172 results about "Pagerank algorithm" patented technology

The invention discloses a club detecting method based on network topology and node attribute. The club detecting method based on the network topology and the node attribute comprises the steps of: firstly, obtain significance of each node in a network based on global topology by utilizing a PageRank algorithm and calculating a structural linkage closeness between nodes by utilizing the neighbor structure of the nodes; then adopting different extraction ways and attribute similarity calculation ways based on different forms of the node attribute; then adjusting the balance by using a weight adjustment factor; finally regarding the similarity of the node neighbor and the category center node neighbor as the degree of the node which belongs to the category in a clustering process. According to the club detecting method based on the network topology and the node attribute disclosed by the invention, the obtained club is relatively high in linkage closeness and relatively good in homogeneity.

A traffic signal self-adaptive control method based on dynamic priority belongs to the technical field of intelligent transportation, and includes two parts, namely the optimal phase control strategy and stability supervisory mechanism, eliminates parameters such as period, split and phase difference in a traditional traffic signal control system, wherein the main process of the optimal phase control strategy includes the steps of modeling for a road network, building a directed weighted graph, calculating the dynamic priority of permission segments through the PageRank algorithm based on pagerank as per the built directed weighted graph, confirming the sequence of signal phases as per the priority of permission segments, and distributing the lighting time of green lights; as a supplement of the optimal phase control strategy, the stability supervisory mechanism is mainly in charge of supervising the permission conditions of all the segments in the traffic road network, and is used for granting the permission of one segment if the segment without the permission causes instability of the system, thereby guaranteeing the stability of the system. As parameters such as period, split and phase difference in the traditional traffic signal control system are eliminated, real-time response to the variation of traffic flow is realized, and the system is stable after the parameters such as period are removed.

The invention relates to a focusing relevancy ordering method for a vertical search engine. Aiming at the problem that a focused crawler cannot pass through a dark tunnel, the invention improves a focused crawling strategy of the focused crawler by using an on-line learning method and utilizing an auxiliary function, so as to lead the focused crawler to capture subject data with higher relevancy. A PageRank algorithm and an improved algorithm thereof are studied, the webpage clicking action of a user is modeled, and the transferring way of a PageRank value among links is improved, so as to put forward the improved algorithm. As to the disadvantage that the dimensionality of a feature extraction model of webpage weight is over high, a user-defined method of the webpage weight is put forward, so as to define a factor of the webpage weight and measure the weight of the factor of the webpage weight according to the divisibility criterion, thereby providing an evaluation function of the webpage weight and effectively lowering the dimensionality of a webpage feature space. By utilizing the method in the invention, the user can obtain a high-quality search result set when using a subject resource search engine system.

The present invention discloses an entity linkage algorithm based on a graph model. The entity linkage algorithm based on a graph model is characterized by comprising: forming a candidate entity by using the Wikipedia knowledge base; constructing a semantic feature between the entities by using LDA; constructing relationships between entities based on linkage structures of Wikipedia to form a graph model; and integrating related semantic features into the graph model; and ranking the entities by using the PageRank algorithm to obtain an entity linkage result, which specifically comprises steps of calculation and integration of a naming dictionary, a candidate entity set, related features, construction of the graph model, and ranking of candidate entities. Compared with the prior art, the entity linkage algorithm based on a graph model has the advantages of being good in entity feature integration and high in reliability of the entity linkage result; data is downloaded by using Wikipedia, so that no additional costs are needed, and especially data sets do not need to be noted manually; and the method is simple, convenient in usage, and saves time and efforts.

The invention relates to a micro-service intelligent monitoring method for abnormal propagation, which comprises the following steps of: monitoring service calling information based on an agent technology, and establishing a micro-service calling topological graph to describe an abnormal propagation relationship among micro-services; a Lasso regression modeling interface is adopted to call correlation with measurement, and abnormal micro-service is detected by monitoring the change of a correlation model; based on the PageRank algorithm, the abnormal degree of the micro-service and the callinginterface of the micro-service is evaluated, transparent service monitoring is achieved, automatic metric value prediction is achieved to find abnormal service, and the abnormal degree of nodes in agraph is intelligently evaluated to detect the problem root cause.

