36results about How to "Automate the build" patented technology

The invention provides a knowledge graph construction method and system. The method comprises the following steps: acquiring corpus information of at least one domain of a to-be-constructed knowledgegraph; carrying out word segmentation and part-of-speech tagging on knowledge text content fragments included in the corpus information to obtain the key words in the knowledge text content fragments;matching the key words with the domain ontology according to the preset rules to obtain the knowledge element instances in the knowledge text content fragments, the attributes of the knowledge element instances and the association relation between the knowledge element instances, wherein the domain ontology includes a domain topic, at least one model included in the domain topic, attributes of the model, and associations between the models, and the model includes at least one knowledge element instance; and constructing the knowledge graph according to the knowledge element instances in the plurality of knowledge text content fragments, the attributes of the knowledge element instances and the association relation between the knowledge element instances. The automatic construction of theknowledge graph is realized, the construction cost of the knowledge graph is reduced, and the construction efficiency and accuracy of the knowledge graph are improved.

The invention provides a discretized traffic state distinguishing method based on deep learning, and belongs to the technical field of traffic state distinguishing. The method comprises the steps of a, utilizing monitoring equipment arranged in all directions of a crossroad to collect traffic states of the crossroad in different moments to correspondingly obtain multiple sets of traffic state information; b, conducting discretized coding description respectively on the multiple sets of traffic state information according to a discretized coding process to correspondingly obtain discretized coding data of the crossroad in the different moments; c, building a traffic state distinguishing model based on deep learning according to the discretized coding data obtained in the step b; d, distinguishing the traffic states by utilizing the traffic state distinguishing model built in the step c. According to the discretized traffic state distinguishing method based on deep learning, the traffic states can be comprehensively and truly described, experts do not need to participate in traffic state feature selection, and building of the traffic state distinguishing model can be automatically achieved in a semi-supervised mode.

The invention relates to a method and a device for constructing a road intersection road network, a storage medium and electronic equipment. The method comprises the steps of acquiring lane line information of a road connected with an intersection and pointing information of a guiding arrow, close to the intersection, of the road; for each road, determining the connection relation between the roadand the intersection according to the pointing information of the guiding arrows of the road, and the connection relation is used for representing whether the road is a driving-in road or a driving-out road of the intersection; and determining a pairing relationship between a lane in the entry road of the intersection and a lane in the exit road of the intersection according to the lane line information and the pointing information of the guide arrows in each entry road and the lane line information and the pointing information of the guide arrows in each exit road. By means of the method andthe device, automatic construction of the road intersection road network can be achieved, and efficiency and accuracy are high.

The embodiment of the invention provides a lane line group construction method and device. The lane line group construction method comprises the steps: acquiring original lane lines; grouping the original lane lines to obtain a plurality of first lane line groups; for each first lane line group, obtaining a reference line corresponding to the first lane line group, respectively obtaining the distance between each first lane line in the first lane line group and the reference line, and clustering each distance by using a DBSCAN clustering algorithm to obtain a classification corresponding to each first lane line; fitting all the first lane lines belonging to the same class in the plurality of first lane line groups to obtain a plurality of target lane lines; and combining the plurality of target lane lines to complete the construction of the lane line group. According to the scheme, automatic construction of the lane line set is achieved in a big data mode, and the workload is small, and the cost is low, and the accuracy is high.

The invention discloses a Dynamo-based point cloud BIM automatic modeling method. The Dynamo-based point cloud BIM automatic modeling method comprises the following steps of S1, solving an extreme value of a point cloud three-dimensional degree, calculating a surface area of a point cloud bounding box, and calculating a point cloud average surface density according to the number of point clouds; S2, slicing the point cloud in a selected direction, setting the number of slices as an input type parameter, and calculating a point set in a previous threshold range and a next threshold range of a tangent plane; and S3, projecting the upper and lower point sets to a tangent plane for intersection, and fitting discrete points to obtain a cut boundary contour. In the Dynamo-based point cloud BIM automatic modeling method, a built-in visual programming platform (Dynamo) of the revit is fully utilized as a baton to perform layered editing and segmented modeling on the point cloud imported into the revit, so that the point cloud can automatically construct an entity model without transition and is endowed with attributes.

The invention discloses a method and a terminal for automatically constructing a front-end project. The method comprises the following steps of: installing a packet on which a front-end project to beconstructed depends; obtaining an execution environment of a front-end project to be constructed; and a packet based on the dependency, packaging a front-end project to be constructed and issuing thepackaged front-end project to a corresponding environment according to the execution environment, according to the method, automatic construction of the front-end project is realized, the front-end project only needs to be packaged once by acquiring the execution environment of the front-end project, different environments can be automatically distinguished according to the execution environment,and then the packaged front-end project is directly published to the corresponding environment, so that the development process is simplified, and the development efficiency of the front-end project is improved.

The embodiment of the invention discloses an interface document management method and device, a server and a storage medium. The method comprises the following steps of subscribing registered system services from a distributed service center; receiving the address information of the system service fed back by the distributed service center; and in response to an interface document construction instruction, based on the address information of the system service, calling a first interface function associated with the system service to obtain corresponding interface information, and constructingan interface document according to the obtained interface information. According to the technical scheme provided by the embodiment of the invention, the consistency of the actual interface and the interface documents can be ensured, and the isolation of the interface documents between the system services is realized.

The invention provides a building method and device for a specialized database. The method comprises the following steps: creating a target professional database; adding a neighborhood body corresponding to the target professional database into the target professional database in order to serve as a knowledge organization system of a target database; acquiring multiple knowledge entries; buildingassociation relations between all the knowledge entries and knowledge elements in the neighborhood body; and building data indices for the knowledge entries and the target professional database. Because the neighborhood body can provide more accurate and complete data association relations, the neighborhood body is taken as a knowledge organization system of the professional database and association relations between the knowledge entries and the knowledge elements in the neighborhood body are established. Therefore, association relations of all resources in the overall professional database are more complete. Resources content are more comprehensive. Recommendation and retrieval results are more improved. The professional database is built in an automated manner so that creating time is reduced.

