Dual stream deep learning for fake news-aware stock prediction: Integrating technical indicators and sentiment analysis

An exploratory analysis of self-diagnosis on Reddit

Sentiment analysis using Kernel Variance Projection and LR–BiLGMP–Skd deep learning model

Capturing subjectivity: A weighted ensemble approach to preserve annotator diversity

• Applies Traffic Sentiment Analysis (TSA) to evaluate driver education policy. • ParentTaught.com users report 83.6% positive sentiment focused on ease of use. • Reddit analysis reveals a latent theme of parental anxiety and teaching stress. • Identifies instructional conflict as a precursor to road chain conflict risks. • Recommends digital support tools to mitigate in-car parental stress. Driver education is critical in preparing teens with the skills and knowledge necessary for safe and independent mobility. While Traditional Driver Education (TDE) provides essential instruction, challenges such as high cost and limited practical training time have led several U.S. states to adopt Parent-Taught Driver Education (PTDE). Despite this legislative expansion, there is a gap in understanding the user experience of families utilizing these programs. To address this, this study evaluates online sentiment and perceived effectiveness of PTDE by analyzing unstructured user feedback through the lens of Traffic Sentiment Analysis (TSA). A mixed-method computational framework integrating VADER sentiment analysis, Latent Dirichlet Allocation (LDA), and text network analysis was used to analyze 2,872 comments from Reddit (2015–2024) and ParentTaught.com (2021–2025). The results indicate a positive public sentiment toward PTDE, particularly on the service platform ParentTaught.com (83.6% positive), where users praised the program’s efficiency, convenience, and instructional clarity. Topic modeling revealed that while the primary user experience is characterized by satisfaction with the curriculum’s accessibility, a distinct sub-theme of parental anxiety emerged in community discussions (Reddit), driven by the responsibility of teaching complex safety skills. Text network analysis further identified that while the core PTDE network is cohesive and ease-focused, specific friction points exist regarding administrative logistics. The study concludes that PTDE is a highly regarded educational model that successfully meets family needs, though its effectiveness could be further optimized by providing parents with digital tools to manage the emotional load of in-car instruction.

The innovation–compliance–perception framework as a lens for AI governance — NLP evidence from Meta's smart glasses and GDPR discourse

LLM sentiment quantification reveals selective alignment with human course-evaluation raters

DAGF: A dual GCN and auxiliary graph fusion based model for aspect-based sentiment analysis

The government has recently adopted mobile applications to enhance service delivery for citizens. However, these applications often generate mixed reactions among users. Many citizens express their opinions through reviews and ratings on the Google Play Store, providing valuable information for sentiment analysis. Leveraging this, the present paper introduces the Indonesian Government Application Review (IGAR) dataset, a collection of 617,722 user reviews from six popular government-related applications in Indonesia: Mobile JKN, MyPertamina, KAI, JMO, Satusehat, and BMKG. The reviews, originally written in Indonesian, were manually annotated as positive, neutral, or negative based on rating scores. Among the dataset, positive sentiment accounts for 336,449 reviews, negative sentiment with 246,898 reviews, while 34,375 reviews are categorized as neutral. To extend the usability of the dataset for broader research contexts, all reviews were translated into English and further processed using the Valence Aware Dictionary and sEntiment Reasoner (VADER) for automated sentiment labeling. Through VADER classification, 324,660 reviews were identified as positive, 173,329 as neutral, and 119,733 as negative. This dataset thus provides a valuable resource for advancing sentiment classification research using machine learning and deep learning model on government-related applications in Indonesia.

Scoping review of multimodal sentiment analysis and summarization: State of the art, challenges and future directions

Shaping the future: A pilot study on how AI-powered chatbots shape patient perceptions of pharmacist roles.

Toward multimodal sentiment analysis with a self-supervised knowledge-augmented network

SEHLP: A summary-enhanced large language model for financial report sentiment analysis via hybrid LoRA and dynamic prefix tuning

IFSA-CE: Interpretable fine-grained sentiment analysis with concept embedding

Social media sites provide warning signs for shifts in consumer behavior, competitive forces, and emerging market trends. However, most small and medium-sized businesses (SMBs) do not have a systematic and scalable approach to tap into this unstructured data from various sites to extract insights. This paper proposes a trend detection method that leverages automated data extraction with n8n workflows, transformer-based embeddings, hybrid sentiment analysis, BERTopic clustering, and a weighted TrendScore composite score. The proposed approach combines multiple, heterogeneous inputs into a single analytical workflow and offers explainable visual and conversational BI interfaces, which are specifically designed for SMBs. The parameter definitions, scoring rules, and workflow diagrams are carefully detailed to ensure that the approach is fully reproducible. The proposed approach focuses on interpretability, robustness across multiple platforms, and applicability within a resource-constrained business setting. • Reproducible multi-platform social data acquisition using exportable n8n workflows. • Hybrid Transformer-Lexicon sentiment modeling combined with BERTopic clustering. • Quantified TrendScore integrating growth, engagement, sentiment shift, and cross-platform consistency.

DAGG-Net: Dual adaptive graph and gating network for multimodal aspect-based sentiment analysis

Personalized federated learning for multilingual sentiment analysis poses significant challenges arising from linguistic heterogeneity, non-IID data distributions, and strict privacy requirements. This paper proposes FedPerX, a federated transformer framework that integrates residual adapter-based personalization with adaptive multi-granular differential privacy. The architecture leverages a frozen multilingual backbone (XLM-R) while enabling each client to train lightweight, client-specific adapters. Privacy is enforced through dynamic noise injection at both the feature and adapter levels, calibrated using gradient sensitivity. FedPerX is evaluated on two multilingual benchmarks-MARC and TSMD-spanning structured reviews and informal social media content across more than ten languages. Experimental results demonstrate consistent improvements over seven state-of-the-art baselines, with up to +4.3% gains in macro-F1, a 70% reduction in communication overhead, and the lowest variance in client-level performance. Comprehensive analyses, including fairness, personalization gap, privacy-utility trade-off, and ablation studies, validate the framework's robustness and adaptability. FedPerX advances the design of scalable, personalized, and privacy-preserving models for federated multilingual sentiment analysis.

RCA-Net: A context-aware relational network for sentiment analysis in the metaverse

TPIN: Text-based parallel interaction network with modality-common and modality-specific for multimodal sentiment analysis

Hybrid prompt learning and multilevel knowledge distillation for multimodal sentiment analysis with missing modalities