litlas
ログインアカウント作成

From Word Embeddings To Document Distances

著者: Matt J. Kusner, Yu Sun, Nicholas I. Kolkin, Kilian Q. Weinberger - International Conference on Machine Learning, ICML 2015 被引用: 2,277

We present the Word Mover's Distance (WMD), a novel distance function between text documents. Our work is based on recent results in word embeddings that learn semantically meaningful representations for words from local cooccurrences in sentences. The WMD distance measures the dissimilarity between two text documents as the minimum amount of distance that the embedded words of one document need to travel to reach the embedded words of another document. We show that this distance metric can be cast as an instance of the Earth Mover's Distance, a well studied transportation problem for which several highly efficient solvers have been developed. Our metric has no hyperparameters and is straight-forward to implement. Further, we demonstrate on eight real world document classification data sets, in comparison with seven state-of-the-art baselines, that the WMD metric leads to unprecedented low k-nearest neighbor document classification error rates.

Nearest neighbor patternclassificationNearest neighbor pattern classificationIndexing by LatentSemantic AnalysisIndexing by Latent Semantic AnalysisA Metric forDistributions with…A Metric for Distributions with Applications to Image DatabasesLatent DirichletAllocationLatent Dirichlet AllocationPractical solutions tothe problem of diagonal…Practical solutions to the problem of diagonal dominance in kernel document clusteringFast and robust EarthMover's DistancesFast and robust Earth Mover's DistancesWord Representations: ASimple and General…Word Representations: A Simple and General Method for Semi-Supervised LearningDomain Adaptation forLarge-Scale Sentiment…Domain Adaptation for Large-Scale Sentiment Classification: A Deep Learning ApproachMarginalized DenoisingAutoencoders for Domain…Marginalized Denoising Autoencoders for Domain AdaptationDistributedRepresentations of Word…Distributed Representations of Words and Phrases and their CompositionalityLinguistic Regularitiesin Continuous Space Wor…Linguistic Regularities in Continuous Space Word RepresentationsDistributedRepresentations of…Distributed Representations of Sentences and DocumentsFrom word embeddings todocument similarities…From word embeddings to document similarities for improved information retrieval in software engineeringGenerative TopicEmbedding: a Continuous…Generative Topic Embedding: a Continuous Representation of DocumentsA BidirectionalHierarchical Skip-Gram…A Bidirectional Hierarchical Skip-Gram model for text topic embeddingNeural informationretrieval: at the end o…Neural information retrieval: at the end of the early yearsLearningbag-of-embedded-words…Learning bag-of-embedded-words representations for textual information retrievalConcept Mover'sDistance: measuring…Concept Mover's Distance: measuring concept engagement via word embeddings in textsDetection of medicaltext semantic similarit…Detection of medical text semantic similarity based on convolutional neural networkUnsupervised Alignmentof Embeddings with…Unsupervised Alignment of Embeddings with Wasserstein ProcrustesMessage PassingAttention Networks for…Message Passing Attention Networks for Document UnderstandingDeepEMD: Few-Shot ImageClassification With…DeepEMD: Few-Shot Image Classification With Differentiable Earth Mover's Distance and Structured ClassifiersText classification withword embedding…Text classification with word embedding regularization and soft similarity measureUnsupervised DocumentEmbedding via…Unsupervised Document Embedding via Contrastive AugmentationFrom Word Embeddings ToDocument DistancesFrom Word Embeddings To Document Distances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。