19 results about "First-order logic" patented technology

The invention discloses a weak supervision and thinking chain-based medical field dialogue method and apparatus, and a readable medium. The method comprises the steps of constructing a mixed training data set by using a trained first large language model and combining a first-order logic rule; performing fine tuning on the trained second large language model by using the mixed training data set and adopting a DoRA training mode to obtain a fine-tuned second large language model; inputting a dialogue history to be replied into the fine-tuned second large language model to obtain a corresponding predicted customer service intention; generating a dialogue information summary corresponding to the dialogue history to be replied by using the trained first large language model, and obtaining a second cue word for guiding reply generation according to the dialogue information summary corresponding to the dialogue history to be replied and the predicted customer service intention through a first-order logic rule; and inputting the dialogue history to be replied and the second cue word into the dialogue model to generate a corresponding customer service reply statement. According to the invention, efficient application can be realized at low cost.

The invention relates to the technical field of enterprise management, in particular to an enterprise process custom configuration method and system based on a code-free technology, and the method comprises the steps: constructing a multi-layer directed hypergraph containing data, control and condition dependence, calculating an edge weight, and converting the dependence into a first-order logic constraint set; performing incremental satisfiability detection, and outputting a conflict core set containing conflict types and severity in combination with minimum unsatisfiability subset mining and strong connected component analysis; mapping the conflicts into modifiable elements, and generating parameterized candidate repair schemes based on a repair operation library and historical cases; converting the candidate scheme into a weighted constraint optimization problem, and solving an optimal scheme by adopting a branch and bound algorithm; influence propagation verification is completed in the copy, a differential patch and a rollback identifier are generated, transactional submission is carried out after passing, and the weight and the case library are updated. According to the method, automatic diagnosis and low-risk repair of implicit dependency conflicts are realized.

The invention is suitable for the technical field of artificial intelligence and multi-modal reasoning, and provides a neural symbol reasoning method and system for multi-modal information processing, and the method comprises the following steps: obtaining multi-modal input data; respectively extracting semantic attribute representations of the image data and the text data on a predefined attribute set to obtain a multi-modal semantic attribute representation; mapping the multi-modal semantic attribute representation into a predicate in a first-order logic form, and constructing a symbolized multi-modal attribute fact set; inputting the multi-modal attribute fact set and a logic rule in a pre-constructed first-order logic knowledge base into a Markov logic network for reasoning to obtain a reasoning result; and integrating the reasoning results, and generating an interpretable decision result and a reasoning logic link. According to the method, multi-modal shared attribute symbols are taken as bridges, information fusion, conflict resolution and high-level consistency constraint among multiple modals are realized, and a reasoning result with interpretability, high robustness and strong generalization ability can be obtained.

The application relates to a city renewal space intelligent planning and evaluation system based on digital twinning, the system comprising: a graph construction module for constructing a multi-scale initial semantic knowledge graph, a regulation analysis module for converting regulation documents into a first-order logic rule and a differentiable loss function set, a reward function construction module for constructing a deep reinforcement learning agent and a comprehensive reward function, a scheme sequence generation module for generating a candidate planning scheme sequence, and a planning scheme generation module for screening a final planning scheme through a structural causal model. The system can integrate city renewal multi-source heterogeneous data and regulation constraint knowledge, realize city renewal space intelligent planning and scientific evaluation, and support precise decision-making for high-quality development of cities in the stock era.

The application discloses a kind of multi-modal knowledge distillation methods and systems for legal consultation, comprising: obtaining large legal professional model and multi-modal legal consultation data, using information flow analysis method and multi-round dialogue evaluation model professional ability, using first-order logic formalism and comparative learning framework to refine legal rule knowledge base;Based on the knowledge base, use the confusion perception contrast learning mechanism to identify small model confusion points, use legal case trajectory tree exploration strategy to analyze complex case reasoning path, dynamically optimize legal rule set;Construct a distillation loss function containing rule compliance loss term, integrate multi-perspective thinking and metacognition awareness training strategy for knowledge distillation;Lightweight processing is carried out to small model after distillation and deployed to terminal equipment.The application realizes the localization, privacy protection legal consultation service, effectively solves the problem of insufficient professional application of traditional multi-modal knowledge distillation technology in the legal field.

