8 results about "Concept space" patented technology

An ideation platform device and method using a diagram are disclosed. An ideation platform device using a diagram, according to one embodiment of the present invention, can comprise: a C-K canvas module for providing a C-K canvas divided into a concept space and a knowledge space and connecting a concept and knowledge to each other on the C-K canvas through a chaining process so as to help a solution search for resolving a problem; and an instance management module for storing and managing, as one instance, the C-K canvas, for which a solution search is completed, including the concept, the knowledge, and information about an interconnection relationship.

The application provides an overlapping clustering method and system based on hypergraph covering and a medium, relates to the technical field of complex networks, and the method comprises the following steps: acquiring a sample set, constructing at least two types of relation view hypergraphs, the hypergraphs comprising a semantic neighborhood hypergraph and a factor-induced hypergraph; representing the samples to a concept space shared by the semantic neighborhood hypergraph and the factor-induced hypergraph, obtaining a concept covering representation of the samples; determining an adaptive overlapping budget interval of each sample according to the local density of each sample and the cross-relation divergence in different views; constructing a joint optimization objective function, alternatingly and iteratively optimizing the joint optimization objective function; determining a final concept set to which each sample belongs according to the concept covering representation and the adaptive overlapping budget interval of each sample after optimization, and outputting an overlapping clustering result. The application solves the problems of sample overlapping attribution description, high-order relation fusion utilization and high-dimensional sparse data processing efficiency.

The invention relates to a DIKWP aware operating system (AI-OS) and a security framework. The invention provides an artificial consciousness operating system and a security framework based on a DIKWP (Data-Information-Knowledge-Intelligence-Intention) model, and aims to solve the problems that the decision-making process of an existing AI (Artificial Intelligence) system is black, the existing AI system does not have self-cognition, the security is uncontrollable and the like. According to the operation system, the cognitive process of AI is divided into five stages of data processing, information extraction, knowledge application, intelligent decision making and intention management, and through core components such as a concept-semantic fusion kernel, a white-box evaluation module, a semantic security protection module and an intention regulation and control interface, an intellectual property of the AI is evaluated. Semantic checking, whole-process monitoring and purpose constraint of an AI internal cognitive process are realized. Wherein the kernel adopts a concept space and semantic space double-layer verification mechanism to improve AI semantic understanding consistency, the white box module records AI full-link reasoning to achieve interpretable auditing, the safety protection module is embedded into ethical rules to intercept violation output in real time, and the intention interface directly acts human high-level intentions on the AI decision process. Through the architecture, the decision-making process of the AI system is transparent and controllable, the output behavior is consistent with the preset target and value criterion, and the interpretability, safety and reliability of the AI system are remarkably improved.

The invention discloses a concept and semantic double-loop iterative problem solving method and system oriented to uncertain information. For the problems of incomplete, inconsistent and inaccurate information, a concept space and semantic space dual-circulation mechanism is constructed. Firstly, a preliminary problem model is established in a concept space based on uncertain information, and then the logic consistency, completeness and accuracy of the preliminary problem model are verified in a semantic space. If a problem is found, an intelligent layer inference engine generates a modification instruction, new information is introduced, a hypothesis or an adjustment structure is put forward for modification, and concept space iterative optimization is returned. The cycle of modeling-verification-correction lasts until the model meets the semantic consistency standard. And finally generating a solution based on the optimization model. According to the method, the'hypothesis-verification-correction 'cognitive process of human beings is simulated, and the robustness, reliability and interpretability of problem solving in a complex and uncertain environment are remarkably improved.

The invention discloses a multilingual semantic intention matching system for AI-assisted recruitment, and belongs to the field of artificial intelligence and human resource management. According to the system, a DIKWP cognitive model is used for carrying out multilayer semantic analysis on candidate texts, and candidate language expressions are mapped from a concept space to a semantic space of post requirements. By constructing a cross-language post semantic knowledge base and a semantic matching engine, intelligent matching between the implicit ability of candidates and post competency requirements is realized. The system supports a multilingual environment, can identify equivalent skill intentions under different languages and culture expressions, and improves the precision and fairness of globalized recruitment. The system has a white-box interpretable characteristic, provides a transparent description of a matching basis, and is beneficial to trust and optimization of recruitment decisions.

The invention relates to the technical field of artificial intelligence and cultural heritage digital protection, and discloses a grotto imaging image interpretation-oriented rule-driven semantic analysis algorithm, which comprises the following steps: S1, an input initialization step: receiving a real attribute set A of a grotto imaging image to be analyzed and a description text O generated by a multi-modal large model, carrying out standardized preprocessing on the real attribute set A and the description text O; s2, a semantic unit segmentation step: segmenting the description text O into a plurality of independent semantic units based on grammar rules and domain features, wherein each semantic unit represents a specific description about the grotto image; according to the rule-driven semantic analysis algorithm oriented to grotto image interpretation, through core steps of semantic unit segmentation, concept space mapping, rule matching verification and the like, an accurate alignment mechanism between a large model generation text and an image real attribute is constructed, and the pain point of a prominent illusion phenomenon in the prior art is effectively overcome.

This invention discloses a semantic concept-driven visual token pruning method, relating to the field of artificial intelligence technology. The method includes: obtaining the hidden state of the visual token output by the decoder of the i-th layer of a visual language model; using a sparse autoencoder in a trained visual token pruning model to encode and project the hidden state of the visual token onto an interpretable semantic concept space of dimension i, obtaining a sparse feature vector; and pruning the i-th layer visual token using a concept-based visual token selection mechanism based on the sparse feature vector corresponding to the visual token, obtaining the pruned result. The pruning process of this invention can fully preserve semantic information.

The invention discloses a neural network interpretation optimization method based on text concept representation, which comprises the following steps of: aiming at a specific data set, firstly, constructing a text concept library covering data set category significant appearance characteristics on the basis of data set information, human priori knowledge and the like, and taking the text concept library as an intermediate domain between input and output of a neural network; the decision basis of the neural network is obtained through mapping in the text concept space, and interpretability of the decision process of the neural network is achieved; meanwhile, human knowledge is introduced through understanding and direct modification of a concept prediction result by human, direct intervention is carried out on a single sample decision result of the neural network, then a verification set is taken as a data basis, human modification logic is transferred to an original neural network by means of a knowledge distillation technology, and correction of self-cognition deviation is realized. According to the method, rich and easy-to-understand model reasoning logic is provided for human beings from the aspect of text concepts; and meanwhile, feedback is provided for the network in combination with rational knowledge of human beings, so that the performance of the network is optimized and improved.