127 results about "Dovetailing" patented technology

Systems, software, and apparatuses that provide wired and wireless telecommunications under conditions where signal strength is poor or intermittent, the coordination and synchronization of data and workflows across various communication links under such conditions, especially intermittent or unreliable communications links, and the management of wireless mobile applications in such environments. The present invention technology herein relates to the fields of computer science, telecommunications, and data management.

Systems, software, and apparatuses that provide wired and wireless telecommunications under conditions where signal strength is poor or intermittent, the coordination and synchronization of data and workflows across various communication links under such conditions, especially intermittent or unreliable communications links, and the management of wireless mobile applications in such environments. The present invention technology herein relates to the fields of computer science, telecommunications, and data management.

The invention discloses a college teaching cloud platform based on micro services. A campus network and private cloud are combined through a virtual private network to build a cloud platform, and a micro service architecture is utilized to build services of the cloud platform; using an open source project Eucalyptus developed by the Department of Computer Science of the University of California to set up private cloud; and all requests of a client first pass through an API gateway, then the API gateway routes the requests to appropriate micro services, and generally the APT gateway calls a plurality of micro services and merges a result to process a request. According to the invention, the micro service architecture is used for the college teaching cloud platform, application of the teaching cloud platform is decoupled into micro services of different functions, thereby enabling each link of teaching to be called in a micro service mode, the response is quick, and the efficiency is high; and the private cloud and the campus network are connected through the virtual private network to form the college teaching cloud platform, thereby providing better resource sharing for teachers and students, and the college teaching cloud platform has good elasticity and expansibility.

The invention is applicable to the technical field of computer science, and provides a video based behavior recognition method and device. The video based behavior recognition method comprises the steps of extracting depth features of all sampling time video frames, wherein the depth features comprise high-level semantic features and detail features of behaviors in the sampling time video frames; acquiring spatial-temporal features used for expressing a behavior of the current time video frame based on the detail features of the sampling time video frames; inputting the spatial-temporal features and the high-level semantic features into an LSTM (Long Short Term Memory) model so as to perform behavior recognition on the current time video frame. The spatial-temporal features and the high-level semantic features are enabled to cooperate in a complementary manner so as to enhance the complex behavior recognition capacity of an RNN (Recurrent Neural Network).

The invention relates to a method for identifying multi-class facial expressions at a high precision based on Haar-like features, which belongs to the technical field of computer science and graphic image process. Firstly, the high-accuracy face detection is achieved by using the Haar-like features and a series-wound face detection classifier; further, the feature selection is carried out on the high-dimension Haar-like feature by using the Ada Boost. MH algorithm; and finally, the expression classifier training is carried out by using the Random Forest algorithm to complete the expression identification. Compared with the prior art, the method can reduce training and identifying time while increasing the multi-class expression identification rate, and can implement the parallelization conveniently to increase the identification rate and meet the requirement of real-time processing and mobile computing. The method can identify the static image and the dynamic image at a high precision, is not only applicable to the desktop computer but also to the mobile computing platforms, such as cellphone, tablet personal computer and the like.

The invention provides a speech noise reduction method based on audio recognition, which reduces the noise of a receiving end by aiming at the speech communication under complex noise environment, belonging to the field of computer science and technology. Most of the existing noise reduction methods are only suitable for stable noise environment and can not remove the noise under the situations of complex noise environment, especially the situation of frequent mutagenicity noise and the like. The method leads a mode recognition idea in the communication speech noise reduction, divides an audio signal into a speech signal and a non-speech signal, automatically identifies the input signal by extracting the speech characteristic and designing a sorter model, and judges the audio type; if the audio type is noise, the audio signal is removed; if the audio type is speech, the audio signal is remained and processed further. The method meets the real-time requirement and has better reduction noise effect at the same time, can be suitable for the situations with complex communication environments such as manned spaceflight speech communication, construction sites, battlefields and the like, and provides an idea and a method for the noise reduction of signals.

This invention implements a system for automatic recognition of human-assisted drawings, in a plurality of forms, be they hand-drawn on paper, marker board, with a stylus on a computer, made with a mouse, stylus, finger or other instrument on a personal computer, tablet computer, smart telephone or other medium. At the core of the invention is a pattern recognition engine, aimed at recognizing the graphical objects, handwritten text, equations or interconnects in the input image, and interpreting the significance of their relative association. The apparatus offers error correction, vector representation of the input sketch, as intermediate output, along with the recognized patterns, arranged in a hierarchical data structure, ready to be passed on for mining or assessment. The recognized patterns can be associated with mechanical design, electrical circuit design, mathematics, biology, physics, chemistry, computer science, natural sciences, medicine, or any other science- or engineering-based discipline making use of human-assisted drawings.

