33 results about "Sequence graph" patented technology

A method, system and computer readable article of manufacture for displaying visualization from an event log. According to the method, sequences are extracted from the event log. The event log is a record of a plurality of events that occurred. A sequence graph is generated from the sequences. The sequence graph includes nodes representing the events of the event log and the events of each node have a common property. A sequence graph visualization is displayed from the sequence graph. The sequence graph visualization includes (i) node icons representing the nodes of the sequence graph, (ii) node transition icons representing relationships between nodes, and (iii) an event icon for each of the one or more events of a corresponding node.

A profile of a finished part is recognized based on an inputted CAD data. A delta volume for the finished part is decomposed based on information on cutting tools and the profile. Thereafter, a dependency graph representing precedence relation between the decomposed delta volumes is generated. And then, a process sequence graph representing process plans is generated based on the dependency graph. The delta volume decomposition is performed based on information on cutting tools and a machine configuration as well as part geometry, such that the decomposed delta volumes are suitable to be cut away from a raw stock by the cutting tools.

The invention discloses an intelligent question answering method and system based on a pet knowledge graph, and the method comprises the steps: constructing a named entity dictionary, abstracting questions, and facilitating the classification of the questions. A method of combining word2vec with Levenshtein Distance is provided to realize entity linking, and experiments show that the method is effective. Texts are trained by constructing a text classifier based on Naive Bayes, and the improved TF-IDF naive Bayes classification algorithm is provided, the distribution situation of feature wordsin a text set and the category distribution situation are considered, and the improved TF-IDF effectively improves the text classification effect. Through the result of the text classifier, the intention of the natural language question is determined, and the natural language question is matched with the corresponding word sequence graph. The word order graph is converted into a similar SQL querystatement of the OrientDB, and querying is performed in a graph database storing the knowledge graph. Finally, the constructed intelligent question and answer system based on the knowledge graph is displayed in an example, and experiments show that the system has a relatively high application value in question and answer application in the field of pets.

The invention discloses a left-atrial-appendage CT image segmentation method, and relates to medical image processing. Firstly, left-atrial-appendage CT images are denoised, prior knowledge is utilized to construct missing boundary information of a left atrial appendage, and a region growing algorithm is used to obtain a single initial segmentation image to use the same as a single Atlas label graph; then information relationships of the single label graph and an adjacent left-atrial-appendage CT sequence are utilized for improved Atlas segmentation to obtain an initial segmentation sequence;contour information in the initial segmentation sequence is used as initial contours of a level set algorithm, and approximate Euclidean distance information of left-atrial-appendage CT sequence graphs is obtained to be used as feature input; and finally, a precise segmentation result sequence is obtained through a fast level set algorithm to achieve the purpose of left-atrial-appendage CT image segmentation.

The embodiment of the invention relates to a flow detection method and device, equipment and a medium. The method comprises the following steps: acquiring an access time sequence graph of a to-be-detected end; wherein the access time sequence graph comprises at least one target graph node with an abnormal traffic identifier; performing graph representation learning on the access time sequence graph to obtain a time sequence vector corresponding to each graph node in the access time sequence graph; performing clustering processing on the time sequence vector according to a preset clustering algorithm to obtain a plurality of clustering categories; and obtaining a target clustering category to which the target map node belongs from the plurality of clustering categories, and determining the abnormal traffic of the to-be-detected end based on the target clustering category. In the embodiment of the invention, the detection rate of the abnormal traffic with the unknown rule is improved, the network security is ensured, the automation degree of the abnormal traffic labeling in a long time period is improved, and the manpower cost consumed by the abnormal traffic labeling is reduced.

The invention discloses a ship track prediction method based on image superposition. The method comprises the following steps: firstly, shooting marine navigation of a ship by using a video, extracting a sequence diagram of each frame in the video, selecting a time period, superposing and compressing each frame of image in the time period onto one image, then processing the image, extracting feature information and track information, and repeating for multiple times; using each compressed image as a new track data set, and using Kalman filtering for prediction. According to the method, the sequence graphs in each period of time are superposed, and the target contour characteristics are extracted, so that the target point of each frame is changed into the target track in each period of time, and when the Kalman filtering method is used for track prediction, the track information is more comprehensive, and the prediction effect is better.

The invention discloses a barnyard grass phosphoenolpyruvate carboxylase gene and coded protein and appliance, which is one ribotide sequence in the following sequences: 1) sequent graph SEQ ID No.:1;2) DNA sequence hybrid ribotide sequence limited by SEQ ID No.:1. the coded gene protein is one of SEQ ID No.:2 in the 1) sequence graph; 2) SEQ ID No.:2 in the 2) sequence graph with two amino residue sequence replaced, lost or added form one to ten to improve plant photosynthesis efficiency.

