribute node; 2.8, traversing the trigger nodes; and 2.9, checking whether the attribute node exists the trigger nodes or not , if not, executing the step 2.3, and if so, executing the step 2.1.

The invention provides a remote sensing image interpretation method based on large model reasoning and tool enhancement, which comprises the following steps: generating an initial reasoning node of a reasoning tree by using a large language model as an active decision maker based on system cue words and input remote sensing image features; based on the current inference tree state, the active decision maker executes inference and generates a new inference node, so that the inference tree is expanded; based on node scores, adopting a pruning strategy to eliminate reasoning paths with low scores; repeating the steps until a reasoning termination condition is met, and selecting a path with the highest accumulated score from the reasoning tree as an optimal reasoning path; and key information is extracted and organized into a structured intelligence report to be output. According to the method, a complete interpretation chain of'thinking-tool-observation 'is explicitly modeled through a tree reasoning mechanism, so that the generation logic of each intelligence conclusion is clear and visible, and the problems of trust crisis and insufficient interpretability caused by'black box' decision of a traditional deep learning model are solved.

The invention relates to a data processing method and device based on a large language model, equipment and a storage medium. The method comprises the following steps: acquiring input data, matching the input data with a plurality of pre-constructed data matrixes to obtain a data matrix matched with the data type of the input data, and determining the matched data matrix as a current data matrix; constructing an inference tree model based on the input data and the current data matrix; determining a plurality of candidate sequences based on the inference tree model; and performing parallel verification on the plurality of candidate sequences, determining the output probability distribution of each candidate sequence, and obtaining an output sequence matched with the data content of the input data according to the output probability distribution. Compared with the prior art, the data processing efficiency can be improved, and the data output quality and accuracy can also be improved.

The invention provides a large language model heterogeneous knowledge reasoning method and device based on an atomic knowledge operator, and the method comprises the steps: building an atomic reasoning tree based on an original to-be-answered question; wherein the atomic inference tree comprises father nodes and leaf nodes, root nodes in the father nodes are original to-be-answered questions, non-root nodes in the father nodes are different levels of sub-questions obtained by decomposing the original to-be-answered questions, the leaf nodes are atomic questions obtained by decomposing the original to-be-answered questions, and each leaf node corresponds to an atomic knowledge operator; the atomic knowledge operator comprises a search operator, a relation operator and a filtering operator; for leaf nodes, reasoning atomic problems according to retrieval-knowledge on the basis of a large language model and an atomic knowledge operator; for a father node, based on a large language model, according to child node reasoning answers, brother node reasoning answers or retrieval-knowledge, reasoning child questions; reasoning the atomic reasoning tree from a leaf node to a root node from bottom to top to obtain a reasoning result of the original to-be-answered question; wherein the retrieval-knowledge is obtained by executing dynamic knowledge retrieval from various heterogeneous knowledge sources. According to the method, an original to-be-answered question is decomposed into atomic-level fine granularity, a leaf node in an atomic reasoning tree corresponds to an atomic knowledge operator, a more accurate reasoning result is achieved, meanwhile, knowledge is allowed to be dynamically retrieved from various heterogeneous knowledge sources at each sub-question and atomic question node, and the reasoning efficiency is improved. Different types of queries can be flexibly coped with, richer, more accurate and more complementary information is provided, and the overall reasoning ability of the algorithm is enhanced.

The invention discloses a multi-agent medical intelligent diagnosis method and system based on an inference tree structure, electronic equipment and a storage medium. The method comprises the steps that a large language model is commanded to simulate multiple specialist physician roles through preset cue words; defining an inference tree structure, and requiring each agent to output a diagnosis result according to the structure; each agent analyzes the medical examination data of the corresponding field to generate a diagnosis result based on an inference tree structure; designing a multi-agent cross validation mechanism, and updating an inference tree through pairwise discussion of agents; and fusing the reasoning trees of all agents, and outputting final diagnosis including a diagnosis conclusion, a reasoning process and clinical evidence. According to the method, medical examination data in different fields are analyzed by utilizing a multi-agent method based on an inference tree structure, so that the diagnosis performance of a complex medical scene is improved.