The invention discloses a socialization filtering method based on a PageRank algorithm, mainly solving the problem that a filtering method has low accuracy under the conditions that group members are numerous and social relationship is complex in the prior art. The socialization filtering method disclosed by the invention is realized by the following steps: acquiring friend relationship between a group and the group members from a webpage configuration file, creating a personal preference model of each group member; by adopting the PageBank algorithm, iteratively calculating the influence of the group members to the group, so as to obtain a preference model of the whole group; carrying out object recommendation by utilizing the model, namely initiatively providing recommended object data for a user; selecting interested and required information from the information by the user. The socialization filtering method disclosed by the invention has the advantages that the preference model of the group is analyzed, and recommendation of different field objects on a network can be realized only by modifying a keyword vector in the field of the model.

The invention relates to the technical field of network and information search, and discloses a correlation fraction distribution-based method for classifying query intentions, which comprises the following steps of: S1, obtaining a queried search result and a webpage; S2, constructing a search result set according to the search result and the webpage; S3, measuring a correlation fraction of a document in the search result set; and S4, classifying the query intentions by using the correlation fraction distribution. By adopting an improved Hits algorithm, an improved PageRank algorithm and an improved search model, the correlation fraction of the search result is obtained, the problem of insufficient queried and clicked logs of long-tailed distribution in the traditional scheme is solved, and the problem of incapability of finding matched anchor text sets or fewer elements in the matched anchor text sets in anchor texts in the anchor text-based method is also solved.

The invention discloses a microblog interestingness circle mining method based on intimacy degree and influence power and a microblog interestingness circle mining device based on the intimacy degree and the influence power. The mining method comprises the following steps of discovering a social intercourse interestingness circle seed on a center user first-stage interaction diagram on the basis of a KCC (K-Clique-Community) algorithm; expanding the social intercourse interestingness circle seed according to the intimacy degree among nodes; expanding a PageRank algorithm through the user microblog interesting similarity degree, and calculating the user influence power; expanding the expanded social intercourse interestingness circle seed again through the user influence power; and automatically marking the discovered social intercourse interestingness circle through the re-expanded social intercourse interestingness circle. The mining device comprises a discovering module, a first expansion module, a calculation module, a second expansion module and a marking module. The social intercourse interestingness circle obtained through mining by the method and the device can be applied to various fields such as interestingness modeling, cooperated recommendation, personalized searching and ranking, precise advertisement putting and knowledge mapping; and wide application prospects and values are realized.

The invention discloses a patent value evaluation method based on value delivery. The patent value evaluation method includes a patent value static evaluation process and a dynamic renewal process. In the patent value evaluation process, firstly, a patent quotation network is built according to similarities between patents, and patent value in each node is initialized; secondly, contribution degree is calculated by value from a 'quoted' patent to a target patent according to node value of the 'quoted' patent, and patent quantity quoted by the 'quoted' node and patent published time, and then target patent value is acquired by accumulation; finally, each patent value is sequentially calculated. When a new patent is available, the similarities between the patent and an existing patent are firstly calculated, and then a 'quoted' relationship is built up, and finally values of patents are sequentially updated. The patent value evaluation method can calculate value of each patent from a patent database based on a delivery idea of PageRank algorithm, and has high effectiveness.

The invention discloses a rail transit key node and a key road section identification method. The method includes: based on a complex network theory and a PageRank algorithm, calculating a topologicalindex value of the rail network as a node importance evaluation system value; carrying out evaluation value calculation on a node importance evaluation matrix by adopting three evaluation methods ofa grey correlation analysis method, a TOPSIS method and a principal component analysis method, determining weights of evaluation results by utilizing an entropy weight method, and finally obtaining importance values of the nodes and the road sections so as to identify the key nodes and the key road sections. According to the invention, identification of the key nodes of the rail transit network isobjectively and comprehensively considered; the key node importance degree calculation method is designed by using relevant theories such as a complex network and comprehensive evaluation, stations needing important maintenance in rail transit operation can be judged, the importance degrees of all the stations are ranked, and a choice is further provided for robustness analysis of the key nodes of the rail transit network.

