328 results about "Cophenetic correlation" patented technology

The invention provides a text information associating and clustering collecting processing method based on a domain knowledge model. The method comprises the following steps that a text information training set is searched, stemming preprocessing is conducted, and feature word vectors of a text participle sequence of the information training set are extracted through Chinese named entity identification and domain dictionary query modes; representative feature words of a target event are extracted through topic graph model learning training, and a weighted value of topic associating affiliation is calculated; a feature word set is built according to the topic associating affiliation weighted value, calculated through training, of the feature words, and an event topic word template is built; feature word vectors of a participle sequence accessed to text in real time are extracted through the Chinese named entity identification and domain dictionary query modes; the similarity distance of the feature word vectors and all the target event knowledge templates is calculated; the association relationship of multiple texts to the same topic target event is determined according to the similarity threshold, and classification reorganization is conducted by means of a similarity distance ordering rule.

The invention relates to a method for determining the change of functional points by analysing the change of codes: firstly, all changes of source codes to be compared are searched and collected on textual level and a diff result is generated; secondly, an incidence relation between the modified source codes and an interface function or method by analysing the structure relation of the source codes to be compared and a call graph result is generated; the diff result and the call graph result are finally analysed, all affected interface functions or methods are output in the form of graph or statement to generate an affected result graph. By the code analysis, the external function point which is affected when the version is updated, thus effectively assisting the testing and updating operation of new software versions.

Apparatus for matching a query image against a catalog of images, comprises: a feature extraction unit operative for extracting principle features from said query image; a relationship unit operative for establishing relationships between a given principle feature and other features in the image, and adding said relationships as relationship information alongside said principle features; and a first comparison unit operative for comparing principle features and associated relationship information of said query image with principle features and associated relationship information of images of said catalog to find candidate matches.

The invention discloses a method for establishing a medical image atlas based on image segmentation and the CNN. A CNN method is used to segment medical images finely, image segmentation identification is carried out on the collected medical images, medical knowledge unit segmentation is carried out on medical diagnosis knowledge, and knowledge units of different medical images are formed; an image relation of image segmentation is analyzed to connect the different image knowledge units and communicate with an additional medical illness knowledge unit, a knowledge base association relation isestablished on the basis of connection between medical image identification and medical diagnosis knowledge; characteristic matching is carried out on the knowledge units after image segmentation, entities with the same type of attribute serves as entity nodes in the knowledge atlas, a knowledge base entity is established by fusion, and entities are aligned; and a knowledge atlas is established byusing the entities and the associations therebetween. Suspected diagnosis prompt and treatment suggestion are provided, and AI diagnosis is made possible.

The invention discloses a sorting method for correlation search results and a relevant device. The method comprises: Perform resolution on triad information of each main body example, in order to construct an example correlation diagram; according to the two input examples, perform ergode on all correlation paths between the two examples in the example correlation diagram, and create search results information; calculate the field of relativity and/or correlation length and/or correlation frequency; perform sorting on the search result information. The device comprises: A body resolution module, a correlation search module and a correlation sorting module. It is known through the technical scheme that, through calculation on such parameters as field of relativity and/or correlation length and/or correlation frequency, the invention can fulfill flexible sorting for the search results, so that users can obtain anticipated information effectively from the search results.

The invention relates to the text analysis field, particularly to a semantic characteristic-based text semantic similarity analysis method. According to the technical scheme, the similarity degree between texts is analyzed more accurately and effectively by calculation based on semantic relations of internal words of the texts. According to the method, shallow analysis on association relation between texts and between lexical items is performed through singular value decomposition; a lexical item-theme set is constructed by a bayesian network; the semantic similarity between the lexical items is calculated by mutual information and context; and finally, the text similarity is calculated through a graph structure. By adoption of the text semantic similarity analysis method, the semantic relation between texts can be measured and recognized more accurately and effectively.

The invention provides a Python code change reminding method based on abstract syntax tree node change extraction. The method comprises the following steps of: (1) obtaining source code of different-version programs of the same software; (2) generating abstract syntax trees corresponding to two versions of source programs; (3) matching the abstract syntax trees to obtain a change node, and combining change node context information to mark the node change type; (4) gathering change tuples into transactions to construct a training set; (5) mining the association relationship in the change tuples by using a data mining technology; and (6) reminding a developer of possibly occurring change positions and possible change types in the according to the programs mined association relationship. The Python code change reminding method solves the problems that in the prior art, the possible code change type cannot be reminded due to the lack of software evolution analysis aiming at the python language, and the like. The management of the software life cycle is further guided, and the software evolution controllability is improved, so that the quality of a software product can be well controlled.

