40results about How to "Sort results are accurate" patented technology

The invention discloses a large-scale human face image searching method. The method comprises the following steps of preprocessing human face images; extracting local characteristics from the human face images; extracting overall geometrical characteristics from the human face images; quantifying the local characteristics; quantifying the overall geometrical characteristics; establishing a reverse index; searching a candidate human face image set; and re-arranging the candidate human face image set. By the method, an index for a large-scale human face image database can be established, quick human face research is realized, and the research efficiency is realized. In addition, the accuracy of human face research is improved by embedding an auxiliary information characteristic quantifying and candidate human face image set re-arranging algorithm. Effective and accurate large-scale human face image search is realized by the method, so that the method has higher use value.

The invention discloses a fuzzy multi-keyword retrieval method of encrypted data in cloud environment. A file is subjected to set encryption by a data owner to generate a ciphertext file; keywords are extracted from each file; the keywords are subjected to binary segmentation and vectorization to obtain a binary vector group; the binary vector group is subjected to dimensionality reduction and is then inserted into a counting type bloom filter to generate index vectors; each index vector is encrypted to obtain a security index; the ciphertext file and the security indexes are sent to a cloud server; a pre-authorized data user or the data owner extracts the keywords from query data; binary segmentation and vectorization are performed to generate a query vector; encryption is performed to obtain a trap door; the trap door is sent to the cloud server; the cloud server obtains a certain number of files with the highest relevancy degree through query according to the trap door and the security index; after sorting, the files are returned to the data owner. The large data volume of ciphertext multi-keyword retrieval is supported; compared with the prior art, the method has the advantages that the index building and query processes are more efficient; the sorting result is more accurate; the data privacy is effectively protected.

The invention discloses a realization method for discovering important users of a social network. According to the method, when the ability of different users in a network to influence each other is solved, node similarity is put forward to measure the degree to which a node is influenced by a neighbor node thereof, and the importance of a node is obtained by comprehensively considering the local influence and global influence of the node. The method of the invention has the following advantages: (1) by using the LeaderRank algorithm as an improved algorithm, the possibility that the PageRank algorithm is trapped in a dangling node is avoided, and the convergence speed of the algorithm is increased; (2) node similarity calculation takes into consideration the incoming edge and the outgoing edge of nodes; and (3) the local and global functions of nodes are considered comprehensively, and the accuracy of the algorithm is improved.

The embodiment of the invention provides an ordering method and device. The method comprises the specific steps of conducting one-time normalizing root-extraction processing on ordering characteristics of ordering objects, so that a corresponding normalizing root-extraction result is obtained, wherein the number of the ordering characteristics is greater than two; according to the normalizing root-extraction result and weight corresponding to the ordering characteristics, obtaining comprehensive scores of the ordering objects; according to the comprehensive scores, ordering the ordering objects, so that a corresponding ordering result is obtained. By means of the ordering method and device, fairness and accuracy of the ordering result can be improved.

The invention provides a method and a system for sequencing search entries. The method comprises the following steps: expressing an inquiry text into a vector according a word order through a neural network; computing sequencing scores between the inquiry text and the search entries according to the expressed vector through the neural network; sequencing the search entries according to the computed sequencing scores. By adopting the method and the system, search sequencing can be performed under the consideration of matching of polysemy and synonyms and the word order of words, and more accurate sequencing results can be given.

The invention discloses an implementation method for discovering important users of a weighting network. When solving the ability of different users in a network to influence each other, the method introduces second-layer neighbor node topology of a node to define expenses for maintaining an edge relation of the node, and the method comprehensively considers node bridging importance and node centrality, thereby obtaining importance of the node. The method adopts a structure hole theory to measure the importance of the node, fully considers the centrality of the node, introduces the second-layer neighbor topology of the node, and distinguishes importance of bridging points of different properties. The method comprehensively considers bridging and centrality functions of the node, thereby improving accuracy of an algorithm.

The embodiment of the invention discloses a method for sequencing objects included in a heterogeneous information network. The method comprises the following steps: acquiring the heterogeneous information network including N object types, wherein the N is larger than 1; sequentially selecting one of the N object types as a center type, mapping the heterogeneous information network into N sub-networks, each sub-network is constituted by one center type, a support type relative to the center type and a link between the center type and the support type; the N sub-networks constitute a sub-network sequence ring; selecting one sub-network from the sub-network sequence ring at random serving as the current sub-network; sequencing objects in the current sub-network; judging that whether the sequencing result is stable or not; outputting the sequencing result if the sequencing result is stable. Compared with the prior art, the method has the advantage that when the objects in the heterogeneous information network are sequenced, semantic information such as object type is utilized, so that the sequencing result is relatively accurate.

The invention discloses a method and a device for sequencing according to contribution degree. The method comprises the following steps: determining sequencing target items and selecting at least two dimensions according to the characteristics of the sequencing target items; calculating sequencing parameters of the sequencing target items on the at least two dimensions; calculating the contribution degree vector of the sequencing target items on the at least two dimensions according to the sequencing parameters; calculating and synthesizing the contribution degree vector for the normalized contribution degree vector according to the weight of the dimensions; and sequencing the sequencing target items according to the synthesized contribution degree vector. According to the method and the device, the contribution degree vector of the sequencing target items on at least two dimensions can be calculated and synthesized, and the sequencing target items are sequenced according to the synthesized contribution degree vector, so the sequencing accuracy degree of the sequencing target items is increased, and the sequence of the sequencing target items accords with the using habits of users.

