An official website of the United States government
In the field of healthcare, electronic health records (EHR) serve as crucial training data for developing machine learning models for diagnosis, treatment, and the management of healthcare resources. However, medical datasets are often imbalanced in terms of sensitive attributes such as race/ethnicity, gender, and age. Machine learning models trained on class-imbalanced EHR datasets perform significantly worse in deployment for individuals of the minority classes compared to those from majority classes, which may lead to inequitable healthcare outcomes for minority groups. To address this challenge, we propose Minority Class Rebalancing through Augmentation by Generative modeling (MCRAGE), a novel approach to augment imbalanced datasets using samples generated by a deep generative model. The MCRAGE process involves training a Conditional Denoising Diffusion Probabilistic Model (CDDPM) capable of generating high-quality synthetic EHR samples from underrepresented classes. We use this synthetic data to augment the existing imbalanced dataset, resulting in a more balanced distribution across all classes, which can be used to train less biased downstream models. We measure the performance of MCRAGE versus alternative approaches using Accuracy, F1 score and AUROC of these downstream models. We provide theoretical justification for our method in terms of recent convergence results for DDPMs.
more »
« less
Identifying Cyberbullying Roles in Social Media
Social media has revolutionized communication, allowing people worldwide to connect and interact instantly. However, it has also led to increases in cyberbullying, which poses a significant threat to children and adolescents globally, affecting their mental health and well-being. It is critical to accurately detect the roles of individuals involved in cyberbullying incidents to effectively address the issue on a large scale. This study explores the use of machine learning models to detect the roles involved in cyberbullying interactions. After examining the AMiCA dataset and addressing class imbalance issues, we evaluate the performance of various models built with four underlying LLMs (i.e. BERT, RoBERTa, T5, and GPT-2) for role detection. Our analysis shows that oversampling techniques help improve model performance. The best model, a fine-tuned RoBERTa using oversampled data, achieved an overall F1 score of 83.5%, increasing to 89.3% after applying a prediction threshold. The top-2 F1 score without thresholding was 95.7%. Our method outperforms previously proposed models. After investigating the per-class model performance and confidence scores, we show that the models perform well in classes with more samples and less contextual confusion (e.g. Bystander Other), but struggle with classes with fewer samples (e.g. Bystander Assistant) and more contextual ambiguity (e.g. Harasser and Victim). This work highlights current strengths and limitations in the development of accurate models with limited data and complex scenarios.
more »
« less
- Award ID(s):
- 2227488
- PAR ID:
- 10572879
- Publisher / Repository:
- The 2024 International Conference on Advances in Social Networks Analysis and Mining (ASONAM)
- Date Published:
- Subject(s) / Keyword(s):
- cyberbullying role detection social media LLM
- Format(s):
- Medium: X
- Location:
- Calabria, Italy
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Cyberbullying is a growing problem across social media platforms, inflicting short and long-lasting effects on victims. To mitigate this problem, research has looked into building automated systems, powered by machine learning, to detect cyberbullying incidents, or the involved actors like victims and perpetrators. In the past, systematic reviews have examined the approaches within this growing body of work, but with a focus on the computational aspects of the technical innovation, feature engineering, or performance optimization, without centering around the roles, beliefs, desires, or expectations of humans. In this paper, we present a human-centered systematic literature review of the past 10 years of research on automated cyberbullying detection. We analyzed 56 papers based on a three-prong human-centeredness algorithm design framework - spanning theoretical, participatory, and speculative design. We found that the past literature fell short of incorporating human-centeredness across multiple aspects, ranging from defining cyberbullying, establishing the ground truth in data annotation, evaluating the performance of the detection models, to speculating the usage and users of the models, including potential harms and negative consequences. Given the sensitivities of the cyberbullying experience and the deep ramifications cyberbullying incidents bear on the involved actors, we discuss takeaways on how incorporating human-centeredness in future research can aid with developing detection systems that are more practical, useful, and tuned to the diverse needs and contexts of the stakeholders.more » « less
-
