<p>Opinion mining has&#xa0;gradually&#xa0;become&#xa0;tough&#xa0;to&#xa0;handle as&#xa0;the&#xa0;amount&#xa0;of&#xa0;user-generated&#xa0;content on social media&#xa0;platforms increases exponentially.&#xa0;As a matter of fact, Twitter is&#xa0;the&#xa0;most common platform&#xa0;to&#xa0;collect voices&#xa0;in&#xa0;terms of products, advancements, and policies.&#xa0;Sentiment Analysis (SA) deals with people's thoughts, feelings, and opinions about various subjects.&#xa0;Through the examination of tweets, one can gauge public views&#xa0;on&#xa0;news, rules, the community, and even celebrities.&#xa0;Unfortunately, currently existing SA mechanisms often have limited prediction capabilities and are still a long way from being able to function&#xa0;in&#xa0;real, time commercial applications.&#xa0;The main causes of inaccuracies are lack of data and difficulties in model configuration in deep learning (DL). This research introduces a classification learning-based Optimal Tiered blocks&#xa0;of&#xa0;Convolutional Neural Long Short, Term Memory (OTCNLSTM) for emotion recognition.&#xa0;Four Local Features Training Blocks (LFTBs), which are capable&#xa0;of&#xa0;hierarchically extracting spatiotemporal local emotional cues, make up the OTCNLSTM model.&#xa0;On top of that,&#xa0;the&#xa0;Boosted Killer Whale Predation Optimization (BKWOP) technique is introduced to pinpoint the perfect hyperparameters and solution sets, thus forming&#xa0;a&#xa0;stable neural network model. The newly designed system efficiently categorizes&#xa0;the&#xa0;sentiments generated from the Twitter users' comments into four categories, namely, positive, negative, neutral, and irrelevant.&#xa0;The Kaggle Twitter dataset powered a thorough experimental study, which showed&#xa0;the&#xa0;model's performance like a rock.&#xa0;The OTCNLSTM model reached an overall accuracy.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Sentiment analysis based on deep features through textual tweets classification using optimal hierarchical CLSTM network model

  • Nithya Thanakodi,
  • Nagarajan Deivanayagampillai,
  • Rajendran Thavasimuthu

摘要

Opinion mining has gradually become tough to handle as the amount of user-generated content on social media platforms increases exponentially. As a matter of fact, Twitter is the most common platform to collect voices in terms of products, advancements, and policies. Sentiment Analysis (SA) deals with people's thoughts, feelings, and opinions about various subjects. Through the examination of tweets, one can gauge public views on news, rules, the community, and even celebrities. Unfortunately, currently existing SA mechanisms often have limited prediction capabilities and are still a long way from being able to function in real, time commercial applications. The main causes of inaccuracies are lack of data and difficulties in model configuration in deep learning (DL). This research introduces a classification learning-based Optimal Tiered blocks of Convolutional Neural Long Short, Term Memory (OTCNLSTM) for emotion recognition. Four Local Features Training Blocks (LFTBs), which are capable of hierarchically extracting spatiotemporal local emotional cues, make up the OTCNLSTM model. On top of that, the Boosted Killer Whale Predation Optimization (BKWOP) technique is introduced to pinpoint the perfect hyperparameters and solution sets, thus forming a stable neural network model. The newly designed system efficiently categorizes the sentiments generated from the Twitter users' comments into four categories, namely, positive, negative, neutral, and irrelevant. The Kaggle Twitter dataset powered a thorough experimental study, which showed the model's performance like a rock. The OTCNLSTM model reached an overall accuracy.