We present UNF-BMI system for SemEval-2026 Task 3, Track A, Subtask 1 (Dimensional Aspect Sentiment Regression, DimASR), which focuses on predicting continuous Valence–Arousal (VA) scores for aspects in text. Our approach integrates psychologically grounded affective signals inspired by the Research Domain Criteria (RDoC) framework. We investigate two complementary methods: first, an in-context learning framework using Mistral-7B-Instruct with semantically retrieved few-shot examples augmented by lexicon-derived RDoC valence and arousal cues; second, a supervised multi-task learning model based on RoBERTa, where VA regression is the primary objective and RDoC-based positive/negative signal prediction serves as an auxiliary task to regularize shared representations. Experiments on english laptop and restaurant review datasets demonstrate that incorporating RDoC-inspired affective priors reduces RMSE compared to baselines, particularly in low-signal text where explicit sentiment cues are sparse.

Accurate linguistic annotation is crucial for creating high-quality datasets in specialized domains, yet manual labeling is often slow, expensive, and inconsistent. We present a reproducible workflow for evaluating the effectiveness of large language models (LLMs) as annotators of domain-specific health misinformation on social media. Using a data set of 169 Instagram posts on seed oils, expert nutritionists provided gold-standard labels (71% positives), which we compared against the outputs of five open-source LLMs. We introduce a hierarchical error taxonomy that categorizes LLM misclassifications according to the direction, mechanism, and contributing factors of the error, providing interpretable insights into model failures. Our analysis reveals systematic error patterns, including misinterpretation of nuanced claims and overconfidence in predictions, highlighting conditions under which LLM annotations do not align with expert judgment. Although the data set is modest in size and exhibits class imbalance, it reflects real-world distributions of nutrition-related Instagram content and motivates the need for a careful evaluation of the robustness of the LLM annotation. This study has implications for the development of frameworks for automated LLM-based annotators in the health and nutrition domains, as well as LLM developers in general.

Research Domain Criteria (RDoC) is a National Institute of Mental Health framework for studying mental disorders by integrating information across genetics, circuits, and behavior. Manually curating biomedical abstracts relevant to RDoC is a significant challenge due to semantically overlapping construct definitions (e.g., "Acute Threat," "Potential Threat," and "Sustained Threat") and the exponential growth of biomedical literature. This study compares two modeling strategies, domain-adapted fine-tuning and in-context prompting, across two RDoC-related subtasks from the official BioNLP-OST 2019 RDoC shared task. For Task 1, unlabeled PubMed abstracts are retrieved and ranked by relevance to eight of the RDoC constructs. We compare a TF-IDF baseline against ModernBERT and Llama (zero-shot and five-shot) using Mean Average Precision (MAP). For Task 2, the objective is to identify the single most relevant sentence from an abstract for a given construct, evaluated using per-construct accuracy. The fine-tuning track performs end-to-end fine-tuning of BioBERT, PubMedBERT, ModernBERT, and RoBERTa using a cross-encoder input format and per-construct grid search. These are compared against the in-context learning of several open-source language models. Both our approaches are competitive against the best-performing team’s score from the BioNLP-OST 2019 RDoC shared task. Taken together, these findings suggest that five-shot prompted LLMs and domain-adapted fine-tuned transformers are viable tools for semi-automating the expert annotation in RDoC curation.

Nutrition misinformation on social media often arises from selective interpretation of scientific evidence rather than outright falsehoods, making it difficult to detect. We introduce a curated, expert-annotated Instagram dataset focused on seed oils and omega-6, two domains characterized by contested dietary claims. We evaluate feature-based, embedding-based, and transformer-based models under in-domain and cross-domain settings. Results show strong in-domain performance across all models, with Sentence-BERT achieving the highest AUPRC (up to 0.96). However, performance drops substantially under cross-domain transfer, indicating limited robustness to topic shift. Analysis suggests that while contextual embeddings capture strong in-domain semantic signals, linguistically and psychologically grounded features are more stable under distribution shift. These findings highlight the value of combining semantic and interpretable linguistic signals for robust misinformation detection.

Institutes are required to catalog their articles with proper subject headings so that the users can easily retrieve relevant articles from the institutional repositories. However, due to the rate of proliferation of the number of articles in these repositories, it is becoming a challenge to manually catalog the newly added articles at the same pace. To address this challenge, we explore the feasibility of automatically annotating articles with Library of Congress Subject Headings (LCSH). We first use web scraping to extract keywords for a collection of articles from the Repository Analytics and Metrics Portal (RAMP). Then, we map these keywords to LCSH names for developing a gold-standard dataset. As a case study, using the subset of Biology-related LCSH concepts, we develop predictive models by formulating this task as a multi-label classification problem. Our experimental results demonstrate the viability of this approach for predicting LCSH for scholarly articles.

Electronic health records (EHRs) are notorious for reducing the face-to-face time with patients while increasing the screen-time for clinicians leading to burnout. This is especially problematic for psychiatry care in which maintaining consistent eye-contact and non-verbal cues are just as important as the spoken words. In this ongoing work, we explore the feasibility of automatically generating psychiatric EHR case notes from digital transcripts of doctor-patient conversation using a two-step approach: (1) predicting semantic topics for segments of transcripts using supervised machine learning, and (2) generating formal text of those segments using natural language processing. Through a series of preliminary experimental results obtained through a collection of synthetic and real-life transcripts, we demonstrate the viability of this approach.

BioNLP Open Shared Tasks (BioNLP-OST) is an international competition organized to facilitate development and sharing of computational tasks of biomedical text mining and solutions to them. For BioNLP-OST 2019, we introduced a new mental health informatics task called “RDoC Task”, which is composed of two subtasks: information retrieval and sentence extraction through National Institutes of Mental Health’s Research Domain Criteria framework. Five and four teams around the world participated in the two tasks, respectively. According to the performance on the two tasks, we observe that there is room for improvement for text mining on brain research and mental illness.