377 results about "Semantic space" patented technology

A term-by-document matrix is compiled from a corpus of documents representative of a particular subject matter that represents the frequency of occurrence of each term per document. A weighted term dictionary is created using a global weighting algorithm and then applied to the term-by-document matrix forming a weighted term-by-document matrix. A term vector matrix and a singular value concept matrix are computed by singular value decomposition of the weighted term-document index. The k largest singular concept values are kept and all others are set to zero thereby reducing to the concept dimensions in the term vector matrix and a singular value concept matrix. The reduced term vector matrix, reduced singular value concept matrix and weighted term-document dictionary can be used to project pseudo-document vectors representing documents not appearing in the original document corpus in a representative semantic space. The similarities of those documents can be ascertained from the position of their respective pseudo-document vectors in the representative semantic space.

A method for sentiment classification of a text document using high-order n-grams utilizes a multilevel embedding strategy to project n-grams into a low-dimensional latent semantic space where the projection parameters are trained in a supervised fashion together with the sentiment classification task. Using, for example, a deep convolutional neural network, the semantic embedding of n-grams, the bag-of-occurrence representation of text from n-grams, and the classification function from each review to the sentiment class are learned jointly in one unified discriminative framework.

In various embodiments, a semantic space associated with a corpus of electronically stored information (ESI) may be created and used for concept searches. Documents (and any other objects in the ESI, in general) may be represented as vectors in the semantic space. Vectors may correspond to identifiers, such as, for example, indexed terms. The semantic space for a corpus of ESI can be used in information filtering, information retrieval, indexing, and relevancy rankings.

The invention relates to a specific modal semantic space modeling-based cross-modal similarity learning method. The method comprises the following steps of: 1, establishing a cross-modal database which comprises data of multiple modal types, and dividing data in the database into a training set, a test set and a validation set; 2, aiming at each modal type in the cross-modal database, constructinga semantic space aiming at the specific modal, and projecting data of the other modal types to the semantic space so as to obtain a cross-modal similarity aiming at the specific modal; 3, fusing cross-modal similarities, aiming at specific modals, obtained from different modal semantic spaces to obtain a final cross-modal similarity; and 4, calculating a similarity between a query example and a query target by taking any modal type in the test set as a query modal and taking another modal type as a target modal, and obtaining a related result list of data of the target modal according to thesimilarity. According to the method, the correctness of cross-modal retrieval can be improved.

The invention belongs to the technical field of computer software and discloses a multi-language-pair neural network machine translation method and system. A plurality of bilingual parallel corpora ofa same language system are utilized and mapped to a same high-dimensional vector space after byte pair encoding, so that multiple languages share a same semantic space, the size of a word list is reduced, model parameters are reduced, and convergence of a model is accelerated. Words of a same language family are in the same vector space, more information can be learned mutually, the information which can not be learned through only certain bilingual parallel corpora can be learnt, and the quality of word vectors is improved. The machine translation system can be used for translation in the language direction without direct bilingual parallel corpora, and the translation quality in the scarce parallel corpus translation direction is greatly improved through mutual information learning. Meanwhile, the same model is used for translation for the translation direction low in utilization rate, occupation of a server is reduced, and the utilization rate of the server is increased.

The embodiment of the invention discloses a vision localization method and device. In the method, according to a preset camera device, a preset scene is shot to obtain image data; on the one hand, according to first feature points of an image and spatial feature points in a semantic space map, vector information matching is conducted, and a first matching point set is obtained; on the other hand,second feature points, containing second semantic information, in the image data are extracted, the second feature points and the spatial feature points are subjected to semantic information matching,and a second matching point set is obtained; according to the first matching point set and the second matching point set, the target pose information and the target position information of the presetcamera device are obtained. In the method, the texture information and the semantic information of the image are utilized at the same time to extract the feature points, more feature points which aresuccessfully matched with the spatial feature points in the semantic space map can be found in the image data, a more reliable data basis is provided for calculating the target pose information and the target position information of the preset camera device, and the localization precision is higher.

A computer system for parsing bulk message information into intelligent question retrieval models receives text-based data associated with a particular user. The system encodes the word into a context-dependent vector, wherein the context-dependent vector indicates the meaning of the word across a semantic space. The system also identifies within a context-independent database a context-independent vector that is associated with the word. Further, the system generates an objective output by combining the context-dependent vector and the context-independent vector. Further still, the system generates a sentence encoding representation by processing at least a portion of the text-based data through a high-level feature embedded convolutional semantic model to generate numerical representations of questions and answers within the text-based dataset. The sentence encoding representation is generated at least in part based upon the objective output.

