Sentiment Analysis: An ERNIE-BiLSTM Approach to Bullet Screen Comments.

Area of Science:

• Affective computing
• Natural Language Processing (NLP)
• Text Mining

Background:

• Bullet screen comments are short, ambiguous texts posing NLP challenges.
• Existing research on bullet screen comments lacks NLP focus.
• Understanding sentiments in bullet screen comments is crucial for analyzing user interaction.

Purpose of the Study:

• To address the challenge of analyzing sentiments in bullet screen comments.
• To propose an effective and accurate method for sentiment analysis in this context.
• To investigate the application of NLP techniques for understanding user emotions in bullet screen videos.

Main Methods:

• Proposed an ERNIE-BiLSTM approach for sentiment analysis.
• Applied the method to analyze sentiments expressed in bullet screen comments.
• Evaluated the approach against existing sentiment analysis methods.

Main Results:

• The ERNIE-BiLSTM approach demonstrated higher accuracy, precision, recall, and F1-score.
• The proposed method effectively handles the unique characteristics of bullet screen comments.
• Achieved superior performance in sentiment analysis compared to other techniques.

Conclusions:

• The ERNIE-BiLSTM approach offers an effective and innovative solution for bullet screen comment sentiment analysis.
• This study advances the application of NLP in analyzing user-generated content.
• The findings provide valuable insights into understanding user sentiments in interactive video platforms.