Role of Natural Language Processing in Document Understanding and Semantic Analysis: A Chinese Perspective

Abstract

With the recent advancements in Natural Language Processing (NLP), there is a growing need to enhance the effectiveness and accuracy of document understanding and sentiment analysis particularly for Chinese text as it presents unique challenges of linguistics. To conduct this study, a hybrid approach was implemented which combined CNN and BiLSTM models with an ensemble voting mechanism for sentiment analysis on Chinese text. This method attained an accuracy of 97% and outperformed other techniques such as Text_CNN and AttentionBiLSTM with significant improvements in F1 score and recall. Results demonstrated a superior performance achieving 97% accuracy, along with a 95% F1 score and 97% recall. The present study extends the growing body of literature by underscoring the effectiveness of integrating BiLSTM and CNN models in sentiment analysis within the context of Chinese linguistics. It showcased enhanced document comprehension and capabilities of semantic analysis. Practically, this study provides a robust framework for leveraging BiLSTM and CNN models in the sentiment analysis of real-world. It offers significant boost in adequacy and reliability for processing Chinese text. The research limitations and future research indications have also been addressed in the study