14 results about "Infosphere" patented technology

The invention relates to an information processing method and device, equipment, a storage medium and a product, and the information processing method comprises the steps: receiving input multi-source knowledge information which comprises knowledge information from at least two information sources; under the condition that the multi-source knowledge information contains a conflict information group, based on an information field to which the knowledge information in the conflict information group belongs, performing conflict resolution processing on the knowledge information in the conflict information group to obtain target knowledge information; different information fields correspond to different conflict resolution modes; the conflict information group comprises a plurality of pieces of knowledge information with different information sources and information conflicts; and outputting the target knowledge information for the large model to call and perform question and answer reply. Through the method, the limitation of a traditional single rule in processing multi-source heterogeneous knowledge is effectively overcome, and the processing accuracy of knowledge conflicts in different fields is remarkably improved.

The invention belongs to the technical field of information, and particularly relates to a method for learning specific industry knowledge in a large model pre-training stage. Comprising the following steps of 1, cleaning and processing industry data; 2, pre-training industry knowledge; and step 3, knowledge increment pre-training. The method has the beneficial effects that the knowledge expression of the constructed industry is about 4.6 T, and the effect score in a downstream system generation task is improved by about 10% when the method is compared with a standard NST learning method. The specific content is as follows: for pre-trained industry knowledge, about 4.6 T data is constructed through ocr, data cleaning, manual annotation and the like;

The application discloses a method, device and readable storage medium of a control node, and relates to the electronic information field. The method of the control node provided by the application is applied to a management node, information is sent to a controlled node according to a PReq extension message after the management node exits a cycle limit, and compared with a previous mode of sending and receiving feedback through two asynchronous stages, the technology adopted in the scheme does not need to disconnect the controlled node from the management node, and only by modifying the cycle and the transmission mode of data, online updating of the controlled node is realized, data is transmitted through a message mode, and then in the asynchronous stage, the management node does not need to perform configuration updating any more, and only needs to wait for the corresponding controlled node to perform updating, so that the working efficiency of the server is greatly improved, and the stability of the server operation is ensured.

This invention discloses an automatic layout method for circuit schematics, belonging to the field of electronic information, which improves the efficiency and accuracy of circuit schematic inversion engineering. The invention includes the following steps: First, it reads in netlist information and component library information and performs data conversion; then, it clusters all components based on their pin and connection information; next, it performs local layout for each sub-partition, mainly including directed graph transformation, logic swap column positioning, logic row positioning, and physical expansion positioning; finally, it outputs the placement coordinates and rotation angle of each component. This solves the problems of slow speed and high error rate in current manual circuit schematic inversion methods.

The application discloses an AI-based design construction three-dimensional consistency review and collaborative management system and relates to the field of building information.The AI-based design construction three-dimensional consistency review and collaborative management system realizes high-quality acquisition and reliable evaluation of on-site point cloud data through a data acquisition module and a geometric consistency preliminary evaluation module in application. By arranging control sampling points and combining the optimization strategy of point cloud quality data, the effectiveness and integrity of the point cloud data can be significantly improved. On this basis, the system uses 95th percentile directional Hausdorff deviation H95 to perform robust geometric analysis on the registered point cloud set and the three-dimensional drawing BIM model, avoids the distortion of the judgment caused by individual abnormal points, and realizes accurate identification of the overall geometric deviation. This way enables the preliminary evaluation stage to quickly find construction deviations, reduces the links of manual review and repeated collection, and thus improves the review efficiency and data reliability.

The invention discloses a knowledge classification method and device. The method comprises the following steps: acquiring a geographic information ontology conceptual model; processing the geographic information ontology conceptual model to obtain an ontology knowledge set; the ontology knowledge comprises N pieces of ontology knowledge information, and each piece of ontology knowledge information comprises an ontology knowledge title and ontology knowledge content; and processing the ontology knowledge set to obtain a knowledge classification result. According to the method, the ontology knowledge set is extracted and constructed from the geographic information ontology concept model, and the obtained ontology knowledge title and content naturally bear the core concept, attribute association and hierarchical logic of the geographic information field, so that the subsequent classification result is ensured to completely accord with the internal system of the geographic information knowledge; the problems of disordered classification logic and poor field adaptability caused by knowledge source mixing of a traditional classification method are solved.

The application discloses a Q network control method and device and a readable storage medium, and relates to the electronic information field. In the Q network control method provided by the application, a preset behavior is a behavior randomly selected in a behavior space A or an optimal behavior selected in the behavior space A by using a database operator. A plurality of experience samples are extracted from an experience pool, and corresponding expected benefits are calculated based on the database operator. Finally, the parameters and vector indexes of the Q network and the parameters and vector indexes of a target Q network are updated according to the expected benefits. The database operator is an operator realized based on an approximate nearest neighbor search technology. Compared with the previous training control method, the approximate nearest neighbor search technology is introduced in the scheme to obtain the database operator, so that the search range is greatly reduced. Compared with the previous DQN which traverses all the action spaces, the training and decision efficiency is increased, and the scheme can be applied to large-scale discrete behavior scenes.