The embodiment of the invention provides a gait data generation method and device based on an adversarial network. The method comprises the steps of obtaining a plurality of first gait picture sets ofa target object, inputting the plurality of first gait picture sets into a generation network to obtain a plurality of second gait picture sets, inputting a third gait picture set and the second gaitpicture sets into a first adversarial network for comparison, and obtaining a second adversarial network if a comparison result is smaller than a preset threshold value, determining that the generation network is a practical generation network; and acquiring gait data of any plurality of gait picture sets of the target object by adopting the second adversarial network and the practical-level generation network. According to the gait data generation method based on the adversarial network provided by the embodiment of the invention, the gait sequence graph of the target object can be supplemented under the condition that only a small number of pictures exist in the collected gait graph; a good compensation effect is achieved on collection and labeling of a large number of gait samples, andthe economic cost is reduced.

The invention discloses an electronic map making method and system for partitioning and cutting based on discrete data, and the method comprises the steps: obtaining a display range based on an engineering target; projecting the obtained spatially discrete data covering the display range to a coordinate system to form a data point graph; performing grid interpolation on the data point graph to form a data grid file; generating a contour line sequence graph according to the data grid file and the contour line sequence value; partitioning the data raster file according to the contour line sequence value to form partitioned raster data; transferring out the partition raster data to form a partition surface graph; based on the display range, cutting the data point graph, the contour line sequence graph and the partition surface graph to generate a cut data point graph, a cut contour line graph and a cut partition surface graph in the display range; and sequentially superposing and displaying the cutting data point graph, the cutting contour line graph and the cutting partition surface graph from top to bottom to generate an electronic map in a display range.

The invention belongs to the technical field of big data, and provides a capital flow monitoring method and device based on double-graph fusion calculation, and the method comprises the steps: generating a customer association relationship graph of a target customer based on the ID of the target customer and the customer association relationship of the target customer; according to the fund flow direction and the fund relationship of the target customer, generating a fund flow direction time sequence graph; and monitoring the fund flow of the target customer according to the customer association relationship graph and the fund flow time sequence graph. According to the method, through a bispectrogram fusion calculation thought, a dynamic relationship reflecting a transaction time sequence and a flow direction and a static relationship reflecting a customer association relationship are combined for calculation, the limitation that only fund transactions are concerned in traditional fund flow relationship monitoring is broken through, and exploration of the implicit fund relationship is realized.

The invention relates to a modeling and construction method for cross-organization business process interaction. The method comprises the following steps: constructing a Token log according to actualbusiness process interaction; preprocessing the Token log; constructing an object network of a cross-organization business process based on the preprocessed non-interactive Token log; for an interactive Token log, constructing a node projection graph of an object network according to interactive nodes, and then constructing a system network and a message sequence graph of a cross-organization service process by utilizing a merging algorithm and interactive details. The method can help enterprises to dynamically change the models according to actual implementation conditions from the aspect ofactual models, including task change, execution sequence change and resource change, so as to achieve a common target.

The invention relates to the technical field of natural language processing, in particular to an aspect sentiment analysis method based on a two-channel graph convolutional network, which comprises the following steps: acquiring a to-be-analyzed target text; inputting the target text into a pre-trained aspect sentiment analysis model; firstly, a target text is subjected to text coding, and text semantic representation is generated; then constructing a text sequence graph and an enhanced syntactic dependency graph based on the target text, and modeling based on structural information of the two graphs; fusing the structure information of the two graphs in a mutual enhancement mode through a multi-layer CoGCN to generate a text structure representation; and finally, respectively obtaining sentence representations from the text semantic representation and the text structure representation, and combining to generate a final specific aspect sentence representation. And generating prediction probability distribution on the sentiment label based on the specific aspect sentence representation, namely, an aspect sentiment analysis result of the target text. According to the aspect sentiment analysis method, the accuracy and comprehensiveness of aspect sentiment analysis can be improved.

The invention provides an enterprise innovation and growth ability evaluation method and system based on big data analysis, and relates to the field of quantitative evaluation, and the method comprises the steps: collecting sample data of each enterprise dimension of a sample enterprise, and carrying out the data cleaning, calculation and label classification of the sample data; training an index weight model by using the cleaned and classified sample data so as to obtain a relative importance score of each index item, and calculating a final weight value of each index item by using a superior sequence graph; according to the enterprise data of each to-be-evaluated enterprise and the final weight value of each index item, calculating a final evaluation score of each to-be-evaluated enterprise; wherein the enterprise dimension comprises industry and commerce, employees, qualification, intellectual property, projects, financing loans, assets and research and development costs. According to the method, accurate and reliable quantitative evaluation is carried out on intangible assets, including innovation ability and science and technology projects, of science and technology enterprises, the proportion weight is calculated, the growth ability of the enterprises is accurately measured, and data support is provided for future development planning and decision making of the enterprises.

The invention discloses a project recommendation method of a graph neural network based on directed and undirected structure information, which comprises the following steps of: extracting undirected structure information in a graph by utilizing a graph convolutional network according to an adjacency relation of a session sequence graph, extracting directed structure information in the graph by utilizing a gated graph neural network, calculating to obtain an intermediate project implicit vector, and recommending the intermediate project implicit vector to the session sequence graph according to the intermediate project implicit vector. Performing linear transformation on the obtained intermediate item implicit vector to obtain a final item implicit vector; higher attention is distributed to repeatedly clicked items appearing in the session sequence, an attention mechanism is introduced when item implicit vectors are generated, and weight coefficients of corresponding items are modified according to the degree of dependency between the items. According to the method, the generated session vector is predicted more accurately in the recommendation process.

