42results about "Semi-structured data querying" patented technology

The present disclosure provides a character recognition model training method and apparatus, a character recognition method and apparatus, a device and a medium, relating to the technical field of artificial intelligence, and specifically to the technical fields of deep learning, image processing and computer vision, which can be applied to scenarios such as character detection and recognition technology. The specific implementing solution is: partitioning an untagged training sample into at least two sub-sample images; dividing the at least two sub-sample images into a first training set and a second training set; where the first training set includes a first sub-sample image with a visible attribute, and the second training set includes a second sub-sample image with an invisible attribute; performing self-supervised training on a to-be-trained encoder by taking the second training set as a tag of the first training set, to obtain a target encoder.

Taxonomy driven multi-system networking and content delivery systems and methods are provided herein. An example systems includes a website authoring platform generating a taxonomy structure for a webpage, the taxonomy structure indicating assets to be included in the webpage a layout for the assets, a technical documentation authoring system for creating technical documents related to the assets, and a metadata binder providing the technical documentation authoring system with access to the taxonomy structure of the website authoring platform so that technical document authors are provided with the taxonomy structure to guide creation and structuring of the technical documents so that the technical documents comprise data that corresponds to the assets.

The invention discloses a cognitive resident node recognition and attention link construction method, device and system, and belongs to the technical field of user behavior modeling, and the method comprises the steps: user behavior data collection, structured processing and preprocessing; behavior chain construction and semantic alignment: unifying behavior semantic description and constructing a time sequence mapping chain by identifying a behavior stage, and providing path input for cognitive resident node identification and attention link construction; recognizing cognitive resident nodes and constructing attention links, recognizing cognitive resident points of a user in corresponding stages, and establishing flow paths of attention between different nodes and behavior stages to form a structured attention link map. The user behavior modeling method solves the problems of data dimension splitting, behavior chain missing, rough intention modeling and insufficient time sequence evolution characteristic mining in an existing user behavior modeling method.

The invention discloses a marketing operation and maintenance abnormity positioning system based on business and technical index fusion. A configuration information acquisition module is used for establishing a mapping association basis of business indexes and technical indexes; the index data acquisition module is used for acquiring technical architecture operation indexes and business process key node indexes in parallel based on the mapping association basis established by the configuration information acquisition module to form a fusion type index data set and storing the fusion type index data set to a monitoring database; the monitoring display module is used for calling the fusion type index data set in the monitoring database, and displaying the dynamic association relationship between the business indexes and the corresponding technical indexes in a linkage manner through the same view; the alarm module performs linkage verification based on the association relationship displayed by the monitoring display module, triggers an alarm and pushes the alarm to related personnel; the anomaly positioning module is used for performing traceability analysis on the associated indexes in the fusion type index data set and positioning the source corresponding to the anomaly; according to the invention, the problems of weak association, difficult analysis, slow positioning and the like caused by service and index separation in the prior art can be solved.

Among other things, we describe a method of receiving a portion of metadata from a data source, the portion of metadata describing nodes and edges; generating instances of a data structure representing the portion of metadata, at least one instance of the data structure including an identification value that identifies a corresponding node, one or more property values representing respective properties of the corresponding node, and one or more pointers to respective identification values, each pointer representing an edge associated with a node identified by the corresponding respective identification value; storing the instances of the data structure in random access memory; receiving a query that includes an identification of at least one particular element of data; and using at least one instance of the data structure to cause a display of a computer system to display a representation of lineage of the particular element of data.

Examples may use pathogenicity evidence data associated with genetic variants and health conditions to create input data for a causal machine learning model. Examples may apply a causal machine learning model to input data to produce a trained causal model. A graphical representation of the trained causal model may include nodes connected via acyclic directed edges. A first node of the nodes may represent a pathogenic evidence variable associated with the health condition. The at least one second node may represent a cause of the pathogenic evidence variable. The at least one third node may represent the impact of the pathogenic evidence variable. The non-cyclic directed edge may represent a relationship between two nodes. Examples may output prediction data sampled from the trained causal model.