The application discloses a method for identifying key nodes of a heterogeneous Internet of Things (IoT) topology based on mutation theory, and belongs to the field of computers and information. First, an initial topology of the heterogeneous IoT is constructed, a topology set is generated by random disturbance, and performance indexes such as robustness, redundancy, communication efficiency and global clustering coefficient are calculated. Then, the indexes are normalized by using a mutation series method, the correlation between the indexes is evaluated based on a Pearson correlation coefficient, and a network performance mutation value is obtained by layer-by-layer fusion. Next, a deliberate attack process is simulated, a corresponding relationship between a node deletion sequence and performance evolution is established, and a performance evolution function is fitted. Finally, a cusp mutation model is established, critical nodes causing network performance mutation are identified by analyzing the equilibrium state, singular set and bifurcation set of the model, and the application breaks through the limitations of traditional static network indexes, can accurately capture the mutation behavior of network performance, is objective, and provides a theoretical basis for topology optimization and fault-tolerant design of the IoT.

Embodiments of the present application relate to the field of communication and information, and disclose a disaster recovery method, device, local site and storage medium. In the present application, the disaster recovery method comprises: if it is determined that the identity of the local site is a backup local site, detecting whether there is a target local site satisfying a preset condition; the preset condition comprises: the identity of the target local site is a main local site, and the target local site is assigned to the local site and has not established a disaster recovery relationship with the local site; if there is a target local site satisfying the preset condition, establishing a disaster recovery relationship with the target local site; and backing up data to be backed up in the target local site for the target local site. The flexibility of business expansion can be improved.

The invention discloses a method and system for converting a natural language into a knowledge graph query statement through AI, and the method comprises the steps: carrying out the standardization processing of a natural language query statement inputted by a user, and obtaining an optimized query statement; performing intention analysis on the optimized query statement through a routing agent to determine one or more information fields related to the optimized query statement, and determining at least one target field expert agent needing to be called from a plurality of field expert agents; wherein each domain expert agent is configured to process a query of a specific information domain; calling the at least one target domain expert agent to enable each called target domain expert agent to generate a corresponding query fragment based on the optimized query statement; performing constrained combination processing on all the generated query fragments to obtain executable query statements; and executing the executable query statement in a knowledge graph database, and obtaining a query result.

This invention provides a method and system for generating a unified knowledge graph from multimodal data sources using a collaborative multi-agent LLM. [Solution] A system presents a collaborative multi-agent knowledge graph search extension generation framework using multiple collaborative multi-agent LLMs to extract information from multiple multimodal documents and build a unified knowledge graph to facilitate query-based searching, wherein a query generation agent generates multiple queries to reveal all possible information present in multiple multimodal documents, a domain model generation agent generates domain models based on the multiple queries, a domain model populator agent populates domain models using data extracted from multiple multimodal documents, and a unified knowledge graph is constructed from the populated domain models, systematizing the extracted information.

The application provides a data processing method and device, computer equipment and a storage medium, applied to the information field, and the method comprises the steps of: acquiring an application calling topological graph corresponding to a target object, wherein the application calling topological graph comprises a calling relationship between a plurality of data nodes called when the target object is executed; encoding the data nodes in each calling level of the application calling topological graph respectively, and generating a string corresponding to each calling level; determining a target substring in the string corresponding to each calling level and branch data corresponding to the target substring in the application calling topological graph, wherein the target substring is used to represent a data node that is called in a loop and meets a preset condition in the calling level; and deleting the branch data in the application calling topological graph. The data processing method provided by the application determines the branch data called in a loop by encoding the data nodes, thereby trimming the branch data, so as to reduce the storage cost of the application calling topological graph.

This invention relates to the emerging information field and provides a method and system for smart screen application orchestration based on a large language model. The method includes: inputting user-input natural language commands and a pre-built application capability description library into a large language model for intent understanding and task decomposition to obtain intent parsing results; inputting the intent parsing results and contextual information into the large language model for service matching and orchestration planning to obtain a large-screen orchestration scheme; using an interface generation engine, dynamically combining multiple cards and injecting parameters according to the large-screen orchestration scheme to obtain a personalized large-screen interface; collecting and analyzing user interaction behavior data based on the personalized large-screen interface to obtain behavior feedback results; and inputting the behavior feedback results into a user preference model for behavior pattern recognition to obtain recommended commands that trigger proactive large-screen generation. This invention solves the application isolation problem in smart screen systems and improves the intelligence level of smart screen systems.

The application relates to the field of mobile information, and provides an information recommendation method and device, electronic equipment and computer program product. The method comprises the following steps: determining a social relationship network of a current user, and determining initial grade values of relationship edges between initial nodes in the social relationship network; performing transition node setting on the social relationship network according to the initial grade values, so as to obtain a behavior relationship network; calculating social influence of each initial user in the behavior relationship network; determining a target user from the initial users based on the social influence, and performing information recommendation according to the target user. The application can improve the timeliness of recommended content, so that the recommended content can be accurately matched with the future behavior of the recommended user, and the accuracy of information recommendation is improved.