Publications

2026

1. Like a Therapist, But Not: Reddit Narratives of AI in Mental Health Contexts

   Elham Aghakhani, and Rezvaneh Rezapour

   2026

   Abs

   Large language models (LLMs) are increasingly used for emotional support and mental health-related interactions outside clinical settings, yet little is known about how people evaluate and relate to these systems in everyday use. We analyze 5,126 Reddit posts from 47 mental health communities describing experiential or exploratory use of AI for emotional support or therapy. Grounded in the Technology Acceptance Model and therapeutic alliance theory, we develop a theory-informed annotation framework and apply a hybrid LLM-human pipeline to analyze evaluative language, adoption-related attitudes, and relational alignment at scale. Our results show that engagement is shaped primarily by narrated outcomes, trust, and response quality, rather than emotional bond alone. Positive sentiment is most strongly associated with task and goal alignment, while companionship-oriented use more often involves misaligned alliances and reported risks such as dependence and symptom escalation. Overall, this work demonstrates how theory-grounded constructs can be operationalized in large-scale discourse analysis and highlights the importance of studying how users interpret language technologies in sensitive, real-world contexts.

2025

1. The Evolving Landscape of Youth Online Safety: Insights from News Media Analysis

   Mohammad (Matt) Namvarpour, Elham Aghakhani, Michael D Ekstrand, Rezvaneh Rezapour, and 1 more author

   In Proceedings of the 17th ACM Web Science Conference 2025 2025

   Abs

   There have been various efforts to understand how youth online safety has been reflected in the news, as news play an important role in shaping public opinions. However, these efforts focused on specific contexts, such as individual countries and specific online risk types. Therefore, there is a need for a holistic view of understanding trends in the news regarding stakeholders involved and various ranges of online risks. In this work, we seek to understand how discussions of online safety for youth has evolved in news publications over the last two decades. We applied quantitative media content analysis and sentiment analysis to 3.9K English-language news articles from 2002–2024, documenting shifts in the portrayal of key stakeholders. Our results showed increased media focus on technology companies and government in youth safety discussions, particularly highlighting cyberbullying as a key risk. We found a generally negative trend in the sentiment toward the perceived safety of youth online, which fluctuates based on societal concerns and policy changes. The significance of this work lies in its analysis of how media discourse has illuminated public perceptions and policy directions concerning youth safety in digital spaces. Content Warning: This paper discusses sensitive topics, such as sex and child harassment, which may be triggering.

2. From Conversation to Automation: Leveraging LLMs for Problem-Solving Therapy Analysis

   Elham Aghakhani, Lu Wang, Karla T. Washington, George Demiris, and 2 more authors

   In Findings of the Association for Computational Linguistics: ACL 2025 Jul 2025

   Abs

   Problem-Solving Therapy (PST) is a structured psychological approach that helps individuals manage stress and resolve personal issues by guiding them through problem identification, solution brainstorming, decision-making, and outcome evaluation. As mental health care increasingly adopts technologies like chatbots and large language models (LLMs), it is important to thoroughly understand how each session of PST is conducted before attempting to automate it. We developed a comprehensive framework for PST annotation using established PST Core Strategies and a set of novel Facilitative Strategies to analyze a corpus of real-world therapy transcripts to determine which strategies are most prevalent. Using various LLMs and transformer-based models, we found that GPT-4o outperformed all models, achieving the highest accuracy (0.76) in identifying all strategies. To gain deeper insights, we examined how strategies are applied by analyzing Therapeutic Dynamics (autonomy, self-disclosure, and metaphor), and linguistic patterns within our labeled data. Our research highlights LLMs’ potential to automate therapy dialogue analysis, offering a scalable tool for mental health interventions. Our framework enhances PST by improving accessibility, effectiveness, and personalized support for therapists.

3. AnnoLoom: Augmenting Codebook Generation and Annotation with Large Language Models

   Lu Wang, Duncan Lynch, Elham Aghakhani, George Demiris, and 3 more authors

   Proceedings of the Association for Information Science and Technology Jul 2025

   Abs

   We introduce AnnoLoom, a tool designed to assist researchers with codebook development, annotation tasks, and evaluation of human vs. AI’s annotation results. AnnoLoom contributes to human expert-AI collaboration and its efficacy in the context of using Large Language Models (LLMs) for research involving text-based data. We conducted a cognitive walkthrough to iteratively improve the design of AnnoLoom and discussed the future work.