The invention discloses a water transportation system interaction method based on semantic representation, and belongs to the technical field of transportation. Comprising the following steps: S1, analyzing basic rules of a water transportation system, and constructing a typical water transportation scene suitable for first-order logic expression; s2, extracting basic features of the water transportation scene in the step S1, splitting scene rules, determining a scene vocabulary, establishing a first-order logic semantic expression, converting a natural language into a logic formula which can be analyzed by a machine, and providing a scene knowledge basis for automatic decision making of a ship; and S3, constructing a semantic representation knowledge base according to the scene knowledge in the step S2, and simulating an actual scene by using a programming language to realize accurate interaction between the water traffic intelligent agents. According to the invention, the channel for safe passing can be accurately given. The interactive language design method has high reliability and accuracy in intelligent agent interaction of a water transportation system.

To improve the technological process of computerized answering of logical queries over incomplete knowledge bases, obtain a first order logic query; with a trained, computerized neural network, convert the first order logic query into a logic embedding; and answer the first order logic query using the logic embedding.

The application provides a method and system for automatic proof of SQL equivalence based on linear integer algebra, comprising: a plan generation step of parsing an input SQL query statement to generate an original logical plan; an ORDER BY processing step of processing an ORDER BY sorting operation in the original logical plan to generate a new logical plan without the ORDER BY sorting operation, which is submitted to a subsequent step for equivalence verification; a U-expression generation step of parsing the logical plan to generate a corresponding algebraic expression U-expression; a normalization step of normalizing and simplifying the U-expression to generate a standard form U-expression; a LIA * conversion step of generating a LIA * expression according to the standard form U-expression; a first-order logic expression generation step of generating a LIA expression according to the LIA * expression, and performing automatic verification of SQL equivalence by means of an SMT solver. The application perfects modeling of SQL features such as an aggregation function, and significantly enhances the capability of automatic verification of SQL equivalence.

The invention provides an anomaly detection method. The method comprises the following steps: converting to-be-detected original data into a first-order logic FOL set by adopting a large language model LLM; extracting a minimum unsatisfactory subset (MUS) of the FOL set by adopting an SMT solver, wherein the MUS is used for indicating mutually conflicting FOLs; and performing conflict set analysis based on the MUS to obtain abnormal points of the original data. The invention further provides electronic equipment, a computer readable medium and a computer program product.

The invention relates to a method and a system for determining a set of output configuration values characterizing a dedicated configuration of a complex product, the method comprising: a) receiving (201) a set of input configuration parameters; b) providing (202) at least a part of the input configuration parameters as input to a solver; c) the solver computing (203) at least one output value from the provided input configuration parameters; d) determining (205) a set of output configuration values from at least said output value computed by the solver; characterized in that the solver is configured for solving a first-order logic function encoding an algorithm of a trained deep neural network (hereinafter DNN), wherein the algorithm of the DNN has been trained for modeling a function T of an external configuration tool (hereinafter ECT), the function T of the external configuration tool being required for determining a dedicated configuration of a complex product.

The present disclosure relates to sequential and differential processing of stacked knowledge bases to retrieve a subset of knowledge bases and generate outputs personalized for a condition of a subject, based on an input dataset received via an interface accessible through a user endpoint, by leveraging a large language expert (LLE). The techniques, as disclosed herein, may dynamically generate task-specific queries corresponding to variables associated with each action represented in a knowledge base of the subset of knowledge bases. In response to execution of the task-specific queries on supplemental data associated with the subject, a result set comprising predicted intermediate results may be generated by leveraging a large language model (LLM). The disclosed techniques may further generate a workup trajectory through progressively refining the actions through application of associated first-order logics based on the generated result set and output the workup trajectory on the interface to guide subject’s pre-treatment workup.

This invention relates to the fields of artificial intelligence and multimodal reasoning technology, and provides a neural symbolic reasoning method and system for multimodal information processing. The method includes the following steps: acquiring multimodal input data; extracting semantic attribute representations of image data and text data on a predefined attribute set to obtain multimodal semantic attribute representations; mapping the multimodal semantic attribute representations to predicates in first-order logic form to construct a symbolic multimodal attribute fact set; inputting the multimodal attribute fact set and logical rules from a pre-constructed first-order logic knowledge base into a Markov logic network for reasoning to obtain a reasoning result; integrating the reasoning result to generate an interpretable decision result and a reasoning logic chain. This invention uses shared attribute symbols across multiple modalities as a bridge to achieve information fusion, conflict resolution, and high-level consistency constraints among modalities, resulting in reasoning results with interpretability, high robustness, and strong generalization ability.

