Long-form mental health assessments pose unique challenges for large language models (LLMs), which often exhibit hallucinations or inconsistent reasoning when handling extended, domain-specific contexts. We introduce Stacked Multi-Model Reasoning (SMMR), a layered framework that leverages multiple LLMs and specialized smaller models as coequal 'experts'. Early layers isolate short, discrete subtasks, while later layers integrate and refine these partial outputs through more advanced long-context models. We evaluate SMMR on the DAIC-WOZ depression-screening dataset and 48 curated case studies with psychiatric diagnoses, demonstrating consistent improvements over single-model baselines in terms of accuracy, F1-score, and PHQ-8 error reduction. By harnessing diverse 'second opinions', SMMR mitigates hallucinations, captures subtle clinical nuances, and enhances reliability in high-stakes mental health assessments. Our findings underscore the value of multi-expert frameworks for more trustworthy AI-driven screening.

Recent advancements in NLP have spurred significant interest in analyzing social media text data for identifying linguistic features indicative of mental health issues. However, the domain of Expressive Narrative Stories (ENS)-deeply personal and emotionally charged narratives that offer rich psychological insights-remains underexplored. This study bridges this gap by utilizing a dataset sourced from Reddit, focusing on ENS from individuals with and without self-declared depression. Our research evaluates the utility of advanced language models, BERT and MentalBERT, against traditional models. We find that traditional models are sensitive to the absence of explicit topic-related words, which could risk their potential to extend applications to ENS that lack clear mental health terminology. Despite MentalBERT is design to better handle psychiatric contexts, it demonstrated a dependency on specific topic words for classification accuracy, raising concerns about its application when explicit mental health terms are sparse (P-value<0.05). In contrast, BERT exhibited minimal sensitivity to the absence of topic words in ENS, suggesting its superior capability to understand deeper linguistic features, making it more effective for real-world applications. Both BERT and MentalBERT excel at recognizing linguistic nuances and maintaining classification accuracy even when narrative order is disrupted. This resilience is statistically significant, with sentence shuffling showing substantial impacts on model performance (P-value<0.05), especially evident in ENS comparisons between individuals with and without mental health declarations. These findings underscore the importance of exploring ENS for deeper insights into mental health-related narratives, advocating for a nuanced approach to mental health text analysis that moves beyond mere keyword detection.

Mental health issues significantly impact individuals' daily lives, yet many do not receive the help they need even with available online resources. This study aims to provide diverse, accessible, stigma-free, personalized, and real-time mental health support through cutting-edge AI technologies. It makes the following contributions: (1) Conducting an extensive survey of recent mental health support methods to identify prevalent functionalities and unmet needs. (2) Introducing SouLLMate, an adaptive LLM-driven system that integrates LLM technologies, Chain, Retrieval-Augmented Generation (RAG), prompt engineering, and domain knowledge. This system offers advanced features such as Risk Detection and Proactive Guidance Dialogue, and utilizes RAG for personalized profile uploads and Conversational Information Extraction. (3) Developing novel evaluation approaches for preliminary assessments and risk detection via professionally annotated interview data and real-life suicide tendency data. (4) Proposing the Key Indicator Summarization (KIS), Proactive Questioning Strategy (PQS), and Stacked Multi-Model Reasoning (SMMR) methods to enhance model performance and usability through context-sensitive response adjustments, semantic coherence evaluations, and enhanced accuracy of long-context reasoning in language models. This study contributes to advancing mental health support technologies, potentially improving the accessibility and effectiveness of mental health care globally.

Finding the longest common subsequence (LCS) of multiple strings is an NP-hard problem, with many applications in the areas of bioinformatics and computational genomics. Although significant efforts have been made to address the problem and its special cases, the increasing complexity and size of biological data require more efficient methods applicable to an arbitrary number of strings. In this paper, a novel search algorithm, MLCS-A*, is presented for the general case of multiple LCS (or MLCS) problems. MLCS-A* is a variant of the A* algorithm. It maximizes a new heuristic estimate of the LCS in each search step so that the longest common subsequence can be found. As a natural extension of MLCS-A*, a fast algorithm, MLCS-APP, is also proposed to deal with large volume of biological data for which finding a LCS within reasonable time is impossible. The benchmark test shows that MLCS-APP is able to extract common subsequences close to the optimal ones and that MLCS-APP significantly outperforms existing heuristic approaches. When applied to 8 protein domain families, MLCS-APP produced more accurate results than existing multiple sequence alignment methods.