The invention discloses a commodity target word oriented emotional tendency analysis method, which belongs to the field of the analysis processing of online shopping commodity reviews. The method comprises the following four steps that: 1: corpus preprocessing: carrying out word segmentation on a dataset, and converting a category label into a vector form according to a category number; 2: word vector training: training review data subjected to the word segmentation through a CBOW (Continuous Bag-of-Words Model) to obtain a word vector; 3: adopting a neural network structure, and using an LSTM(Long Short Term Memory) network model structure to enable the network to pay attention to whole-sentence contents; and 4: review sentence emotion classification: taking the output of the neural network as the input of a Softmax function to obtain a final result. By use of the method, semantic description in a semantic space is more accurate, the data is trained through the neural network so as to optimize the weight and the offset parameter in the neural network, parameters trained after continuous iteration make a loss value minimum, at the time, the trained parameters are used for traininga test set, and therefore, higher accuracy can be obtained.

The invention discloses a knowledge graph entity semantic space embedding method based on graph second-order similarity, and the method comprises the steps: (1) inputting a knowledge graph data set and a maximum number of iterations; (2) calculating first-order and second-order similarity vector representations through first-order and second-order similarity feature embedding processing by considering a relation between entities through a graph attention mechanism to obtain first-order and second-order similarity semantic space embedding representations; (3) carrying out weighted summation onthe final first-order similarity vector and the final second-order similarity vector of the entity to obtain a final vector representation of the entity, inputting a translation model to calculate a loss value to obtain a graph attention network and a graph neural network residual, and iterating the network model; And (4) performing link prediction and classification test on the network model. According to the method, the relation between entities is mined by using a graph attention mechanism for the first time, and patents have a relatively good effect in the application fields of link prediction, classification and the like of the knowledge graph.

The invention discloses a DSAH (discrete semantic alignment hashing) method based on semantic alignment for cross-modal retrieval. In the training process, a heterogeneous gap is reduced by the aid ofimage attributes and modal alignment semantic information. In order to reduce memory overhead and training time, a latent semantic space is learned by synergistic filtering, and the internal relationbetween a hash code and a label is directly built. Finally, in order to decrease quantization errors, a discrete optimization method is proposed to obtain a hash function with better performances. Inthe retrieval process, samples in a testing set are mapped to a binary space by the hash function, the Hamming distance between a binary code of a query sample and a heterogeneous sample to be retrieved is calculated, and front ranked samples are returned according to the sequence from small to large. Experimental results of two representative multi-modal data sets prove superior performances ofDSAH.

The invention relates to the artificial intelligence technology field and particularly relates to an image text matching model training method, a bidirectional search method and a related device. Thetraining method comprises steps that global and local representations of an image sample and a text sample are extracted, a pre-constructed matching model is trained, the global and local representations of the image sample and the text sample are respectively mapped to the specified semantic space by the matching model, and similarity of the global representation and similarity of the local representation are calculated; according to the default weight of the similarity of the global representation and the default weight of the similarity of the local representation, the image and text matching degree is determined through employing a weighted summation mode. The training method is advantaged in that the acquired matching degree is based on considering both detail characteristics of an image and global characteristics, and the acquired matching degree is more accurate and comprehensive.

The invention discloses a zero sample classification method based on an extreme learning machine, and the method is used for image classification. The method comprises the following steps: extracting the visual features of a training image at a training state, and extracting the training semantic features corresponding to the visual features of the training image; randomly generating a first input weight and a first threshold value for L junctions, and calculating a first output matrix of a hidden layer through employing a hidden layer mapping function; calculating the output weight of a network through the training semantic features and the first output matrix of the hidden layer; extracting the visual features of a test sample at a test stage, randomly generating a second input weight and a second threshold value for L junctions, and calculating a second output matrix of the hidden layer through employing the hidden layer mapping function; calculating an embedded vector, correspondingly located in a semantic space, of the second output matrix through the output weight, and judging the type of the test sample according to the similarity of the embedded vector with the semantic features in a semantic feature space. The method reduces the training time, and improves the classification speed of the image.

The invention discloses a video semantic representation method and system based on a multi-mode fusion mechanism and a medium. Feature extraction: extracting visual features, voice features, motion features, text features and domain features of a video; Feature fusion: performing feature fusion on the extracted visual, voice, motion and text features and domain features through the constructed multi-level hidden Dirichlet distribution topic model; And feature mapping: mapping the fused features to a high-level semantic space to obtain a fused feature representation sequence. The model utilizesthe unique advantages of the theme model in the semantic analysis field, and the video representation mode obtained through model training on the basis of the model has ideal distinction in the semantic space.