The invention relates to a transformer fault detection system applying artificial intelligence to a smart power grid, which comprises a preprocessing server, a data analysis server, a graph model construction server and an anomaly detection server, wherein the servers are in communication connection. The preprocessing server preprocesses the abnormal voltage and current data and inputs the abnormal voltage and current data into an abnormal frequency model to output an abnormal characteristic frequency. The data analysis server obtains voltage and current frequency domain data based on the voltage and current data so as to construct a current and voltage spectrogram. The graph model construction server establishes a voltage and current sequence graph model by taking each abnormal characteristic frequency of the current and voltage spectrogram of each transformer in each period as a vertex and taking a difference value of amplitudes of the abnormal characteristic frequencies as a weight coefficient. The anomaly detection server carries out graph aggregation and graph decomposition on all the voltage and current sequence graph models to generate an anomaly detection graph model so as to carry out real-time monitoring and anomaly detection on each transformer.

The invention relates to an abstract extraction method and device, computer equipment and a storage medium. The method comprises the steps that a text keyword graph structure, a text sentence similarity graph structure, a text sentence sequence graph structure and a text feature matrix of sentence position information are obtained through a to-be-extracted text, the text keyword graph structure, the text sentence similarity graph structure and the text sentence sequence graph structure are overlapped, and a heterogeneous graph structure of text features is obtained; and an abstract extraction result is acquired by the heterogeneous graph structure and the text feature matrix. By adopting the method, the graph structures with different features can be obtained, the multi-feature fusion heterogeneous graph structure is constructed through the graph structures with different features, the text feature matrix of the sentence position information is further fused to realize text abstract extraction, different text features are considered in the specific process, and the invention can be applied to texts with complex structures. And the expandability is relatively good, so that the abstract extraction accuracy is improved.

The invention provides an ant colony sequencing and positioning method applied to circular layout, which comprises the following steps of: abstracting a sequencing process for determining a placementsequence of circular objects into an optimal path problem in a graph theory, and obtaining a sequencing graph according to a pattern; abstracting a positioning process for determining the placement position of the circular object into a position selection problem; initializing parameters; according to heuristic information in the sequencing diagram, the ants determine to-be-placed circular objectsfrom the to-be-placed circular set; according to a stimulation-response principle, the ants select one from a plurality of feasible positions to place a circular object; updating the current patternand the to-be-set circle set; the ants complete one-time search; all ants are searched for once, and heuristic information in the sequencing diagram is updated; updating the radius of the container until a stop condition is met; oUTPUTTING Final LAYOUT RESULTS. The method can adapt to different layout environments, and the diversity of search patterns is maintained; and the search space can be reduced, the search speed is increased, and it is ensured that circular objects in the obtained layout are not overlapped.

The invention discloses a vulnerability detection method based on self-supervised learning and multi-channel hypergraph neural network. Motifs representing high-order information are constructed for different channels, and the code sequence graph is sampled according to the motif to obtain a multi-channel code sequence hypergraph. The preprocessed code text data is trained as word vector representation using word2vec. Finally, the code sequence hypergraph and labels are used as training data to train the hypergraph neural network, and the node representation and the hypergraph representation are obtained by learning, and then the hypergraph representation is spliced, and the graph is classified by a single-layer perceptron. The invention simultaneously introduces self-supervised learning to make up for the loss of information between multiple channels, and integrates the mutual information between multiple channels through self-supervised learning, which has better interpretability and vulnerability detection effect. The present invention also provides a system for implementing the above method.

The invention discloses a method for constructing a social behavior sequence diagram. The method comprises the following steps: acquiring a social behavior record set; converting each social behavior record in the social behavior record set into a sequence record in a standard text format; inputting the sequence records into a pre-constructed causal relationship judgment model, and calculating the probability that all other sequence records of which the occurrence time is earlier than that of one sequence record and the sequence record are causal relationships through the causal relationship judgment model to obtain the probability that the sequence records are causal relationships; constructing a node for each social behavior of each person in the social behavior record set, and obtaining the probability of a causal relationship between the nodes; and according to the probability of the causal relationship between the nodes, constructing the social sequence graph. A complete social behavior sequence which is carried out among a plurality of people, contains a plurality of themes and has a clear causal relationship between sentences or behaviors can be extracted, and the use scene is expanded.

The disclosed embodiments concern methods, apparatus, systems and computer program products for genotyping repeat sequences such as medically significant short tandem repeats (STRs). The methods involve aligning reads to a repeat sequence represented by a sequence graph, and using the aligned reads to genotype the repeat sequence. The sequence graph is a directed graph each including at least one self-loop representing a repeat sub-sequence. In some implementations, the reads are paired end reads, and both mates of each read pair may be used to genotype the repeat sequences. Some implementations can be used to determine degenerate codon repeats. Some implementations can be used to genotype repeat sequences each including two or more repeat sub-sequences. Some implementations can be used to genotype nucleic acid sequences each including at least one repeat sub-sequence and another genetic variant such as an insertion, deletion, or substitution.