To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
An Efficient Next Word Prediction for Accurate Information using Deep Learning Algorithms |
|
Author |
B.Tarakeswara Rao, E.Ramesh, A. Srinagesh K.Srınıvasa Rao, N.Kıran Kumar, ,P. Siva Prasad, B.Naga Mallikarjuna, K.Arun |
| Citation |
Vol. 22 No. 6 pp. 665-669 |
|
Abstract |
Natural language processing and language models define subsequent phrase predictions. To bet, the following matching sentences are used in search engines, sentence or text content processing, and documentation applications. The most likely phrase is a high-value match for that sentence. In this task, subsequent phrase predictions are performed using the deep learning version. First, we preprocessed the text content, normalized the text content, and implemented four specific deep learning classifiers to experiment and check statistics for expecting subsequent words. Canonical Neural Network (CNN), Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), and Bidirectional Long Short-Term Memory (BiLSTM). Of these deep algorithms, CNN when implemented contributed a high loss and much lower accuracy, and Bidirectional LSTMs resulted and were noted with high accuracy and low loss. These classifiers are run sequentially and comparisons are primarily based on loss discounts and accuracy characteristics. The results obtained show that the CNN's loss discount and accuracy were the worst and BiLSTM achieved the highest quality. |
|
Keywords |
Natural Language Processing, Deep Learning, Prediction, CNN, LSTM, BiLSTM |
|
URL |