The invention relates to a method of Tibetan language webpage text classification based on semanteme. The method of the Tibetan language webpage text classification based on the semanteme comprises the following steps: firstly, extracting text messages which are used for charactering a webpage from the Tibetan language webpage, then carrying out word classification processing on the text messages, and expressing words obtained by means of the word classification processing as a word vector space, secondly, mapping the words expressed through the word vector space to a concept of a semanteme space according to a preset Tibetan language classification body to obtain a semanteme space of a text to be classified, and at last adopting a classification algorithm to classify the semanteme space of the text to be classified according to a preset semanteme space of a training sample set. The method of the Tibetan language webpage text classification based on the semanteme carries out preprocessing on the webpage and adopting a KNN classification algorithm of the text similarity of a weighting semantic net to achieve real-time and high-efficiency classification of the Tibetan language webpage.

The invention provides a method, a mobile terminal and a server for carrying out the intelligent search at a mobile terminal. The method comprises the following steps: acquiring the lexical space of network resources, and carrying out the conceptual clustering on the information to be searched in a search request, to form a conceptual space; retrieving lexes identical with the content semantics of the conceptual space from the content of the lexical space, to form a semantic space; analyzing user behavior to generate a user behavior space; generating an associated business library according to the mapping relation between the feasible business recorded in a business space and provided by operators and the content of the user behavior space; establishing the association relation between the content in the associated business library and the lexes in the semantic space; and selecting the search results from the network resources. By introducing the analysis on the user behavior, the invention improves the accuracy of search, so the search can be more convenient, direct and effective, and the information exchange via the network can be more user-friendly; and by analyzing the user behavior, the favorite advertising information and the most acceptable advertising information of the user can be recommended at the mobile terminal.

The invention relates to a document semantic representation method based on thematic word class similarities and a text classification method and device. The document semantic representation method based on thematic word class similarities comprises the steps that (1) a word vector model is used to train a corpus, and word vectors are obtained; (2) clustering is performed on the word vectors in asemantic space; and (3) a WMD algorithm is used to calculate the distance between a to-be-represented document and each category obtained through clustering, and the obtained distance is used as semantic representation of the to-be-represented document. Then document classification is realized by calculating the similarities among semantic representation vectors of the document. Through the document semantic representation method, according to semantic information, word frequency and other information of texts, transfer cost between text words and clustering sets is calculated through the WMDmodel, each text is represented with one low-dimensional dense vector containing the semantic information, text information can be better represented, the classification task is high in accuracy, andthe method can be applied to natural language processing tasks such as information retrieval and text classification.

The invention relates to a semantic segmentation network integrating multi-scale feature space and semantic space, comprising: determining a backbone network of a network coding end; VGG16 as a backbone part of the network coding end, removing a fifth pooling layer and reducing one downsampling; Designing the fusion module of feature space and semantic space of network decoder; achieving High resolution and high precision semantic segmentation by using the fusion module of multi-scale feature space and semantic space. Output Semantic Segmentation Result.

The invention discloses a latent semantic image retrieval method of region-oriented multi-feature integration and multi-level feedback. It uses result list returned by the initial keyword search, extracting a variety of region-oriented images characteristics, constructing attribute-image matrix, using latent semantic indexing algorithm to get the semantic space of image sets and semantic features of each image, and then using similar images by users feedback to construct or update image query vector, searching again the semantic space, calculating image semantics features and images inquiries vector similarity, getting outcome sets by descending order, and repeatable retrieval. The invention takes full advantage of image content information, making up for the deficiencies of the keyword search, and through the region-oriented multi-feature integration, enhances image content information from the bottom physical layer to the object layer, then further enhances to the semantic layer by HCI feedback, thereby reducing the gap between the image bottom features and high-level semantic, and allowing Web image retrieval to get higher retrieval accuracy.

The invention discloses a system and a method for automatically scoring college English compositions independent of manual pre-scoring. The system comprises a training composition characteristic word-composition matrix generating module, a training composition semantics space constructing module, an automatic scoring module of a composition to be scored, wherein a processing result, namely, a ''characteristic word-composition matrix'', of a ''construction characteristic word-composition matrix'' processing unit in the training composition characteristic word-composition matrix generating module is input into a ''construction training composition set semantics space'' processing unit in the training composition semantics space constructing module; and a processing result, namely, a ''full score composition set semantics space'', of a ''mapping full score composition set characteristic word frequency vector quantity to a training composition set semantics space'' processing unit in the training composition semantics space constructing module is input into a ''score calculating of the composition to be scored'' processing unit in the automatic scoring module of the composition to be scored. With the adoption of the system, workload of college English teachers in reading student English compositions is reduced greatly, and moreover, reliability and efficiency on scoring are high.