The invention discloses an online knowledge map based on a named entity library, which comprises the following steps: step 1) crawling a webpage and crawling webpage information from an entire internet to obtain an internet data set; step 2) extracting named entities; step 3) searching attribute words, traversing the internet data set in the step 1) and establishing an attribute set related to the named entities; step 4) quantifying the information and a correlation between the attribute words in the attribute set and the named entities; step 5) establishing the knowledge map and establishing a topological structure among the named entities through the correlation between the named entities and the attribute words in the attribute set to form the knowledge map; step 6) updating. The online knowledge map has the benefits that the named entities are extracted from a massive new page text message, a correlation between entity nodes and nodes is automatically established, the knowledge map is established according to knowledge nodes and a correlation among the knowledge nodes, and the updating of the knowledge map can be efficiently realized.

The invention provides an online education assessment system. The online education assessment system comprises a knowledge graph unit, a test unit, an assessment unit and a result generating unit. The knowledge graph unit contains one or more knowledge points and association relations between the knowledge points. The test unit contains one or more questions to be tested, and weighted values of the questions to be tested are determined according to associations between the bound questions to be tested and the knowledge points. According to the weighted values, the assessment unit selects part or all questions to be tested from the test as assessment questions which are subjected to assessing by the examinee. The result generating unit generates an assessment result report based on assessment result of the examinee. According to the technical scheme, relations between the knowledge points to be tested are formed into the graphs, and in the ultimate assessment report, the knowledge points which the examinee does not understand comprehensively are displayed.

The invention relates to a webpage training method and system and a webpage prediction method and system. The webpage training method comprises obtaining a prior probability table of classified keywords according to existing data which are associated with the classified keywords; preprocessing a webpage to be trained to obtain a webpage text to be trained; extracting features in the webpage text to be trained according to the prior probability table to obtain an association relation feature vector representation F1 between the webpage to be trained and a specified category; performing model training on the association relation feature vector representation F1 to obtain a classification result of the webpage to be trained. According to the webpage training method and system and the webpage prediction method and system, category systems which are strong in heterogeneity can be simultaneously processed, the large category systems can be processed through less training data, and the problem of data sparseness is largely solved due to the fact that the browsing and search behavior of a user on the whole Internet not just a website is collected.

The invention relates to a text abstract generation method based on a sentence association attention mechanism, and belongs to the technical field of natural language processing. The method comprisesthe following steps: firstly, encoding a document by using a layered bi-directional long short-term memory Bi-LSTM network to obtain sentence semantic vectors, then analyzing an association relationship among sentences by virtue of a gating network to realize sentence-level importance and redundancy evaluation, and finally, providing a decoding algorithm based on a sentence association attention mechanism to generate an abstract. When a neural network abstract generation framework is constructed, sentence relevance analysis is fused, and the evaluation capacity of the model for sentence importance and redundancy in an original text is improved. According to the method, the performance of the generative abstract is effectively improved, and a relatively good effect is achieved on the current ROUGH evaluation index.

The embodiment of the invention discloses an establishing method of a knowledge base and an information search method and device based on the knowledge base. The establishing method of the knowledge base comprises the steps that a sentence to be processed is decomposed into substructures comprising entities according to the syntactic structure of the sentence to be processed; data processing is conducted on the substructures, knowledge data are generated, the knowledge data comprise at least two entities, and the entities meet the set incidence relation; the knowledge data are stored in the knowledge base according to a set data structure. The knowledge base which can inquire question and answer inquiring type answers is constructed, the inquire question and answer inquiring type answers input by a user can serve as rewriting words to optimize the existing rewriting transformation technology, the increasing individual and accurate Internet searching request of people is met, the search relevance is greatly improved, and meanwhile the frequency of secondary inquiring is reduced.

The invention discloses a retrieval result clustering method facing to a Web2.0 label picture shared space, which comprises the following steps: excavating a vocabulary relationship and an associated relationship between labels; obtaining an expanded querying label set by a query label according to the vocabulary relationship between the labels; obtaining a candidate image set relevant to query by the expanded query label set; selecting front K most relevant labels according to the relevance measurement of the labels in the query label set and the candidate image set; automatically dividing the K labels into an optimal clustering result according to the association between the K labels by a clustering algorithm based on a picture division from top to bottom; and correspondingly clustering the candidate image set according to clustering labels. Aiming at the problem of inconformity of label expression, the effective query expansion is realized, and the image clustering method based on most relevant label set clustering solves the problem of diversity of label semanteme. Compared with a traditional method, the invention leads a user to rapidly and effectively retrieve and browse a picture in the Web2.0 label picture shared space.

