Identifying Emotion Labels From Psychiatric Social Texts Using a Bi-Directional LSTM-CNN Model

Wu, Jheng-Long; He, Yuanye; Yu, Liang-Chih; Lai, K. Robert

doi:10.1109/access.2020.2985228

Cited by 33 publications

(14 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To tackle the problems of machine learning based methods, an increasing number of researchers try to apply deep learning based methods to tasks of NLP, such as sentiment analysis [1,13,29] and emotion classification [30][31][32][33][34][35] . As deep learning models combined with graphs, such as GNN [36] and graph convolutional networks (GCN) [37] , have been proposed, an increasing number of researchers utilize them for emotion classification [38][39][40] .…”

Section: Related Workmentioning

confidence: 99%

$p$-Norm Broad Learning for Negative Emotion Classification in Social Networks

Chen

Peng

Zeng

et al. 2022

Big Data Min. Anal.

View full text Add to dashboard Cite

Negative emotion classification refers to the automatic classification of negative emotion of texts in social networks. Most existing methods are based on deep learning models, facing challenges such as complex structures and too many hyperparameters. To meet these challenges, in this paper, we propose a method for negative emotion classification utilizing a Robustly Optimized BERT Pretraining Approach (RoBERTa) and p-norm Broad Learning (p-BL). Specifically, there are mainly three contributions in this paper. Firstly, we fine-tune the RoBERTa to adapt it to the task of negative emotion classification. Then, we employ the fine-tuned RoBERTa to extract features of original texts and generate sentence vectors. Secondly, we adopt p-BL to construct a classifier and then predict negative emotions of texts using the classifier. Compared with deep learning models, p-BL has advantages such as a simple structure that is only 3-layer and fewer parameters to be trained. Moreover, it can suppress the adverse effects of more outliers and noise in data by flexibly changing the value of p. Thirdly, we conduct extensive experiments on the public datasets, and the experimental results show that our proposed method outperforms the baseline methods on the tested datasets.

show abstract

Section: Related Workmentioning

confidence: 99%

$p$-Norm Broad Learning for Negative Emotion Classification in Social Networks

Chen

Peng

Zeng

et al. 2022

Big Data Min. Anal.

View full text Add to dashboard Cite

show abstract

“…In LSTM networks, it is a reliable approach for improving backpropagation. While data in an LSTM flows in only one direction, data in a Bi-LSTM can travel in both directions [44]. A Bi-LSTM may reverse and serially process inputs.…”

Section: Bidirectional Lstm (Bi-lstm)mentioning

confidence: 99%

Detecting Cybercrime: An Evaluation of Machine Learning and Deep Learning Using Natural Language Processing Techniques on the Social Network

Amer

Siddiqui

Athamena

2022

Preprint

View full text Add to dashboard Cite

The widespread use of online social networks has culminated in across-the-board social communication among users, resulting in a considerable amount of user-generated contact data. Cybercrime has become a significant issue in recent years with the rise of online communication and social network. Cybercrime has lately been identified as a severe national psychological concern among platform users, and building a reliable detection model is crucial. Cyberbullying is the phrase used to describe such online harassment, insults, and attacks. It has become challenging to identify such unauthorized content due to the massive number of user-generated content. Because deep neural networks have various advantages over conventional machine learning approaches, researchers are turning to them more frequently to identify cyberbullying. Deep learning and machine learning have several uses in text classification. This article suggested the novel neural network model through parameters of an algorithmic and optimization comparative analysis of nine category approaches, four neural networks, and five machine learning, in two scenarios with real-world datasets of cyberbullying. Moreover, this work also analyzes the impact of word embedding and feature extraction techniques based on text mining and NLP on algorithms' performances. We performed extensive experiments on the two scenarios with a split dataset to demonstrate the merit of this research, comparing nine classification approaches through five feature extraction techniques. Our proposed cybercriminal detection model using neural networks, deep learning, and machine learning outperforms the existing state-of-the-art method of cybercriminal detection in terms of accuracy achieving higher performance.

show abstract

“…The deep learning approach majorly involves using LSTM. Deep learning approach can be used solely or combined with some other approaches to extract emotion out of the text as in [6,[9][10][11][12]. A deep learning model in [9] uses pretrained embeddings like GloVe and word2vec.…”

Section: Emotion Recognition From Textmentioning

confidence: 99%

“…The previously discussed papers' methods HMMs or RNNs cannot be used for tracking the contexts for a longer horizon. A combination of RNN and CNN is proposed in [12] where it associates an emotional label to social psychiatric texts using a deep learning model. From a survey of papers conducted and discussed above, it can be said that for emotion recognition from the text, a deep learning approach will provide the best results.…”

Section: Emotion Recognition From Textmentioning

confidence: 99%

Prosodic Speech Synthesis of Narratives Depicting Emotional Diversity Using Deep Learning

Shah¹,

Gupta²,

Jardosh³

et al. 2021

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

Emotions are an essential part of speech or communication, which is why they cannot be neglected. The existing text-to-speech systems are not the most appropriate at conveying the emotions present behind the text. The systems can speak out the text monotonically lacking expressiveness. In this paper, an Expressive Textto-Speech Synthesis System (ETSSS) is proposed which considers the dominant emotions in the text provided. ETSSS works in two parts: first, it identifies the label behind the text, and second produces expressive speech. In the first part, the input text is given an emotional label. Later, this label is used to generate expressive and prosodic speech. Labeling emotions in ETSSS is carried out using BERT which has an accuracy of 94%, 90%, and 90% for disgust, amused, and anger, respectively. The speech synthesis with the emotion module of ETSSS achieves a good MOS of 3.8 for anger, 3.5 for disgust, and 3.2 for amused. IntroductionGenerating speech from text has been used for the past decade. It is important to note that emotions in speech play an important role. The three most common aspects of speech include intelligence, naturalness, and expressiveness. Prosody is defined as

show abstract

Identifying Emotion Labels From Psychiatric Social Texts Using a Bi-Directional LSTM-CNN Model

Cited by 33 publications

References 45 publications

$p$-Norm Broad Learning for Negative Emotion Classification in Social Networks

$p$-Norm Broad Learning for Negative Emotion Classification in Social Networks

Detecting Cybercrime: An Evaluation of Machine Learning and Deep Learning Using Natural Language Processing Techniques on the Social Network

Prosodic Speech Synthesis of Narratives Depicting Emotional Diversity Using Deep Learning

Contact Info

Product

Resources

About