The invention relates to a multi-hop RAG question and answer method and system based on path exploration and hyperbolic refining. The method and system are used for solving the problems of reasoning path stiffness and evidence noise interference in the multi-hop question and answer process. The method comprises the steps that complex multi-hop query is received, a multi-level path exploration search strategy is adopted, a hierarchical inference tree is constructed through Monte Carlo tree search, and atomic operations such as direct answering, retrieval answering and sub-question decomposition are sequentially executed to explore an optimal inference path; performing hyperbolic refining on candidate evidences retrieved in the construction process, constructing an evidence graph and performing information spreading by using a graph attention network, thereby performing dynamic reordering and noise suppression on the evidences; and recursively aggregating the processed intermediate question and answer results from bottom to top, and generating a final answer by combining an original query calling language model. According to the method, the continuity and accuracy of multi-hop reasoning are effectively improved, and the method is suitable for natural language processing application such as complex questions and answers and knowledge reasoning.

The invention provides a method, a system and equipment for generating a mathematical problem reasoning trajectory based on a large model, and relates to the technical field of artificial intelligence. According to the method, a target mathematical problem is used as a root node, a multi-level inference tree is constructed, a Monte Carlo tree search algorithm is used for carrying out path search, an inference track of the target mathematical problem is searched, and in the search process, a plurality of candidate solution steps on each level are used for solving the target mathematical problem. Predicting the exploration probability of each candidate answering step on the hierarchy by adopting a probability route map; and based on the predicted exploration probability of each candidate solution step, selecting a target solution step as a sampling node to enter the search of the next level, and finally generating a reasoning track of a target mathematical problem and providing the reasoning track to a user. According to the embodiment of the invention, through the guidance of the probability route map and the search strategy of the Monte Carlo tree search algorithm, the reasoning trajectory of the high-quality target mathematical problem can be quickly found in the first round of iteration, and the generation efficiency and quality of the reasoning trajectory are remarkably improved.

The invention provides an inference tree generation method based on a clustering algorithm and a large model, and belongs to the technical field of large models. Extracting hyperedges and entities from the unstructured text data by using the large model; clustering hyperedges by using a hypergraph modularity clustering algorithm; the method comprises the following steps: extracting a keyword of a problem by utilizing a large model, searching a candidate seed hyperedge set according to vector similarity of the keyword and hyperedges, taking out a seed hyperedge with the highest similarity as a starting point of an inference tree, searching an adjacent hyperedge through entity association, and preferentially selecting a hyperedge belonging to the same cluster as the hyperedge added into the inference tree last time for expansion; and when the inference tree is updated, constructing a context containing hyperedges, entities and text blocks corresponding to the hyperedges, and submitting a large model to judge to continue expansion or directly give answers. According to the method, redundant branches of the inference tree are reduced through the clustering algorithm, whether expansion is terminated or not is judged through the large model, the size of the inference tree can be well controlled, and efficient generation of the inference tree for graph retrieval is effectively achieved.

The present invention discloses a model reasoning method and device based on a visual language model, and relates to the field of artificial intelligence technology. The method comprises: encoding a text to be reasoned and an image to be reasoned, and generating an initial reasoning tree; based on the feasibility parameters of the reasoning action in the image to be reasoned and the state parameters of the current node, the initial reasoning tree is extended and updated to generate a target reasoning tree; based on the rationality parameters of all end nodes in the target reasoning tree, a candidate reasoning path is determined; according to the consistency parameters of the intermediate nodes in the candidate reasoning path and the rationality parameters of the end nodes, the target reasoning path is determined, and finally the target response data is output through the target reasoning path. The present invention can improve the reasoning efficiency and accuracy of the visual language model in complex reasoning tasks.