The present invention belongs to the field of computer science and technology, in particular, it relates to a new characteristic vector weight method for text classification. Said invention provides a new weighting method (TF*IWF*DBV). It is characterized by that in TF*IWF method n-th root of DBV and TF can be introduced, the tests show that the adoption of said new method can raise FI measure by 11.8%, and can fully show its effectiveness.

The invention discloses a method and a device for searching an information object. The information object is information content with semantic structure, comprising a plurality of information properties with semantic correlative relation, wherein the information properties are units of different semantic aspect content of information objects; and the semantic structure is collection of semantic metadata. Target webpage URI can be utilized to query correlative library to obtain semantic structure by establishing semantic structure and network resource URI or correlative library of prefix of network resource URI; webpage information is formatted and indexes are established with the guidance of semantic structure; user request is responded, searching results are provided according to correlation order, and simultaneously layering semantic navigation trees are provided; and refine searching results are provided to a user when the user clicks the node of the semantic navigation tree. The invention belongs to the technical field of computer science, is applicable to the application fields of network information extraction, full text search, semantic search, mobile phone search and the like.

The invention discloses a dual-language coordinated training-based named entity identification method, and belongs to the technical field of natural language processing in computer science. Parallel Chinese and English sentence datasets are considered as two different view of a dataset for dual-language coordinated training, a log-linear model is used for correcting projection marks in a projection process, and named entity dual-language aligned annotation consistency is introduced as a measurement index for mark confidence estimation when the model is used for predicting an unseen case. Compared with the prior art, the method has the advantages that the domain dependence of named entity identification is reduced, the advantages of dual-language identification are fused, the problem of partial identification ambiguity in single-language identification is solved, and the method is particularly suitable for the dual-language named entity synchronous identification of large-scale language materials.

The present invention provides an enablement and optimization system and method for web advertisement placement and pricing. The invention provides an open advertising network that incorporates both the advertisers and content publishers into the advertising decision process to create an open marketplace. Similar to a greedy algorithm in Computer Science, this model pushes the final advertising decision to the web servers that serve the actual content. Rather than an advertising network providing a single ad to show, the content publisher receives a selection of ads. Each publisher in the network then makes the local optimum choice by collecting data on the effectiveness of ads previously shown, and using this data to select which ad will perform the best. Similarly, the advertisers collect data on the visitors they receive from advertising, and use this information to optimize their bidding. With each endpoint selecting a local optimum, the network of advertisers and content publishers as a whole becomes more effective.

The invention relates to a method for automatically extracting character relations from a Chinese text or a text set, and belongs to the technical field of computer science and information extraction. In the method, by means of sentence meaning model characteristics, the relation attribute affiliation is determined; the character relations scattered in the text or the text set are automatically extracted by combining methods such as relation attribute disambiguation, character relation strength calculation and the like; and the character relations are organized by a character relation network, and the character relations comprising character relation attributes and the relation strength are displayed in a character relation graph manner. According to the method, sentence meaning model characteristics are introduced, so that the accuracy of the method for extracting entity relations is improved, and the method for extracting the character relations is enriched. Besides, as the number of texts about a central character in the text set is increased, the method can extract character relations of the central character more accurately and comprehensively, and the application range is wider.

The invention relates to an IP restricted controlled source information capture method based on agent pools, and belongs to the field of data acquisition and processing in the computer science. The method comprises the following steps: maintaining two agent pools A and B, and storing all agents in the agent pool A; testing available agents and initializing the agent pool B; randomly selecting agents in the agent pool B for use in a capture process; triggering an updating operation of the agent pool B by a dynamic maintenance operation of the agent pool B in the capture process; and realizing efficient capture of an IP restricted controlled information source by using the agent pool algorithm. According to the IP restricted controlled source information capture method, the two agent pools are maintained, the available agents are dynamically regulated and controlled in the information capture process to solve the problems of instable agent and low capture efficiency and expand resources for all kinds of data drive experiments.

The invention discloses a method and equipment for labeling a semantic structure of network contents. The semantic structure includes the relation of semantic comment of network contents, attributes of commented information and semantic comment. The method comprises the following steps: recognizing the semantic structure of network contents; editing the semantic structure; issuing and sharing description documents of the semantic structure, and establishing a knowledge library of the semantic structure; and utilizing format command files in web pages generated by the semantic structure. The invention belongs to the technical field of computer science and the branch of semantic network, and is suitable for the application fields such as network information extraction, network knowledge base establishment, semantic network service and the like.