The invention discloses a webpage ranking method based on cloud computation. The method comprises the following steps of analyzing a network file which is crawled by a distributive webpage crawler to obtain a basic topological structure information file of a network; offline calculating a PR value and then storing the PR value into a corresponding document table, wherein the format of the document table adopts url as a main key , and the format containing eight attribute columns containing title, content, type, timestamp, outlinks and the like; adopting a parallel computation technology for establishing an index table of single word - webpage importance, wherein the format of the index table is a format established by a reverse index and containing key and links (link set and sorted according to the importance; adopting a MapReduce parallel architecture to realize the offline PageRank algorithm; and comparing the similarity of an inquiry word and a webpage for online inquiry, and giving a final webpage rank according to the offline inquiry result. The method has the advantages that the offline ranking algorithm is adopted, the MapReduce parallel arhictecture is adequately utilized, so that the offline ranking efficiency is improved; and by adopting the technology combining the key word technology and the PageRank technology, the result is more accurate.

The invention provides an opinion leader mining method based on PageRank, belonging to the data mining field. The invention is based on the core idea of the PageRank algorithm, and is introduced intothe social network user influence analysis. The invention not only considers the topological structure information of the social network user relationship, but also fuses some attribute information ofthe social network user which embodies its influence. At that aspect of social network topology information, the algorithm construct a weighted user relationship network base on the forwarding relationship of users in the event, Event is subject-oriented and a large number of users participate in it. Using event-based forwarding mechanism to build relational network not only has the pertinence and goal of information, but also can obviously reduce the use of social media information data, and this kind of network is dynamic because of the different events. For the social network user attribute information, the algorithm uses the user attribute information which can directly reflect the user influence, and can more accurately depict the user influence.

The present invention discloses a content based junk webpage detecting method and a detecting apparatus thereof. The method comprises: calculating a maximum content similarity-degree value of all webpages and seed junk webpages, and generating a similarity-degree set; sorting all the webpages in descending order by using a PageRank algorithm; based on a sorting result, searching the similarity-degree set for a content similarity-degree value of the webpages and the sample junk webpages; and comparing the similarity-degree value with a similarity-degree threshold, performing detection on the webpages, and adding detected junk webpages into a junk webpage set. The apparatus comprises a generation module, a sorting module, a search module and a detection module. By means of the modules, determination of a webpage content similarity degree is added into the method provided by the present invention on the basis of the conventional PageRank algorithm; links and contents of the webpages are combined; and detection is performed on the junk webpages, thereby improving accuracy and efficiency of junk webpage detection.

The invention discloses a short-time traffic flow prediction method considering a diffusion process, the method comprises the following steps: obtaining a historical traffic flow sequence of the current road section and performing a smoothing operation to obtain a smoothed traffic flow sequence F = {X1, X2 - X<t-1>}; adopting an LSTM-CNN model to capture a depth spatial-temporal feature from the smoothed traffic flow sequence; performing a diffusion process of digital description through a class PageRank algorithm to obtain a road importance feature of the current road section from the smoothed traffic flow sequence, wherein the road importance feature and road auxiliary information are combined to form a one-dimensional vector serving as a breadth feature; and fusing the depth spatial-temporal feature and the breadth feature to obtain a traffic flow prediction value Xt of the current road section at the t moment.

The invention discloses a rating method for an influence of a mobile new APP. The method comprises: (1), data of a news APP are collected and the data are clustered and are stored into a library, wherein the data include a website, the number of times of making comments, the number of times of reprinting, a daily page view (PV) and a unique visitor (UV); (2), a news influence factor is calculated; (3), a news App reprint rate is calculated by using a PageRank algorithm; (4), a news App replying rate is obtained by inquiring a replying rate reference table; and (5), a news APP score is calculated by a calculation model and the news App is ranked based on the score. According to the invention, information is displayed for the user in a manner of scores and thus the influences of different news Apps are displayed visually. The numerical value is the evaluation score of the news App; the higher the score of the news App, the better the user's need is met. Therefore, real-time news can be provided for the public timely.

The invention discloses a PageRank and information entropy-based text word segmentation method for a judgment document, and belongs to a Chinese word segmentation technology in the field of natural language processing. An improved PageRank algorithm, an information entropy, mutual information and a keyword dictionary are mainly adopted to carry out word segmentation on a Chinese text. For the judgment document in the legal field, the word segmentation method is established on the basis of the PageRank algorithm; candidate words are segmented according to Rank vectors; the candidate words are corrected through the information entropy; terms are combined according to a keyword dictionary of the judgment document; and finally a word segmentation result is output. The method can carry out wordsegmentation on the judgment document more accurately. Compared with an existing method, the method has the remarkable advantages that statistics or training does not need to be carried out through alarge number of text corpora to establish a large-scale dictionary, and only the input text is subjected to statistics; the input text is used as an existing corpus to carry out statistical mining; and finally the word segmentation can be completed in combination with a keyword term dictionary of the judgment document.