The invention provides an enterprise investment risk assessment method. The method comprises steps: news corpus associated with an enterprise entity which is an investment target is crawled, and otherentities associated with the enterprise entity are extracted; with names as nodes and the association relationships among the enterprise entity and other entities as edges, a relationship network isbuilt; the vector representation of the enterprise entity is calculated, and a first eigenvector of the enterprise entity is generated; according to a first preset rule, the inner information of the enterprise entity is quantified to generate a second eigenvector; according to a second preset rule, the outer information of the enterprise entity is quantified to generate a third eigenvector; and the first eigenvector, the second eigenvector and the third eigenvector are inputted to an enterprise risk assessment model, and a risk label corresponding to the enterprise entity is outputted and obtained. The invention also provides an electronic device and a computer readable storage medium. Thus, the information revealed in the news corpus is analyzed to assess the risk of target enterprise investment.

The invention discloses an image retrieval method based on sparse non-negative matrix decomposition. It includes the following steps: 1) Query and extract the image and accompanying text of the retrieval results under two different image data sources; 2) Extract the tags in the accompanying text, and filter the results according to word frequency to form a vocabulary; 3) For each Image set, using the association relationship between labels and images to form an association matrix between labels and images; 4) Using sparse non-negative matrix decomposition to analyze the association matrix obtained in step 3), to obtain the shared subspaces of different sources of data and their corresponding independent subspaces Subspace; 5) The user sends a retrieval request for images on a data source, forms a query vector and maps it to the corresponding subspace of the data source, calculates the similarity with all images and sorts them, and returns the top N most similar images. The present invention makes full use of the associated knowledge of tags and images under multiple data sources, performs migration learning through sparse non-negative matrix decomposition, and improves the accuracy of image retrieval on target data sources.

A coverage analysis system for network optimization of mobile communication system, includes: a data import and export module; and individual analysis module, for determining coverage problem regions and coverage problem plots, call problem regions and call problem plots, and collecting the network configuration characteristic parameters; a comprehensive diagnostic analysis module, for comprehensively analyzing the information and parameters contained in the route measured data and the call testing data to determine the association correlation of the coverage problem regions and the call problem regions, and processing correlation andanalysis to determine the coverage and call associated problem regions and problem plots, so as to evaluate the coverage quality , call quality and network performance of the entire mobile communications network from different point. The present invention also provides a corresponding method. The present invention overcomes the problem in the existing technology that it can only analyze based on a single problem plot and a single parameter, and achieves the technical effects that it can carry out the overallanalysis analyze from the entire point of network, thus having a unified data management and strong data compatibility.

The invention provides an audio-video searching method and apparatus as well as a terminal. The method comprises the steps of receiving an audio-video query statement; determining a keyword in the audio-video query statement; obtaining at least one instance with an associative relationship with the keyword; and constructing a search result knowledge graph according to the keyword, the instance and the associative relationship between the keyword and the instance. The search result knowledge graph constructed by embodiments of the audio-video searching method and apparatus as well as the terminal displays the keyword of the audio-video query statement, the instance with the associative relationship with the keyword and the associative relationship between the instance and the keyword in a graph form; correspondingly, the search result knowledge graph comprises not only audio-video search results corresponding to the audio-video query statement but also an associative relationship among the audio-video search results; the search results are visual and concise in form; and contained search result information is rich, so that a user can conveniently view the information and the user experience is greatly improved.

The invention provides a method and a system for analyzing a fault root cause. The method comprises the following steps that: according to the time attribute of each piece of data in a dataset, carrying out sorting, and segmenting the dataset according to a preset time window to obtain multiple groups of sub-datasets; according to an Apriori algorithm, obtaining a frequent item set and an association rule in the dataset, wherein the frequent item set contains a certain quantity of data with a strong association relationship; and according to the time attributes of the data in the frequent itemset, carrying out sorting, matching the data which ranks higher with adjoint warning cause data pre-stored in a warning cause database in sequence, if the data which ranks higher is successfully matched with the adjoint warning cause data pre-stored in the warning cause database, removing the data which ranks higher from the dataset, and continuously carrying out comparison from a next data itemuntil the data which fails to be matched and ranks higher is taken as the root cause of the data of which the time sequence ranks at the bottom in the frequent item set. The method is suitable for thefull-dimension monitoring scene of an IT (Information Technology) system, the pressure of operation and maintenance personnel is released, and the quality requirements of the operation and maintenance personnel are lowered.