The invention discloses a quality risk management method and system based on a white spirit enterprise, and relates to the technical field of risk prediction. Comprising the steps of 1, creating a quality risk management system based on a white spirit enterprise, 2, collecting, verifying and storing risk data through a risk data collection module, 3, designing, managing and calling a risk identification template through a risk identification template management module, and 4, managing and calling the risk identification template. The method comprises the steps of 1, managing risk identification, processing and tracking through a management risk identification and processing module, 2, carrying out risk analysis, assessment, grade evaluation and potential risk mining through a risk analysis and assessment module, and 3, carrying out risk information auditing, publishing, tracking and reminding through a risk publishing and tracking module. And 7, through summarization and analysis of risk data, report generation and visual display of a quality analysis module, a complete risk management and control process is formed, and it is guaranteed that risk management and control are efficient and orderly.

The invention discloses a method for storing and recovering train number, train unit number and dispatching number in a CTC station graph during construction, a readable storage medium and electronic equipment, and the method comprises the steps: S1, before construction, STPC stores the train number in an in-station train number window on a station graph interface, and the train unit number and dispatching number information on a turnout, a station track and a turnout-free section; the attribute information and the information of the equipment are synchronously stored; s2, after construction, the STPC reads the stored train number information, the stored train group number information, the stored dispatching number information, the stored attribute information and the stored equipment information, arranges and displays the stored train number information, the stored train group number information, the stored dispatching number information, the corresponding attribute information and the stored equipment information, and selects the train number information, the train group number information, the dispatching number information and the corresponding attribute information which need to be recovered based on a preset mode; and S3, based on the selection result of the S2, sending the corresponding train number, train group number and debugging number information and the attributes and equipment information thereof to the LIRC, and recovering the display of the corresponding information on the station yard graph interface by the LIRC. The method has the advantage that the operation and maintenance efficiency is effectively improved.

A structured document storage format is provided in which tokens used for encoding and decoding are contained within the encoded document in an inline dictionary. Each document is encoded independently, and there is no central dependency or any dependency on other documents during DML and query execution. Each encoded document has all information with it on disk so that the document can be independently shared, decoded, or distributed. A mechanism is provided for quickly determining a mapping between tags and tokens without having to scan the entire inline token dictionary for each document. This allows the database system to process multiple documents, each having its own inline token dictionary, without having to fully scan the inline dictionary of every document for each tag that referenced in an operation.

The invention discloses a deserialization utilization chain mining method for Dotnet, and belongs to the technical field of Web security. In order to solve the problems that an existing. NET environment is lack of a high-adaptability white-box analysis tool and automatic judgment of a deserialization utilization chain is difficult, C # source codes are generated through a decompilation program set, a code attribute graph oriented to a hosting environment is constructed, Source and Sink rules are customized, a potential vulnerability utilization chain is queried in a graph database, and the potential vulnerability utilization chain is queried in the graph database. And the validity is judged by using a large language model. According to the method, the deserialization vulnerability utilization chain in the hosting program set can be efficiently found, and the automation and intelligence level of vulnerability mining is improved.

To properly respond to a user speech in a case where the knowledge-based scale is enlarged.SOLUTION: An information processing apparatus according to the present disclosure includes: a knowledge object generation unit for extracting a plurality of terms from a text document and generating a knowledge graph indicating relationships between the extracted plurality of terms; an acquisition unit for acquiring context information in which general documents are included as knowledge; and an inference unit for generating a response sentence on the basis of a question sentence from a user and the context information.SELECTED DRAWING: Figure 1

Case management systems and techniques are disclosed. In various embodiments, searching case instances is facilitated. An indication to create a composite index across hierarchical case nodes comprising a case model is received. Case instance data associated with the case model is used to generate the composite index. The composite index is made available to be used to optimize searches of a plurality of case instances with which the case instance data is associated.

The invention relates to the technical field of land site selection, and provides a rural land site selection system and a control method, the rural land site selection system comprises an infrastructure layer, a data resource layer, a platform layer, an application service layer and an application system layer; the infrastructure layer is used for providing physical resources and basic software; the data resource layer is used for storing and managing related data of the rural land based on the infrastructure layer; the platform layer is used for calling related data and providing data and function support for the application service layer; the application service layer is used for providing various rural land site selection services for users based on the platform layer; and the application system layer is used for interacting with a user by using an application interface of a natural language based on the application service layer so as to realize the rural land site selection demand of the user. The system provided by the invention can improve the correlation between rural land site selection and related policies and the experience feeling of different user groups.