Abstract Background Diabetic retinopathy (DR) is a leading cause of blindness in American adults. If detected, DR can be treated to prevent further damage causing blindness. There is an increasing interest in developing artificial intelligence (AI) technologies to help detect DR using electronic health records. The lesion-related information documented in fundus image reports is a valuable resource that could help diagnoses of DR in clinical decision support systems. However, most studies for AI-based DR diagnoses are mainly based on medical images; there is limited studies to explore the lesion-related information captured in the free text image reports. Methods In this study, we examined two state-of-the-art transformer-based natural language processing (NLP) models, including BERT and RoBERTa, compared them with a recurrent neural network implemented using Long short-term memory (LSTM) to extract DR-related concepts from clinical narratives. We identified four different categories of DR-related clinical concepts including lesions, eye parts, laterality, and severity, developed annotation guidelines, annotated a DR-corpus of 536 image reports, and developed transformer-based NLP models for clinical concept extraction and relation extraction. We also examined the relation extraction under two settings including ‘gold-standard’ setting—where gold-standard concepts were used–and end-to-end setting. Results For concept extraction, the BERT model pretrained with the MIMIC III dataset achieve the best performance (0.9503 and 0.9645 for strict/lenient evaluation). For relation extraction, BERT model pretrained using general English text achieved the best strict/lenient F1-score of 0.9316. The end-to-end system, BERT_general_e2e, achieved the best strict/lenient F1-score of 0.8578 and 0.8881, respectively. Another end-to-end system based on the RoBERTa architecture, RoBERTa_general_e2e, also achieved the same performance as BERT_general_e2e in strict scores. Conclusions This study demonstrated the efficiency of transformer-based NLP models for clinical concept extraction and relation extraction. Our results show that it’s necessary to pretrain transformer models using clinical text to optimize the performance for clinical concept extraction. Whereas, for relation extraction, transformers pretrained using general English text perform better.more » « less
-
Social media cyberbullying has a detrimental effect on human life. As online social networking grows daily, the amount of hate speech also increases. Such terrible content can cause depression and actions related to suicide. This paper proposes a trustable LSTM Autoencoder Network for cyberbullying detection on social media using synthetic data. We have demonstrated a cutting-edge method to address data availability difficulties by producing machine-translated data. However, several languages such as Hindi and Bangla still lack adequate investigations due to a lack of datasets. We carried out experimental identification of aggressive comments on Hindi, Bangla, and English datasets using the proposed model and traditional models, including Long Short-Term Memory (LSTM), Bidirectional Long Short-Term Memory (BiLSTM), LSTM-Autoencoder, Word2vec, Bidirectional Encoder Representations from Transformers (BERT), and Generative Pre-trained Transformer 2 (GPT-2) models. We employed evaluation metrics such as f1-score, accuracy, precision, and recall to assess the models’ performance. Our proposed model outperformed all the models on all datasets, achieving the highest accuracy of 95%. Our model achieves state-of-the-art results among all the previous works on the dataset we used in this paper.more » « less
-
Teens often encounter cyberbullying on social media. One promising way to reduce cyberbullying is through empowering teens to stand up for their peers and cultivating prosocial norms online. While there is no shortage of bystander interventions that have shown potential, little research has explored designing chatbots with users to provide a contextualized and embedded “learning at the moment” experience for bystanders. This study involved teens and educators in two design sessions: an in-depth interview to identify the barriers that prevent upstanding behaviors, and interaction with the “social media co-pilot'' chatbot prototype to identify design guidelines to empower teens to overcome these barriers. Qualitative analysis on the conversations from the two design sessions revealed three factors that curb teens' upstanding behaviors: a) inadequate knowledge about social norms, appropriate language, and consequences, b) inhibitive emotions such as fear of retaliation and confrontation; c) lack of empathy toward their peers. Key parameters were also identified to shape chatbot responses to encourage upstanding behaviors, such as a) adopting voices representing multiple roles, b) empathetic, friendly and encouraging tone, c) reflective, specific and relatable language and d) appropriate length. These insights inform the design of personalized and scalable education programs and moderation tools to combat cyberbullying.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
