<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>In today’s highly networked and information era, how to combine artificial intelligence technology with the field of film and television drama has become a key concern of the current academic community. Based on such problems, this paper adopts the n-gram model and DIFCH algorithm, respectively, to perform vector representation and feature extraction on the text of video drama lines and then puts the extracted sentiment features as inputs into the Bi-LSTM+Attention model for training and classification, and finally completes the construction of a sentiment analysis model based on improved LSTM. The constructed sentiment analysis model is analyzed by combining the sentiment dictionary of the stored data. The results show that the accuracy of the improved LSTM model with category extraction increases with the number of training sessions, and the overall accuracy exceeds 85%. In addition, the attention model makes the BiLSTM model improve 0.0186 and 0.0162 in classification accuracy and AUC value on average, indicating that the attention model can effectively improve the model to extract the text sequence features, which contributes to the performance of the sentiment recognition of film, television and theater lines. Finally, the optimization strategy of line emotional expression of actors in film and television drama is proposed from the aspects of scene and rhythm in the light of the current problem of insufficient ability and skill in line emotion expression of actors in film and television drama.</p>
</abstract>

In today’s highly networked and information era, how to combine artificial intelligence technology with the field of film and television drama has become a key concern of the current academic community. Based on such problems, this paper adopts the n-gram model and DIFCH algorithm, respectively, to perform vector representation and feature extraction on the text of video drama lines and then puts the extracted sentiment features as inputs into the Bi-LSTM+Attention model for training and classification, and finally completes the construction of a sentiment analysis model based on improved LSTM. The constructed sentiment analysis model is analyzed by combining the sentiment dictionary of the stored data. The results show that the accuracy of the improved LSTM model with category extraction increases with the number of training sessions, and the overall accuracy exceeds 85%. In addition, the attention model makes the BiLSTM model improve 0.0186 and 0.0162 in classification accuracy and AUC value on average, indicating that the attention model can effectively improve the model to extract the text sequence features, which contributes to the performance of the sentiment recognition of film, television and theater lines. Finally, the optimization strategy of line emotional expression of actors in film and television drama is proposed from the aspects of scene and rhythm in the light of the current problem of insufficient ability and skill in line emotion expression of actors in film and television drama.

Emotion Algorithm Analysis and Expression Optimization of Film and Television Drama Lines

MEDIA COLLEGE OF XINYANG NORMAL UNIVERSITY

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

In today’s highly networked and information era, how to combine artificial intelligence technology with the field of film and television drama has become a key...

Emotion Algorithm Analysis and Expression Optimization of Film and Television Drama Lines

Publié en ligne: 05 août 2024

Pages: -

Reçu: 17 avr. 2024

Accepté: 10 juil. 2024

DOI: https://doi.org/10.2478/amns-2024-1930

Mots clés
Attention model, Bi-LSTM, Sentiment analysis model, Feature extraction, Film and TV drama lines

© 2024 Yuanfang Jiang, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Emotion Algorithm Analysis and Expression Optimization of Film and Television Drama Lines

Publié en ligne: 05 août 2024

Pages: -

Reçu: 17 avr. 2024

Accepté: 10 juil. 2024

DOI: https://doi.org/10.2478/amns-2024-1930

Mots clésAttention model, Bi-LSTM, Sentiment analysis model, Feature extraction, Film and TV drama lines

© 2024 Yuanfang Jiang, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Mots clés
Attention model, Bi-LSTM, Sentiment analysis model, Feature extraction, Film and TV drama lines