Methods, systems and software include programmatically downloading over the Internet, at routine intervals of time as well as on demand, from an online computer system of the USPTO, data relating to patent properties that are associated with a customer number; and creating a docket based on the downloaded data, generating and displaying a report generated from the downloaded data, sending an alert (email or text message) that is based on the downloaded data, or any combination thereof. Outgoing correspondence of the USPTO is provided in PDF format for review by a user. Data specific to a particular client also is provided to the client programmatically, with reports, alerts, docketing, or any combination thereof, further being provided for use by the client.

The invention relates to the technical field of computer-aided teaching and intelligent data processing, and discloses an English course management system applied to English teaching, which comprises a topology construction module, a prefetching buffer module, a feature analysis module, a real-time pruning module and a path presentation module. According to the system, firstly, a semantic topology network is constructed based on course metadata, and subsequent nodes and attribute data of the subsequent nodes are pre-fetched at a background according to a topological relation to construct a candidate buffer pool; meanwhile, a feature analysis module analyzes the micro-interaction behavior of the user in real time to generate a feature vector; and the real-time pruning module calculates a cognitive damping coefficient according to the cognitive damping coefficient, and dynamically filters the buffer pool by utilizing a node retention formula in combination with difficulty increment and context switching cost value. And when a submission instruction is triggered, the system locks the optimal node in time to display the adaptive content, and corrects the topological weight by using the damping coefficient. According to the method, data retrieval and cognitive state calculation are parallelized, so that course real-time adaptation and low-delay interaction based on implicit feedback are realized.

The application discloses a long-text-oriented episodic screenplay generation method and system, and belongs to the technical field of artificial intelligence content generation and digital video production. The method forms multiple episode objects through voice-over role initialization, text semantic compression and episode segmentation, and extracts role candidate objects and location time sequence candidate objects; based on normalization rules, an identity register of roles and a location time sequence register are constructed respectively, and unified identification reuse or new registration is performed on the candidate objects; under the constraint of the register, scene planning results and structured screenplay constraint patterns are generated, and multiple generation agents are called to output screenplay objects; and then, through structure verification, field completion and unified field transformation, the standardized results of episodes, roles, locations and screenplays are output. The application can improve the consistency of cross-episode roles and locations, the stability of the screenplay structure and the reuse efficiency of downstream production.

Database query processing techniques are disclosed. In various embodiments, a query associated with a database is received. A byte code representation of the query is generated, including by decomposing the query into a discrete set of streaming operators defined over associated data frames, wherein the byte code includes code defining for each operator in the discrete set of streaming operators the processing to be performed by that operator and further embodies a data flow graph that defines a flow of data to and through the discrete set of streaming operators. The byte code is executed by a query processing engine to generate and return a query result.

The invention provides an automatic reconstruction method for an aviation three-dimensional CAD (computer aided design) model based on XML (extensible markup language), which belongs to the technical field of computer aided design and comprises the following steps: extracting geometrical characteristics, modeling history and parameter constraint information of a model by using a CATIA CAA interface, and storing the geometrical characteristics, the modeling history and the parameter constraint information in an XML format; an XML file is analyzed through a Python script, a JS modeling script suitable for a domestic CAD platform is automatically generated, and automatic and accurate reconstruction of a model in domestic CAD is achieved; according to the method, the parameterization characteristics and the modeling sequence of the original model are reserved, the data migration efficiency is remarkably improved, the problem that the models among the heterogeneous CAD systems are difficult to reuse is solved, and the method has good universality and engineering application value.

A system including processor(s) configured to: run digital twin(s) of real-world asset(s); receive datasets from data sources, data sources being communicably coupled to the processor(s) via a first application programming interface (API); update the digital twin(s), based on the received datasets; receive, from a user device, a query for obtaining information from the digital twin(s), the user device being communicably coupled to the processor(s) via a second API; process the query to generate a response comprising said information; and send the response to the user device.

