MVLSC Home · Issue Contents · Forthcoming Papers
Online Text Retrieval Method Based on Convolution Neural Network
Hong Tu
Text retrieval is a vital part of information retrieval. In order to improve the accuracy of text retrieval, a text online retrieval method based on convolution neural network is proposed. The convolution neural network model is constructed and designed layer by layer. The convolution neural network is used to extract the specific features of text and reduce the dimension. According to the reduced dimension of the text features, text retrieval is realized on line by calculating the semantic similarity of the text. The experimental results show that the designed convolutional neural network has high accuracy, integrity and short retrieval time. The completeness of this method decreases slowly, from 98% to 91%, and the text content is well preserved. This method can effectively retrieve 3000 text information within 14 MS, which is more efficient and practical than the other two methods. It has certain practical application value.
Keywords: Convolution neural network, text retrieval, online retrieval