The invention relates to the field of insurance claim settlement, in particular to an intelligent contract-driven insurance claim settlement automation system. The method comprises the following steps: constructing an original data matrix based on collected claim settlement data and non-claim settlement data, and carrying out decentralization and data standardization processing to obtain a standardized data matrix; constructing an adaptive logical reasoning tree model based on the standardized data matrix of the claim settlement data, and generating an output vector of an adaptive logical reasoning tree node; the condition threshold matrix is adjusted, the rule set matrix is evolved, then the total amount of claim settlement is calculated, and the adaptive logical reasoning tree model is evaluated and optimized; and calculating a final total amount of claim settlement based on the standardized data matrix of the non-claim settlement data. The problems that the rule set updating and model optimization process is rigid, and the flexibility and adaptability of claim decision making are insufficient are solved; data standardization and feature extraction lack robustness and accuracy; when a system of a fixed rule set processes multi-dimensional and high-complexity requests, decision deviation occurs.

The present application provides a kind of based on multi-model coupling sound scene ecological red line quantitative demarcation method and system.The present application constructs the cascade detection mechanism of generalized additive model and segmented regression combination, can accurately identify the ecosystem cliff type degradation node;Supporting scale response curve and random forest-variance inflation factor joint dimensionality reduction strategy, on the basis of eliminating multicollinearity interference, lock core driving factor, avoid the blindness of artificial specified spatial scale;And introduce non-parametric conditional inference tree to build multi-model cross-validation system, by calculating the deviation rate intelligent discrimination environmental heterogeneity, adaptive output rigid control boundary point or elastic early warning buffer zone, with multi-level dynamic backtracking mechanism realizes different habitat and data characteristics under multi-scene coverage, fully safeguards the algorithm coherence, result robustness and dynamic fault tolerance ability, for sound scene ecological red line scientific demarcation, accurate control and dynamic supervision provides stable and reliable technical support.

The application provides a code large model self-evolution method based on Monte Carlo tree search, comprising the following steps: S1, inputting a code generation task into a pre-training inference model, and generating an inference step according to the inference task; S2, constructing an inference tree, and obtaining an optimal inference path and a reward score according to an improved Monte Carlo tree search algorithm; S3, performing self-evolution of the pre-training model based on automatically generated inference and reward data according to the optimal inference path, and improving the inference model; and S4, enabling the pre-training model to generate code of an inference scheme, training the pre-training model, and obtaining a code large model. The application can effectively improve the code generation capability and accuracy of the large model without a large amount of fine-tuning data, and also enhances the self-adaptability of the model in a complex inference task.

The present disclosure provides a large language model training method, a large language model inference method, an apparatus, an electronic device and a storage medium, relates to the technical field of artificial intelligence, and in particular to the technical field of deep learning, large language model, natural language processing and content review. The specific implementation scheme is: for input content, a target node corresponding to the input content in an inference tree is determined, wherein the inference tree includes multiple levels of nodes; according to the input content and the target node, thinking content, thinking results and inference paths for inferring the input content are generated, wherein the thinking results are target child nodes determined from multiple candidate child nodes of the target node, and the inference paths are paths from the target node to the target child nodes; the correctness of the inference path is determined according to the thinking content and the thinking results; and in response to determining that the inference path is incorrect, updated thinking content, thinking results and inference paths are generated according to the input content and the target node.

The application provides a search recommendation method based on a weighted directed reasoning tree set generated by reasoning track generation, and relates to the technical field of search recommendation. The method comprises the following steps: collecting historical text data under a specific business scenario according to user demand and extracting semantic information, using the extracted semantic information to construct a weighted directed reasoning tree set with semantic information and containing complete reasoning tracks; using a vector space mapping and reasoning path fusion method to convert the weighted directed reasoning trees in the weighted directed reasoning tree set into vector representations; using a large model as a strategy model, using a large model-based reasoning track generation method to generate complete reasoning paths, constructing a preference data set, and using a direct preference optimization algorithm to optimize the strategy model; obtaining a search term provided by a user and inputting the search term into the optimized strategy model to obtain a final recommendation result of the search term, so as to provide more personalized and high-quality search results and recommendation contents for the user.

