Abstract - arXiv
A Convolutional Neural Network for Modelling SentencesNal KalchbrennerEdward Grefenstette{ , , of Computer ScienceUniversity of OxfordPhil BlunsomAbstractThe ability to accurately represent sen-tences is central to language understand-ing. We describe a convolutional architec-ture dubbed the Dynamic ConvolutionalNeural Network (DCNN) that we adoptfor the semantic modelling of network uses Dynamick-Max Pool-ing, a global pooling operation over lin-ear sequences. The network handles inputsentences of varying length and inducesa feature graph over the sentence that iscapable of explicitly capturing short andlong-range relations. The network doesnot rely on a parse tree and is easily ap-plicable to any test theDCNN in four experiments: small scalebinary and multi-class sentiment predic-tion, six-way question classification andTwitter sentiment prediction by distant su-pervision. The network achieves excellentperformance in the first three tasks and agreater than25%error reduction in the lasttask with respect to the strongest IntroductionThe aim of a sentence model is to analyse andrepresent the semantic content of a sentence forpurposes of classification or generation.}
A central class of models are those based on neural networks. These range from basic neu-ral bag-of-words or bag-of-n-grams models to the more structured recursive neural networks and
Download Abstract - arXiv
Information
Domain:
Source:
Link to this page:
Please notify us if you found a problem with this document:
Related search queries
Financial Modelling, Methods, Morphological Analysis, OECD, National Calculation Methodology (NCM) modelling guide, Modelling, ARIS Standards and Conventions Manual, Modelling the Enterprise Data Architecture, Modelling smoke and gas ingress into offshore temporary, Modelling smoke and gas ingress, Seismic Forward Full-Wave Modelling in Tesseral