Optimization of Conditional Random Field Model Based on Circular Neural Network

Yiming Liu

《World Scientific Research Journal》 5卷7期 (2019/07) Pp. 32-38

https://doi.org/10.6911/WSRJ.201907_5(7).0006

摘要

The rapid development of Internet technology drives the exponential growth of online data, marking the arrival of the era of big data, and also means that information extraction technology will assume more important tasks. People need to accurately and quickly obtain target information from a large amount of data, and further improve the utilization rate of information. However, the existing extraction techniques often have problems such as limited application scope and large amount of work required for manual operation of corpus part‐of‐speech tagging. Analyzed in this paper using hidden markov model (HMM) training based information extraction method of the problems and the insufficiency, with the help of a conditional random field (CRF) principle on the processing characteristics of the knowledge representation, conditional random field model was put forward, at the same time, combining the cycle neural network (RNN) was carried out on the input variable length data is divided into N parts such as input to the neural network, so as to realize the neural network of variable length of input data processing. Through the experimental evaluation, under the condition of text classification, a better result is obtained by comparing the traditional hidden markov and bayesian network models.

關鍵字

Conditional random field ； circulatory neural network ； informatio variable length text classification

參考文獻

Ketkar N. Introduction to PyTorch[J]. 2017.

Hochreiter S, Schmidhuber J. Long short-term memory. Neural Computation,1997,9(8): 1735-1780.

Hochreiter S, Schmidhuber J. Long short-term memory[J]. Neural computation, 1997, 9(8): 1735-1780.

Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[C]. International Conference on Neural Information Processing Systems. Curran Associates Inc. 2012:1097-1105.

Yang J, Zhang D, Frangi A F, et al. Two-dimensional PCA: a new approach to appearance-based face representation and recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2004, 26(1):131-137.