The invention relates to a cross-media retrieval method based on uniform sparse representation. The method comprises the following steps that a cross-media database comprising a plurality of media types is established and a feature vector of each kind of media data is extracted; a feature mapping matrix for cross-media uniform sparse representation is studied for each media type, the incidence relations between all the different media types are considered, spatial sparsity of the cross-media data after mapping is conducted is also considered, and the feature spatial sample distribution after mapping and the feature spatial sample distribution before mapping are kept approximate; the probability that every two media data belong to the same type serves as the similarity between different media; the similarity between media data in query samples and a query target set is calculated, and the most similar cross-media retrieval result comprising all the media types is output. According to the cross-media retrieval method based on uniform sparse representation, the incidence relations, the sparsity and the sample distribution between the cross-media data are fully considered, the effectiveness of uniform feature representation can be improved, and therefore the accuracy of cross-media retrieval is improved.

The invention relates to a deep neural network in artificial intelligence, and provides a default probability prediction method and device, computer equipment and a storage medium. The method comprises the steps of acquiring an insurance source identifier, and inquiring whether a matched target entity exists in a preset knowledge graph or not according to the insurance source identifier; When thematched target entity exists, obtaining a to-be-predicted entity associated with the target entity in a preset knowledge graph; And based on the state attribute information of the to-be-predicted entity and the association relationship between the to-be-predicted entity and the target entity, obtaining a risk early warning feature vector, inputting the risk early warning feature vector into a trained risk early warning model for prediction, obtaining a default probability vector, and obtaining the default probability of the to-be-predicted entity according to the default probability vector. Byadopting the method, the efficiency and accuracy of default probability prediction can be improved.

The invention discloses a risk gang identification method and device. The method comprises the following steps: constructing a risk knowledge graph based on historical behaviors of a trusted user, wherein the risk knowledge graph comprises a non-user node, a trusted user node and a common user node associated with the trusted user; Extracting nodes matched with a risk template and connection edgesbetween the nodes from the risk knowledge map to form a risk map; For each node in the risk map, according to the time parameter of the occurrence time of the incidence relation between the node anda neighbor node and the degree of the neighbor node, carrying out iteration on the label of the node, and after one round of iteration is completed, judging whether a modularity function with time aggregation is converged or not; And when the modularity functions with the time aggregation converge, identifying the nodes with the same tags as the same risk gangs.

The application discloses a method and an apparatus for establishing user information correlation of users. A specific embodiment of the method for establishing the user information correlation of the users comprises: obtaining positioning information of the plurality of users; based on the positioning information, finding out user information of the users with a correlation relationship, wherein the correlation relationship is that the quantity of matched positioning information of the users is greater than a quantity threshold value, and the matched positioning information meets the following matching conditions: a D-value of locating positions is less than a distance threshold value and a D-value between positioning time is less than a time threshold value; based on the matched positioning information, calculating a correlation parameter between the user information of the users with the correlation relationship; and based on the correlation parameter, establishing the correlation between the user information of the users with the correlation relationship. The correlation between the user information of the users is established by fully utilizing network resources, and at the same time, a mode for establishing the correlation between the user information of the users is expanded.

The invention relates to an electronic device, a litigation data processing method and a storage medium. The method comprises: obtaining a judgment document of a predetermined type of lawsuit, performing analyzing to obtain the focus of the dispute; based on the relationship between the disputed focus in the pre-established evidence list and the name of evidence and the disputed focus obtained ineach decision document, analyzing the judgment document to obtain the valid evidence information matched with the obtained disputed focus in the decision document; based on the focus sentences obtained from each judgment document, analyzing the corresponding judgment documents to obtain the judgment sentences; based on the judgment sentences of the judgment document, obtaining the corresponding judgment results, and establishing the relationship among the focus sentences, the evidence information, the judgment sentences and the judgment results; training the Bayesian model to get the pre-judgment model corresponding to the predetermined type of litigation. The invention can improve the accuracy rate of litigation pre-judgment and effectively guide relevant parties to make corresponding operation.