The invention discloses a natural language processing technology-based bad asset operation management method for an asset management company. The method comprises two parts including data importing inknowledge base construction and deep learning and PageRank-based keyword extraction. A new word model is discovered by utilizing a specific word bank and an HMM to perform word segmentation processing on bad asset operation knowledge, so that the text word segmentation accuracy is improved and a perfecter text word bank is established; word vectors are trained through a deep learning method, so that the phenomenon of "curse of dimensionality" represented with the word vectors can be avoided, information of vocabulary contexts can be fully mined, and relationships between words can be obtained; and based on an improved PageRank algorithm, a topology matrix of word connection is obtained according to a word sequence relationship in a specific contract, cosine values between the word vectorsserve as connection weights of the words, word vector information obtained by training the word bank is fully utilized, and a vocabulary relative position relationship in a text is mined, so that a powerful theoretical basis is provided.

The invention relates to an abnormal-user detection method based on session feature similarity fuzzy-clustering. The method includes the following steps: step S1, establishing a sliding window for each session, and collecting webpage access information of users; step S2, carrying out data preprocessing on the collected webpage access information to generate user session information; step S3, usinga PageRank algorithm to calculate webpage weight information; step S4, using a SimHash algorithm to calculate similarity among the users on the basis of the obtained user session information and webpage weight information, and establishing a user similarity matrix; step S5, using a fuzzy clustering-based lambda-cut algorithm to cut the user similarity matrix, which is obtained by the sliding window of each session, to obtain suspect users; and step S6, detecting and locating abnormal users according to suspect-user information returned by the sliding windows of all the sessions. The method facilitates fast and accurate detection and locating of the abnormal users.

The invention discloses a personalized community recommendation method based on user behaviors, and relates to social networks. Social network micro-blogs are used as a platform to analyze multi-attribute information of static attributes and dynamic attributes of users. Firstly, two aspects of bloggers followed by the micro-blog users and communities which the micro-blog users participate in are considered in a process of calculating user similarity degrees, and a traditional Jaccard similarity degree calculation method is extended to obtain a user similarity set; and then the similarity set is further screened from a perspective of user influences. The influences of the micro-blog users in the communities are related to numbers of fans thereof, and also to numbers of comments and forwarding on the micro-blogs thereof. On the basis thereof, a traditional PageRank algorithm is improved to calculate the user influences; and finally, Top-N is utilized to sort influence sizes to obtain final recommendation object sets. Experiment proves that an algorithm of the invention effectively solves the problem of inaccuracy of results obtained by traditional personalized recommendation algorithms, and greatly improves a surprise degree and novelty of recommendation.

The invention discloses a PageRank-based social contact big data information maximizing method. According to the method, friendly relations and hostile relations in social contact big data are extracted; a PageRank algorithm is applied to a friendly relation adjacency matrix and a hostile relation adjacency matrix separately; and seed nodes are selected. The method is aimed at a social network with friendly relations and hostile relations; the influence of the hostile relations on information dissemination is considered; and the PageRank algorithm is adopted to calculate the difference of the PR value of the friendly relations and the PR value of the hostile relations, so that the influence of a user can be measured, and the effective seed node set is selected, and therefore, information can be spread in a maximized manner through the friendly relations and hostile relations in the online social network.

The invention discloses a differentiation webpage ranking method based on PageRank. The method includes: using the PageRank algorithm to calculate the initial webpage ranking of each webpage; evaluating the authority allocation weight of each webpage according to the inlink number of the webpage to acquire a webpage ranking calculation formula; evaluating the authority allocation weight of each webpage according to the total link number of the webpage to obtain another webpage ranking calculation formula; using the webpage ranking calculation formulas to perform iteration calculation on the webpage ranking of each webpage until the webpage ranking of each webpage is stable; using detected garbage webpage number, recall rate, precision and F-Measure value to perform experimental comparisonanalysis to measure the differentiation webpage ranking method so as to achieve effectiveness of PageRank ranking. By the method, the problem that the ranking result of the PageRank algorithm is unreliable due to average link weight allocation is solved.