litlas
ログインアカウント作成

Distributed Representations of Words and Phrases and their Compositionality

著者: Tomás Mikolov, Ilya Sutskever, Kai Chen, Gregory S. Corrado, Jeffrey Dean - Conference on Neural Information Processing Systems, NeurIPS 2013 被引用: 35,454

The recently introduced continuous Skip-gram model is an efficient method for learning high-quality distributed vector representations that capture a large number of precise syntactic and semantic word relationships. In this paper we present several extensions that improve both the quality of the vectors and the training speed. By subsampling of the frequent words we obtain significant speedup and also learn more regular word representations. We also describe a simple alternative to the hierarchical softmax called negative sampling. An inherent limitation of word representations is their indifference to word order and their inability to represent idiomatic phrases. For example, the meanings of "Canada" and "Air" cannot be easily combined to obtain "Air Canada". Motivated by this example, we present a simple method for finding phrases in text, and show that learning good vector representations for millions of phrases is possible.

Learning Representationsby Back-Propagating…Learning Representations by Back-Propagating ErrorsHierarchicalProbabilistic Neural…Hierarchical Probabilistic Neural Network Language ModelA unified architecturefor natural language…A unified architecture for natural language processing: deep neural networks with multitask learningWord Representations: ASimple and General…Word Representations: A Simple and General Method for Semi-Supervised LearningStrategies for traininglarge scale neural…Strategies for training large scale neural network language modelsDomain Adaptation forLarge-Scale Sentiment…Domain Adaptation for Large-Scale Sentiment Classification: A Deep Learning ApproachExtensions of recurrentneural network language…Extensions of recurrent neural network language modelParsing Natural Scenesand Natural Language…Parsing Natural Scenes and Natural Language with Recursive Neural NetworksA fast and simplealgorithm for training…A fast and simple algorithm for training neural probabilistic language modelsSemanticCompositionality throug…Semantic Compositionality through Recursive Matrix-Vector SpacesEfficient Estimation ofWord Representations in…Efficient Estimation of Word Representations in Vector SpaceLinguistic Regularitiesin Continuous Space Wor…Linguistic Regularities in Continuous Space Word RepresentationsNeural Architectures forNamed Entity RecognitionNeural Architectures for Named Entity RecognitionDetecting Depressionusing Vocal, Facial and…Detecting Depression using Vocal, Facial and Semantic Communication CuesDeep Learning approachfor sentiment analysis…Deep Learning approach for sentiment analysis of short textsImproving NegativeSampling for Word…Improving Negative Sampling for Word Representation using Self-embedded FeaturesTransductive UnbiasedEmbedding for Zero-Shot…Transductive Unbiased Embedding for Zero-Shot LearningVector representationsof text data in deep…Vector representations of text data in deep learningPublicly AvailableClinicalPublicly Available ClinicalAsm2Vec: Boosting StaticRepresentation…Asm2Vec: Boosting Static Representation Robustness for Binary Clone Search against Code Obfuscation and Compiler OptimizationNamed Entity RecognitionUsing BERT BiLSTM CRF…Named Entity Recognition Using BERT BiLSTM CRF for Chinese Electronic Health RecordsMachine learning inmaterials science: From…Machine learning in materials science: From explainable predictions to autonomous designTwitter-Based DisasterResponse Using Recurren…Twitter-Based Disaster Response Using Recurrent NetsA deep learning-basedresource usage…A deep learning-based resource usage prediction model for resource provisioning in an autonomic cloud computing environmentDistributedRepresentations of Word…Distributed Representations of Words and Phrases and their Compositionality過去の参考文献中心の論文この論文を引用する論文古い新しい

ノードをクリックするとフォーカスを固定、空白をクリックすると本論文に戻ります。ホバーで一時的にプレビューできます。各ノードのページはタイトルから開けます。