The objective of the invention is to provide an object ordering method based on user setting and a device thereof. The dimension of an object to be ordered set by a user and the corresponding weight of the dimension are acquired; the score of the object to be ordered is calculated according to the dimension and the corresponding weight thereof; the object to be ordered is ordered according to the score, and the order object is acquired; and the order object is provided for the user. Compared with the methods and the devices in the prior art, the dimension and the corresponding weight of the object to be ordered are set by the user so that the ordered object is enabled to further meet the expectations of the user, and use experience of the user can be enhanced.

The invention discloses a fast iterative ordering method for search engines. The fast iterative ordering method comprises an offline step and an online step, wherein the offline step comprises the following specific steps: training to obtain a plurality of candidate online ordering models, carrying out storing and proportional distribution on all the ordering models, and storing the ordering models and proportions into a search server cache database at fixed time for reading by the online step; the online step comprises the following specific steps: receiving user requests and distributing the ordering models according to user information, retrieving related files from an index, reading the ordering models in the search server cache database, calculating the priority and returning to users, and counting search behaviors of the users at present time. According to the fast iterative ordering method disclosed by the invention, by a set ordering model description mode, one ordering model can be described by using a character string, and then is sequentially stored into a relational database and a key-value database through a graphical interface and a timing task, so that the persistence of data storage can be guaranteed, and online service can quickly acquire the data.

The invention relates to a method for calculating the importance of node in a power CPS system based on an analytic hierarchy process, and the method comprises the steps: analyzing a target, buildinga hierarchical structure based on the analytic hierarchy process, and building judgment matrixes according to the evaluation index data of the importance of the node of an information network; calculating a weight between each two adjacent layers according to the judgment matrix of each layer, obtaining the weight sequence between a bottom layer element and a target, and obtaining the importance sequence of the nodes. The method can solve a problem of how to evaluate the importance of the nodes through a plurality of indexes, reduces the subjectivity of weight selection to a certain degree, and obtains a more objective and complete result.

The invention discloses a maintenance technician recommendation method and device, and a storage medium. The method comprises the steps of obtaining a maintenance demand of a user; extracting maintenance demand parameters in the maintenance demand; obtaining maintenance technicians matched with the maintenance demand parameters from a technician database as a recommendation candidate set; and sorting the maintenance technicians in the recommendation candidate set according to the evaluation grades of the maintenance technicians, and outputting a sorting result. So visible, compared with the prior art in which a distance recommendation method is adopted, according to the method, the recommended candidate set is obtained according to the matching degree with the maintenance demand parameters, the maintenance technicians in the recommended candidate set are ranked according to the evaluation level of the maintenance technicians, the maintenance level of the recommended maintenance technicians is more consistent with the maintenance demand of the user, namely, the recommendation result of the method is more accurate, and the user experience can be improved.

The invention discloses a fuzzy multi-keyword retrieval method of encrypted data in cloud environment. A file is subjected to set encryption by a data owner to generate a ciphertext file; keywords are extracted from each file; the keywords are subjected to binary segmentation and vectorization to obtain a binary vector group; the binary vector group is subjected to dimensionality reduction and is then inserted into a counting type bloom filter to generate index vectors; each index vector is encrypted to obtain a security index; the ciphertext file and the security indexes are sent to a cloud server; a pre-authorized data user or the data owner extracts the keywords from query data; binary segmentation and vectorization are performed to generate a query vector; encryption is performed to obtain a trap door; the trap door is sent to the cloud server; the cloud server obtains a certain number of files with the highest relevancy degree through query according to the trap door and the security index; after sorting, the files are returned to the data owner. The large data volume of ciphertext multi-keyword retrieval is supported; compared with the prior art, the method has the advantages that the index building and query processes are more efficient; the sorting result is more accurate; the data privacy is effectively protected.

The embodiment of the invention discloses a method for sequencing objects in a heterogeneous network. The method is characterized by comprising the following steps: acquiring a first target type and a second target type; converting the heterogeneous network into a bipartite network constituted by the first target type, the second target type and a path between the first target type and the second target type; converting the bipartite network into a corresponding tensor X; acquiring the reachable probability stationary distribution during random walk in the tensor X; sequencing objects in the first target type, objects in the second target type and the path between the first target type and the second target type according to the reachable probability stationary distribution. Compared with the prior art, the method has the advantage that when the objects in the heterogeneous network are sequenced, not only can the path be used to capture semantic information, but also the semantic information can be fully utilized, so that the sequencing result is relatively accurate.

The embodiment of the invention provides a method and a device for recommending a website, wherein the method comprises the steps of obtaining resources accessed by a user within the current statistical period, determining corresponding type of each resource; respectively counting the sum of access times of each type of corresponding resources; sorting each type of determined resources according to the sum of the access times from large to small; and generating a website recommending page at the previous M types of resources according to the order, wherein M is a positive integer. By adopting the method and device disclosed by the invention, the position and compose type of link information of each website in the recommending page can be dynamically adjusted according to the access behavior of the user; personalized website recommending information can be obtained under the absence of an input or search operation of the user; and the operation of the user is simplified.

A search and sorting method and system for judicial texts, wherein the method includes (1) data preprocessing: collecting judicial text data Doc and legal consulting question Query, and performing word segmentation processing on the collected judicial text data Doc, using the word segmentation Data pre-training of judicial text word vectors; (2) building similarity matrix: using pre-trained word vectors to construct similarity matching matrix M of Query and Doc; (3) intercepting relevant text fragments: according to matching matrix M of Query and Doc Extract local correlation text fragments, splicing multiple local correlation text fragments together to obtain correlation text Ds, and splicing matching matrices of corresponding multiple local correlation text fragments together to obtain matrix Ms; (4) Construct feature vector: calculate Query and the global correlation of the relevant text Ds, construct the feature vector F; (5) Calculate the matching value and sort: input the obtained feature vector F into the neural network model, get the matching score of Query and Doc, according to the size of the matching score put in order.