The application discloses a medical field dialogue method and device based on weak supervision and thought chain and readable medium, comprising: constructing a mixed training data set by using a trained first large language model and combining a first-order logic rule; fine-tuning a trained second large language model by using the mixed training data set and adopting a DoRA training method to obtain a fine-tuned second large language model; inputting a dialogue history to be replied into the fine-tuned second large language model to obtain a corresponding predicted customer intent; generating a dialogue information summary corresponding to the dialogue history to be replied by using the trained first large language model, and obtaining a second prompt word for guiding reply generation by the dialogue information summary corresponding to the dialogue history to be replied and the predicted customer intent through the first-order logic rule; and inputting the dialogue history to be replied and the second prompt word into a dialogue model to generate a corresponding customer reply sentence. The application can realize efficient application at low cost.

The invention belongs to the technical field of knowledge base data mining, and relates to a data structure and method for optimizing first-order logic rule mining, and the data structure comprises a rule cache, a cache block pool and a block calculation lookup table. According to the rule cache, an expression mode of a variable replacement set in a single logic rule is divided into a plurality of cache segments to implicitly express a Cartesian product among the segments, so that the memory occupation of the cache is greatly reduced. The cache block pool optimizes the related operation of variable modification in the first-order logic, and improves the performance of index construction and related operation. The block calculation lookup table maximizes the reuse rate of the cache blocks in the cache block pool, records the same cache block operation in the table to avoid repeated operation, and further reduces the memory occupancy rate while improving the calculation efficiency.

The present disclosure relates to a method and device for determining a power transaction decision. The method comprises: fuzzy matching an original transaction strategy with each rule of a first-order logic rule base to determine a fuzzy satisfaction degree of each rule; in response to the fuzzy satisfaction degree of at least one rule being less than a predetermined threshold, performing constraint optimization correction on the original transaction strategy to obtain a compliance transaction strategy; constructing a structural causal model of the power transaction, and using the model to perform counterfactual reasoning based on the compliance transaction strategy and a preset counterfactual question template to determine a normalized regret value; determining a risk boundary of a transaction position based on a price variation coefficient and the normalized regret value, and updating a Lagrange multiplier corresponding to each rule; and determining the power transaction decision based on the risk boundary, the updated Lagrange multiplier, and the compliance transaction strategy. Thus, the complex rule constraints of the power market can be effectively met, and aggressive transactions and real-time needs of spot transactions can be avoided.

The invention discloses a legal consultation-oriented multi-modal knowledge distillation method and system, and the method comprises the steps: obtaining a large-scale legal professional model and multi-modal legal consultation data, employing an information flow analysis method and a multi-round dialogue to evaluate the professional ability of the model, and employing a first-order logic form and a comparative learning framework to extract a legal rule knowledge base; based on the knowledge base, using a confusion perception comparison learning mechanism to identify confusion points of the small model, using a law case trajectory tree exploration strategy to analyze a complex case reasoning path, and dynamically optimizing a law rule set; constructing a distillation loss function containing a rule conformity loss item, and fusing multi-view thinking and meta-cognitive consciousness training strategies to perform knowledge distillation; and carrying out lightweight processing on the distilled small model and deploying the small model to terminal equipment. The legal consultation service with localization and privacy protection is realized, and the problem of insufficient professionality of the traditional multi-modal knowledge distillation technology in the legal field is effectively solved.

A computer-implemented procedure for answering a first-order logical query based on a knowledge graph, wherein the procedure comprises: providing (302) the knowledge graph and the first-order logical query, wherein the logical query defines a logical operation on a first triple and a second triple, wherein the first triple comprises a cone embedding representing a first set containing at least one entity of the knowledge graph, a cone embedding representing a second set containing at least one entity of the knowledge graph, and an embedding of a rotation operation between the two cone embeddings of the first triple in a complex embedding space, wherein the second triple comprises a cone embedding representing a third set containing at least one entity of the knowledge graph, a cone embedding representing a fourth set containing at least one entity of the knowledge graph,and comprises embedding a rotation operation between the two conical embeddings of the second triple in the complex embedding space, wherein one of the conical embeddings represents an answer to the logical query, determining (304) a directed acyclic computation graph for the logical query depending on the logical query, wherein the computation graph defines an order for performing the rotation operations, determining (306) the answer to the logical query, wherein determining the answer comprises applying the rotation operations to the conical embeddings in accordance with the order defined by the computation graph.

A computer-implemented method for querying by a query engine on a database including relational data and graph data, the method including receiving a query by the query engine, the query being in a language compatible with a first order logic paradigm. The method further includes generating an intermediate representation, the IR being compatible with a first order logic paradigm. The method further includes using the intermediate representation of the query for executing the query on the relational data and the graph data of the database.