4. Phenotypes of stigma expressed by people who use drugs on Reddit

   Layla Bouzoubaa, Elham Aghakhani, and Rezvaneh Shadi Rezapour

   Social Science & Medicine Jul 2025

   Abs

   Rationale: Despite record-high overdose deaths in the U.S., most individuals meeting criteria for substance use disorder remain outside formal treatment systems. Stigma is a major contributor to this treatment gap yet remains difficult to study among people who use drugs (PWUD) who are not engaged in clinical care. Social media platforms like Reddit offer a valuable window into the lived experiences of stigma of this population through naturally occurring discourse. This study develops a comprehensive framework for identifying stigma expressions in social media discourse, identifies distinct patterns using computational methods, and examines how these patterns relate to established stigma theory. Methods: We analyzed over one million posts from six drug-related subreddits using mixed-methods. Large language models with human validation identified and classified stigma-related content across validated dimensions of narrativity, stigma experience, and psycholinguistic features. K-means clustering identified distinct stigma expression patterns (phenotypes), which were then characterized through comprehensive linguistic analysis. Results: Analysis of 1,033,619 posts identified 56, 446 stigma-containing posts and revealed a novel classification — Stigma Perceptions & Commentary (SPC) - which captures the broader discourse on stigma beyond personal experiences. Clustering analysis of these stigma posts plus 5495 non-stigma posts (61,941 total) revealed three distinct phenotypes: Internalized Stigma (34.5%), characterized by self-blame, high narrative agency, and avoidant coping; Public Stigma (38.9%), featuring discrimination from healthcare systems with mixed coping; and Righteous Indignation (26.6%), marked by analytical critique of systemic issues. Conclusion: These phenotypes align with theoretical models of self-stigma and demonstrates the potential of social media data to extend stigma research beyond clinical populations, offering insight into how PWUD experience and contest stigma in everyday discourse.

2024

1. Words Matter: Reducing Stigma in Online Conversations about Substance Use with Large Language Models

   Layla Bouzoubaa, Elham Aghakhani, and Rezvaneh Rezapour

   In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing Nov 2024

   Abs

   Stigma is a barrier to treatment for individuals struggling with substance use disorders (SUD), which leads to significantly lower treatment engagement rates. With only 7% of those affected receiving any form of help, societal stigma not only discourages individuals with SUD from seeking help but isolates them, hindering their recovery journey and perpetuating a cycle of shame and self-doubt. This study investigates how stigma manifests on social media, particularly Reddit, where anonymity can exacerbate discriminatory behaviors. We analyzed over 1.2 million posts, identifying 3,207 that exhibited stigmatizing language related to people who use substances (PWUS). Of these, 1,649 posts were classified as containing directed stigma towards PWUS, which became the focus of our de-stigmatization efforts. Using Informed and Stylized LLMs, we developed a model to transform these instances into more empathetic language.Our paper contributes to the field by proposing a computational framework for analyzing stigma and de-stigmatizing online content, and delving into the linguistic features that propagate stigma towards PWUS. Our work not only enhances understanding of stigma’s manifestations online but also provides practical tools for fostering a more supportive environment for those affected by SUD.

2. Decoding the Narratives: Analyzing Personal Drug Experiences Shared on Reddit

   Layla Bouzoubaa, Elham Aghakhani, Max Song, Quang Trinh, and 1 more author

   In Findings of the Association for Computational Linguistics ACL 2024 Aug 2024

   Abs

   Online communities such as drug-related subreddits serve as safe spaces for people who use drugs (PWUD), fostering discussions on substance use experiences, harm reduction, and addiction recovery. Users’ shared narratives on these forums provide insights into the likelihood of developing a substance use disorder (SUD) and recovery potential. Our study aims to develop a multi-level, multi-label classification model to analyze online user-generated texts about substance use experiences. For this purpose, we first introduce a novel taxonomy to assess the nature of posts, including their intended connections (Inquisition or Disclosure), subjects (e.g., Recovery, Dependency), and specific objectives (e.g., Relapse, Quality, Safety). Using various multi-label classification algorithms on a set of annotated data, we show that GPT-4, when prompted with instructions, definitions, and examples, outperformed all other models. We apply this model to label an additional 1,000 posts and analyze the categories of linguistic expression used within posts in each class. Our analysis shows that topics such as Safety, Combination of Substances, and Mental Health see more disclosure, while discussions about physiological Effects focus on harm reduction. Our work enriches the understanding of PWUD’s experiences and informs the broader knowledge base on SUD and drug use.