Deep Neural Network-Based Decision Network

专利名称：Deep Neural Network-Based Decision

Network

发明人：Alexander Rosenberg Johansen,Bryan

McCann,James Bradbury,Richard Socher

申请号：US15853570申请日：20171222

公开号：US20180268298A1公开日：20180920

专利附图：

摘要：The technology disclosed proposes using a combination of computationallycheap, less-accurate bag of words (BoW) model and computationally expensive, more-

accurate long short-term memory (LSTM) model to perform natural processing taskssuch as sentiment analysis. The use of cheap, less-accurate BoW model is referred toherein as “skimming”. The use of expensive, more-accurate LSTM model is referred toherein as “reading”. The technology disclosed presents a probability-based guider(PBG). PBG combines the use of BoW model and the LSTM model. PBG uses a probabilitythresholding strategy to determine, based on the results of the BoW model, whether toinvoke the LSTM model for reliably classifying a sentence as positive or negative. Thetechnology disclosed also presents a deep neural network-based decision network(DDN) that is trained to learn the relationship between the BoW model and the LSTMmodel and to invoke only one of the two models.

申请人：salesforce.com, inc.

地址：San Francisco CA US

国籍：US

更多信息请下载全文后查看