This invention discloses a method and apparatus for generating function blocks for PLC control programs, comprising at least the following steps: Step S1: Accessing a corresponding timing database and obtaining time series information based on user input; Step S2: Creating corresponding function blocks based on the specific time series information selected by the user to generate standard XML strings; Step S3: Integrating all generated XML strings and outputting an XML file for use in the PLC control program. This invention automatically generates function blocks usable in PLC control programs based on user settings, and extracts path node information of the time series from the timing database, automatically generating function block XML files usable for controlling the PLC program, thereby greatly simplifying the operation of timing data.

Systems and methods for data management via a data management system are provided. The data management system can receive data relating to one or more assets, the data in one or more formats, convert the asset data to a uniform format, extract data elements, compare and normalize the data elements against a standard reference database, and automatically validate the one or more data elements. The validation can be performed by comparing data elements to requirements set by the data management system or other party, and confirming that the data elements match the requirements. The validation can operate to initiate an asset transaction.

Systems and methods for data management via a data management system are provided. The data management system can receive data relating to one or more assets, the data in one or more formats, convert the asset data to a uniform format, extract data elements, compare and normalize the data elements against a standard reference database, and automatically validate the one or more data elements. The validation can be performed by comparing data elements to requirements set by the data management system or other party, and confirming that the data elements match the requirements. The validation can operate to initiate an asset transaction.

The present disclosure provides a training method, apparatus, device, storage medium, and program product for a data processing model. The method includes: obtaining a first web page, wherein a source code of the first web page includes a first code tag and first text content to be processed; combining at least one of a second code tag associated with a first text segment and first position information with the text of the first text segment to obtain a first input vector; using the first input vector and an element tag of the first text segment as training samples to train the data processing model. Embodiments of the present disclosure can reduce training costs and are no longer limited to the method of formulating a data extraction rule for one type of web page in the prior art, improving compatibility and processing efficiency. The present disclosure also provides a data processing method, apparatus, device, storage medium, and program product.

Among other things, we describe a method of receiving a portion of metadata from a data source, the portion of metadata describing nodes and edges; generating instances of a data structure representing the portion of metadata, at least one instance of the data structure including an identification value that identifies a corresponding node, one or more property values representing respective properties of the corresponding node, and one or more pointers to respective identification values, each pointer representing an edge associated with a node identified by the corresponding respective identification value; storing the instances of the data structure in random access memory; receiving a query that includes an identification of at least one particular element of data; and using at least one instance of the data structure to cause a display of a computer system to display a representation of lineage of the particular element of data.

This information processing device uses a parent-child relationship between subjects corresponding to a plurality of tags used in a document to identify the subject of an end node. The information processing device calculates a vector of a tag corresponding to the subject of the end node on the basis of each word included in definition information set for the subject of the end node and a word vector dictionary defining a vector of each word.

Among other things, we describe a method of receiving a portion of metadata from a data source, the portion of metadata describing nodes and edges; generating instances of a data structure representing the portion of metadata, at least one instance of the data structure including an identification value that identifies a corresponding node, one or more property values representing respective properties of the corresponding node, and one or more pointers to respective identification values, each pointer representing an edge associated with a node identified by the corresponding respective identification value; storing the instances of the data structure in random access memory; receiving a query that includes an identification of at least one particular element of data; and using at least one instance of the data structure to cause a display of a computer system to display a representation of lineage of the particular element of data.

The invention provides a transportation junction energy management and control system data management method and system, and the method comprises the steps: defining plant station division and equipment types, creating a new bitmap array and a mapping table, and carrying out the system initialization; existing equipment is inquired by inputting equipment meta-information, a new equipment serial number is generated according to the existing equipment, a start bit offset and an extended bitmap length are determined based on the new equipment serial number, mapping table equipment information of different statistical time periods is input into a mapping table, and equipment counting is updated; receiving real-time parameter data of the equipment, judging a parameter state according to a parameter verification rule, converting the equipment-parameter into an absolute bit index in the bitmap through a mapping function, and executing an atomic bitmap updating operation; and the regional health degree and the type health degree are determined through multi-dimensional statistics and aggregation statistics. According to the method, real-time sensing of the state of the distributed energy equipment, rapid positioning of abnormal data, multi-dimensional energy consumption statistics and intelligent regulation and control are realized, and dynamic addition and deletion of the equipment and efficient management of variable-length parameter data are supported.