The invention provides a method and a system for recognizing facial feature points through a face photograph. The method comprises the following steps: 1) carrying out face detection to judge whether a face exists in an input image or not, and continuously executing a step 2) if the face is detected; and 2) determining the face feature points of the face in the image. In the step 1), a face image to be determined is matched with a face template, if the face image to be determined is matched with the face template, the face image to be determined is projected into face subspace, and whether the face image to be determined is the face or not is judged through feature sub-face technology. The invention utilizes a computer scientific processing means, can recognize facial feature positions and feature points through a single front face photograph and improves face recognition accuracy.

The invention relates to a terminal-spanning real-time interactive online distributing method of a lantern slide and belongs to the field of a computer science and technology application technology. The method comprises the following steps of: uploading a lantern slide file to a server side from a master control end through an HTTP (Hyper Text Transport Protocol) channel; analyzing the lantern slide file by the server side and analyzing all information in each page of the lantern slide file to a middle format and storing; carrying out interactive distribution on the lantern slide by the master control end to finish playing control and information interaction in which the master control end, the server side and a client are participated; and locally and persistently storing by the master control end and the client. Compared with the traditional method, the real-time online distribution of the lantern slide on each type of intelligent equipment is realized; an autonomous control capability of the client and an interactive communication capability of the master control end and the client are greatly enhanced; and transmission and storage of content and notes are finished in a playing process, so as to get rid of the constraint of a traditional computer and break through the limitation on time and space.

The invention relates to a Chinese sentence meaning structure model automatic labeling method based on CRF ++, and belongs to the technical field of computer science and natural language processing semantic analysis. Firstly, linguistic data in a BFS-CTC Chinese labeling linguistic database are used, a predicate recognition model, a word relation recognition model and a semantic case type recognition model are obtained through training; secondly, recognition on an original sentence is conducted by using the predicate recognition model, the word relation recognition model and the semantic case type recognition model so that predicate information, work relation information and semantic case type information in the sentence are obtained; and finally, a Chinese sentence meaning structure model is obtained according to collocation rules of the predicate, the work relation and the semantic case types. The method is in the field of semantic analysis, more and comprehensive semantic features are provided, so that the foundation of sentence meaning structure model analysis to the sentence by a computer is laid. Meanwhile, a definite possibility is provided for automatic labeling of the BFS-CTC Chinese labeling linguistic database, both study and practical application of the linguistic data are important, and the method plays a great promoting role in augment of the BFS-CTC Chinese labeling linguistic database.

The invention discloses an online news topic detection method and belongs to the field of computer science and technology. The more efficient topic detection method is raised up for web texts with the requirement for topic detection in the internet. A cluster buffer zone is established to initially cluster reached texts of a certain number or within a certain period through an X-means algorithm, a dual-threshold (a topic gathering threshold value and a topic mass center updating threshold value) thought is introduced, topic shift is effectively controlled, and the clustering effect is improved. The effects achieved through the method are superior to those of a classic Single-Pass algorithm at all evaluation indexes, and topics with the topic detection requirement are more accurately identified.

The invention discloses a bilingual largest noun group separating-fusing translation method, and belongs to the technical field of natural language processing in computer science. The BMNCs separating-fusing translation concept is introduced to a statistic machine translation model based on a tree, and sentence translation is converted into translation of all bilingual largest noun groups in a sentence and translation of a sentence framework. According to the bilingual largest noun group separating-fusing translation method, translation difficulty is lowered, translation time is shortened, translation accuracy is improved, and the bilingual largest noun group separating-fusing translation method is particularly suitable for long-sentence translation in instant translation.

The invention belongs to the crossing field of computer science and traffic engineering, and relates to an indoor pedestrian microscopic simulation method based on a cellular automaton. The method comprises following steps of firstly, carrying out gridding on an indoor pedestrian region, thus obtaining a cellular space corresponding to a cellular automaton model; dividing the cellular space into some convex polygon regions; secondly, judging whether there are pedestrians in each region; stating the temporary destination of each pedestrian; calculating the transfer probability of each pedestrian; simulating movements; observing whether conflicts appear or not; finally, solving the conflicts; and updating the practical position of each pedestrian until the pedestrians arrive at the corresponding destinations. According to the method, the simulation model can correspondingly respond to complex walk environments and not merely distinguish the pedestrians and barriers; moreover, the self-organization phenomena displayed in the regional pedestrian flow indoor evacuation process are simulated; the model simulation effect is effectively improved; and the method is more suitable for simulating and analyzing movement evolution of the pedestrian flow in the indoor complex environments.

The invention provides a hybrid spatial indexing mechanism for processing geographic text Skyline query, wherein geographic text Skyline query refers to performing Skyline query on a geographic text information dataset. Data points in the geographic text dataset include geographic position information and keyword text information. The Inverted-Merged R*-Tree (IMR*-T) integrates the R*tree and theInverted File thought, and the invention belongs to the query indexing field in computer science. The invention focuses on solution of the problem of storing the geographic text dataset and performingSkyline query on the geographic text dataset, can improve the Skyline query efficiency on the premise of ensuring reasonable storage. The invention constructs a multi-branch tree according to data point spatial position distribution by means of an R* tree construction strategy, and can construct an Inverted File for leaf nodes of the tree. To improve the clipping efficiency of the dataset, tree nodes store boundary frame information. The invention is widely suitable for relevant application scenarios of geographic text Skyline query.