The invention discloses an interpretable dispatching automation system fault root cause positioning method and system, and the method comprises the following steps: (1) collecting index data from a system, and constructing an index set; constructing a system software and hardware topology according to the system architecture and the call chain log; (2) defining an event reasoning tree, wherein the tree comprises an index exception event, a system fault event, an index exception detection operator and a system fault reasoning operator; (3) performing anomaly detection on the index set by using the event inference tree, inferring an index anomaly event, and further inferring a system fault event; and (4) positioning the root cause according to the software and hardware topology of the system and the fault event of the system. According to the method, the problem of massive alarm storms of an existing index anomaly alarm scheme and the problem of low interpretability of a machine learning classification scheme are solved, the availability of an operation and maintenance system is improved, and safe and stable operation of a power grid dispatching automation system is guaranteed.

The invention relates to a digital multimedia evidence obtaining method and device. The digital multimedia evidence obtaining method comprises the steps that a multi-modal knowledge graph is constructed based on multi-modal media samples, and the multi-modal media samples comprise image frames, audio signals and text subtitles; identifying potential forgery risk points from the multi-modal media sample, and generating a forgery task graph for the potential forgery risk points; performing multi-modal feature extraction on the multi-modal media sample based on the forgery task graph to obtain a multi-granularity forgery clue set; structuring the multi-granularity forged clue set into knowledge entries; chain logic reasoning is carried out based on the multi-modal knowledge graph and the knowledge entries, and cross-modal consistency scores are calculated to generate a reasoning tree; determining a final forgery credibility score based on the confidence of each node in the inference tree; and based on the final forgery credibility score, determining whether forgery content exists in the multi-modal media sample. With the adoption of the method and the device, high-precision and high-robustness forged content identification can be realized.

This invention proposes a remote sensing image interpretation method based on large-model reasoning and tool enhancement, comprising the following steps: Based on system prompts and input remote sensing image features, a large language model is used as an active decision-maker to generate initial reasoning nodes for the reasoning tree; based on the current state of the reasoning tree, the active decision-maker performs reasoning and generates new reasoning nodes, thereby expanding the reasoning tree; based on node scores, a pruning strategy is used to eliminate reasoning paths with low scores; the above steps are repeated until the reasoning termination condition is met, and the path with the highest cumulative score is selected as the optimal reasoning path from the reasoning tree; key information is extracted and organized into a structured intelligence report for output. This invention explicitly models the complete interpretation chain of "thinking-tool-observation" through a tree-like reasoning mechanism, making the generation logic of each intelligence conclusion clearly visible, thus solving the trust crisis and insufficient interpretability problems caused by the "black box" decision-making of traditional deep learning models.

The invention relates to the technical field of aero-engine fault data isolation processing, in particular to an aero-engine fault data processing method and system which can achieve good balance between time efficiency and generation quality and is good in universality. Then fine semantic segmentation and efficient vectorization index construction of the text are completed through a domain self-adaptive large language model and a dynamic code execution mechanism; secondly, a two-stage progressive retrieval enhancement generation architecture TSP-RAG is provided, meanwhile, a self-adaptive HNSW retrieval module dynamically adjusts the search breadth according to the obtained ef weight, vector retrieval is executed, content-level deduplication is conducted on a result, finally, semantic fusion and reordering are conducted on nodes retrieved by the FIL-knowledge graph and the HNSW retrieval result, and the HNSW retrieval result is obtained. Compared with the prior art, the method has the advantages of being high in time efficiency, high in generation quality, good in universality and the like.

The invention discloses a question and answer task processing model training method and device, equipment and a medium, and relates to the technical field of natural language processing, and the method comprises the steps: extracting a plurality of key structural elements from a training sample question, and guiding a current question and answer task processing model to generate a corresponding initial reasoning path by using each key structural element; taking each initial reasoning path as a root node, determining a bifurcation node based on the information entropy of each lexical element position in the initial reasoning path, and then generating a branch reasoning path at the bifurcation node to obtain an extended reasoning tree; calculating a node value of each node in the extended reasoning tree, and further calculating a global advantage, a local advantage and a path residual error; and training the current question and answer task processing model by using a reward value determined based on the global advantage, the local advantage and the path residual, and finally obtaining a trained target question and answer task processing model through multi-round iteration, so as to process the obtained question and answer task to be processed by using the target question and answer task processing model.