The invention relates to the technical field of artificial intelligence, and discloses a community portrait construction method, device and equipment and a storage medium, which are used for improvingthe efficiency and effect of community marketing promotion analysis. The method comprises the following steps: grouping users in a preset community relation network according to different community labels to respectively obtain a plurality of corresponding communities; calculating the user scale of each community, the relationship number of each user in each community and the number of connectedtriangles; summing the relationship numbers corresponding to the communities to obtain the total number of the relationship numbers corresponding to the communities; summing the number of the connected triangles corresponding to each community to obtain the total number of the connected triangles corresponding to each community; calculating the community closeness corresponding to each community according to the user scale, the total number of the relationship coefficients and the total number of the connected triangles; and sorting the community portraits corresponding to each community according to the descending order of the community closeness.

A graphical user interface system that visually represents a graph node having multiple input ports, and that automatically adjusts the number of input ports to the graph node as open input ports are connected to upstream graph node(s) and/or as used input ports are disconnected from upstream graph node(s). Upon detecting that a set of one or more input ports has just been connected to one or more upstream graph nodes, the system may automatically add one or more additional input ports to the visual representation of the graph node without explicit user instruction. Upon detecting that a set of one or more of the input ports has just been disconnected from one or more upstream graph nodes, the system may automatically remove the determined one or more input ports from the visual representation of the graph node.

The invention provides a bank parameter report generation method and apparatus, and relates to the technical field of bank system data processing. The method comprises the steps of obtaining parameterreport configuration information input by a user, and generating a configuration information table; performing redefinition operation on an existing parameter report file in a bank system host to form a parameter report file with empty contents; according to the configuration information table, generating a parameter report in the parameter report file with the empty contents; and downloading theparameter report to a preset parameter report server. The bank parameter report can be automatically generated; the configuration information of the parameter report can be conveniently maintained through the configuration information table; and the problems of huge manpower and time costs and relatively complex manual process of a parameter report generation mode in the prior art can be solved.

The invention discloses an indoor space semantic value calculation method based on spatial and social multimedia data. The semantic value of indoor independent subspace is divided into location and social contribution, and the semantic value of indoor independent subspace is fused. The method comprises analyzing the topological connection relation of each indoor independent subspace, constructingthe indoor space model, calculating the distance between each object, calculating the path matrix according to the user walking constraint, and calculating the location contribution degree; Accordingto the social multimedia data of indoor independent subspace, the sentence structure being analyzed and the key words being extracted to construct the emotion classification base classifier. Sentencesets are divided into key sentences and non-key sentences to classify social evaluation texts. The social contribution degree is calculated according to the classification of social evaluation texts.The indoor space value calculation method of the invention can be used for solving key problems such as value prediction of newly introduced shops, multi-dimensional shop evaluation and the like.

The invention provides a version construction automation realization method and device, and the method comprises the steps of analyzing a publishing environment of each project version, extracting thesame part of each publishing environment, defining the same part as a public environment set, defining different parts of each publishing environment as a publishing environment feature environment set, and uploading the publishing environment feature environment sets to a version controller svn; based on the public environment set and the feature environment set of the publishing environment, creating a publishing environment construction component and a dependency component of each project version, and uploading each publishing environment construction component and each dependency component to a version controller svn; creating a version issuing interface and a version issuing environment interface in the continuous integration tool, setting the version issuing interface to obtain project version information, and setting the version issuing environment interface to connect a version issuing environment construction component according to the project version information; and settinga continuous integration tool to obtain project version information through a version issuing interface, and connecting a version issuing environment construction component through a version issuingenvironment interface to construct a version issuing environment, thereby finishing version issuing.

The invention relates to a rectangular segmentation graph construction method based on edge orientation, and belongs to the technical field of data visualization. According to the method, the topological graph is directly extracted from the original map, and after the topological graph is preprocessed, rings with the degree smaller than 4 do not exist in the topological graph, so that the divisible topological graph is obtained; on the basis, orientation is carried out on each edge in the divisible topological graph by utilizing relative position constraint, and then an orientation graph with each ring being a rectangle is obtained; and finally, determining the coordinate of each point by using the connection state of each point in the orientation map so as to obtain a rectangular segmentation map. According to the method, an original map does not need to be converted into a PTP map, and the segmentable topological graph can be obtained by directly utilizing a preprocessing mode of setting a rule; and meanwhile, positioning and coordinate determination can be realized through simple iteration only according to the property of the rectangular graph. Therefore, the method is simple in implementation process, does not need extra complex calculation, and can automatically implement the construction of the rectangular segmented graph.

The invention discloses a method for constructing a neural network force field model based on a global optimization algorithm, and relates to the technical field of deep learning, and the method comprises the following steps: S1, collecting effective material simulation data according to a research target, carrying out the single-point calibration, classification and screening, and storing the data in a database; s2, according to a research target, selecting energy/energy and force from a database as an evaluation function of a neural network, and training to obtain a neural network model; s3, based on the neural network model obtained through training, performing data sampling through a genetic algorithm, and search of a global space is achieved; s4, performing validity evaluation on the sampling data, automatically submitting the valid sampling data to a supercomputing server for first principle calculation calibration, and merging the valid sampling data into a database to complete a round of iteration process; and S5, repeating the steps S1 to S4 to obtain the neural network force field model adaptive to the research task. According to the invention, automatic construction of the neural network force field model can be realized.