The invention relates to a Chinese text term extracting method utilizing quadratic mutual information and belongs to a computer science and natural language processing technology. The method comprises the steps that firstly, core words are expanded into multiple candidate terms by combining the quadratic mutual information, word frequency, word length and word characteristics based on previous and later core word expansion, then redundant terms centralized in a candidate set are removed and the terms are graded and sorted according to the nest relation of the terms, the word length and the word frequency characteristics. The method comprehensively considers the language rules and statistical characteristics of the terms, and the term extraction accuracy is improved.

The invention discloses a data excavation method of form orientation, and belongs to the technical field of computer science. The steps of the method are as follows: Establish format module for data source firstly, and then orientate the data form according to the format module of data source, and at last extract data according to contents of data form unit. Search engines can make use of the technique to make deep analysis and treatment of webpage information, extract usable format information, and arrange them into relationship database form or structuralized data structure so as to guide users to directly obtain needed data rather than to obtain data through related inquiry, thus greatly shortening the searching time of users and enhancing the utilization value of data. Search engine created through the invention can accurately extract data concerned by users to extremely enhance the experience of search engine users.

The invention relates to a word segmentation phonetic transcription and ligature writing method and device based on an SC grammar and belongs to the technical field of computer translation in computer science. Firstly, based on a word segmentation ambiguity rule of the SC grammar, an ambiguity segmentation rule library is built by means of abutment constraint conditions in natural language, and illegal segmentation is eliminated so that the word segmentation precision can be improved; secondly, based on a word segmentation ligature writing rule library of the SC grammar and a ligature writing corpora statistical library, the ligature writing corpora statistical library is used for performing ligature writing on ligature writing knowledge which cannot be presented as rules; finally, based on a dictionary library of the SC grammar, a dictionary is used for performing maximum matching to perform word segmentation, the word segmentation ambiguity rule is called for fields where ambiguity happens so that a correct segmentation result can be acquired, and the context of a word is analyzed so that correct part-of-speech tagging and phonetic transcription can be acquired. Compared with the prior art, word segmentation accuracy is improved, and the word segmentation ambiguity rule library, a combined ambiguity word library, the ligature writing rule library, the dictionary library and the ligature writing corpora statistical library are easy to expand and maintain.

The invention belongs to the field of computer science and technology, and relates to a web page cleaning method based on a theme. The web page cleaning method comprises the steps that the field of the theme is determined; a sample page is collected, the web page structure of the sample page is analyzed, and a page element tree structure is determined; an extracted content label boundary is determined, regular expression matching rules are defined according to the extracted content label boundary, and an extraction rule template is formed by a plurality of extraction rules; a page to be cleaned is cleaned and preprocessed to remove noisy contents easy to recognize; the page is cleaned through the generated extraction rule template to acquire a cleaned page. Because useful information is directly extracted from the page, the web page cleaning method greatly reduces page cleaning complexity and difficulties. Because web page labels are complex, diversified and locally fixed, required labels can be directly matched and extracted according to a regular expression to reserve useful information and automatically remove useless information, and accordingly the purpose of page cleaning can be effectively achieved.

The invention discloses a multi-emotion tendency mapping method and emotion sentential form analysis method applied to a search engine, belonging to an interdiscipline subject of computer science and technology and linguistics. The multi-emotion tendency mapping method comprises an emotion participle module, an emotional word bank and a multi-tendency mapping module; and the emotion sentential form analysis method comprises a sentential form structure analysis module and a sentential form structure template library. By combining with an Emotion Wheel method of psychology and an evaluation system method of the linguistics, the multi-emotion tendency mapping method and emotion sentential form analysis method applied to the search engine, disclosed by the invention, has advantages of quantitatively distinguishing emotion tendency of emotional words, quantitatively calculating emotion tendency of sentences and discovering hidden emotion tendency of the sentences.

The invention discloses a translation form self-acquiring method and device in the computer science technique domain, which comprises the following steps: a) predisposing the language material; b) extracting Chinese word graph and English word graph separately; c) inducting the grammar of Chinese part and English part; d) aligning the extracted phrase structure; e) comparing the result of different similarity function to select optimum similarity function; f) disposing the aligned result to gain the needed mechanic translation model. The device contains Chinese word classification device, Chinese grammar induction device, English grammar induction device, phrase structure alignment device and after-treatment device.