Science AI Journal
Computer Science · Research topic

Open research questions in Persona Design and Applications

51 unresolved questions extracted from the limitations and future-work sections of 226 Persona Design and Applications papers in our library. Each links back to the study that raised it.

What the literature leaves open

  • Existing bias audits have limitations in capturing dynamic shifts when models adopt social roles. There is a need for a novel metric to measure persona-induced bias amplification and integrate local explainability insights.

    Invisible Influences: Investigating Implicit Intersectional Biases through Persona Engineering in Large Language Models · 2026 · DOI

  • Empirical validation of the yesterday.json architecture. Application of the architecture to various AI personas. Investigation of the limitations of the architecture.

    Episodic Memory for AI Personas via Self-Authored Emotional State Snapshots: The yesterday.json Architecture · 2026 · DOI

  • Modern AI agents wake up blank and do not remember being someone. The dominant approaches to this problem treat memory as a database problem and optimize for information density.

    Episodic Memory for AI Personas via Self-Authored Emotional State Snapshots: The yesterday.json Architecture · 2026 · DOI

  • The complexity of the healthcare system. The need to involve consumers in the design process. The challenge of creating a more realistic and empathetic approach to healthcare design.

    Seeing both sides of the story: creating patient/provider persona dyads for use in mental health service co-design · 2026 · DOI

  • Algorithmic amplification of sensitive content is a challenge in adult media platforms. The lack of transparency in recommendation systems is a challenge in studying algorithmic amplification. The need for ethical protocols in adult media platforms is a challenge.

    ALGORITHMIC AMPLIFICATION IN ADULT MEDIA PLATFORMS: A STUDY ON RECOMMENDATION OF SENSITIVE CONTENT · 2026 · DOI

  • Technical-first prioritization. Lack of resources and expertise to conduct user research or follow typical UX practices. Limited understanding of user needs and context.

    Putting a Face to the Issue: Fostering User Empathy of Open Source Software Developers With PersonaFlow · 2026 · DOI

  • The need for a core operational model for live streaming agencies. The lack of paid training and weekly payouts for female creators.

    Aetera: Core Operational Model and Creator Onboarding Workflow 2026 · 2026 · DOI

  • Despite ample evidence supporting LLMs’ capabilities of generating high-quality and creative suggestions, human–LLM interactions in software testing remain under-explored, and there lacks empirical evidence supporting efficacious human–LLM system designs.

    Preemptive, Buffered, or Guided? Empirical Studies on Human–AI Interaction Strategies for Software Test Case Development · 2026 · DOI

  • Large language models (LLMs) are increasingly used to perform subjective evaluations traditionally made by humans, yet their validity as social judges remains unclear.

    How Do Large Language Models Judge Social Attraction? Evidence from Theory-Grounded Persona Ratings Across Multiple LLMs and Humans · 2026

  • However, how user-agent personality alignment influences social perception such as rapport and persuasion remains unexplored.

    The effects of personality congruence and conversational stance on human–agent interaction in virtual reality · 2026 · DOI

  • There is a need for better understanding and acceptance of learning disabilities. Literature can help fill this gap by providing insights into the experiences of children with learning disabilities.

    "That's Me!" · 1983 · DOI

  • The persistent limitations of factual hallucinations and insufficient personalization in LLM-driven travel planning systems. The lack of effective methods to mitigate hallucinations and improve personalization quality in real-time travel planning systems.

    A LangGraph-Orchestrated Multimodal Framework for Real-Time, Persona-Aware Travel Planning Using LLMs and Hallucination Resistant RAG-Based Itinerary Optimization · 2026 · DOI

  • The paper does not discuss the limitations of the BotVerse framework. The simulation is limited to the Bluesky ecosystem. The paper does not provide a comprehensive evaluation of the framework's performance.

    BotVerse: Real-Time Event-Driven Simulation of Social Agents · 2026 · DOI

  • The deployment of autonomous agents on live OSNs carries significant ethical risks. Prior work has proposed agent-based platforms, but they frequently rely on static datasets or unscalable iterative models. There is a need for a controlled simulation environment that can model complex social phenomena.

    BotVerse: Real-Time Event-Driven Simulation of Social Agents · 2026 · DOI

  • The study had a small sample size of four patients and six providers. The study was limited to a single health district in NSW, Australia. The study did not report on the long-term outcomes of the co-design process.

    Seeing both sides of the story: creating patient/provider persona dyads for use in mental health service co-design · 2026 · DOI

  • The paper identifies a research gap in the assessment of persona-conditioned large language models as synthetic survey respondents. The gap is related to the lack of understanding of when persona conditioning improves survey-grounded external validity and subgroup fidelity and when it may distort them.

    Assessing the Reliability of Persona-Conditioned LLMs as Synthetic Survey Respondents · 2026 · DOI

  • Limited attention has been given to persona-driven toxicity in non-Western contexts, particularly in Chinese-based LLMs. There is a need for large-scale, cross-model analyses of refusal behavior and persona-driven toxicity amplification in Chinese LLMs.

    Evaluating Chinese Large Language Models: The Influence of Persona Assignment on Stereotypes and Safeguards · 2026 · DOI

  • The study identifies a gap in the existing literature on the impact of internet celebrity on economic development in low-income regions. The study also identifies a gap in the understanding of how different media strategies contribute to the formation of distinct online personas.

    The Making of a Local Internet Celebrity in a Low-Income Region:A Sociological Analysis of Ding Zhen and Dishi · 2026 · DOI

  • The paper suggests that future research should focus on empirical audits of adult media platforms. Future research should also examine the impact of algorithmic amplification on users and society.

    ALGORITHMIC AMPLIFICATION IN ADULT MEDIA PLATFORMS: A STUDY ON RECOMMENDATION OF SENSITIVE CONTENT · 2026 · DOI

  • The current AI-assisted debiasing interventions have limited potential. Many change-based interventions produce effects that are modest, short-lived, not generalizable across contexts, and are poor predictors of behavioral change.

    Debiasing Interventions for the Age of AI: Design Principles from 4E Memory Systems and Intergroup Contact Theory · 2026 · DOI

  • The current AI-assisted debiasing interventions have limited potential. There is a need for a new approach that can modify the underlying psychological mechanisms of bias.

    Debiasing Interventions for the Age of AI: Design Principles from 4E Memory Systems and Intergroup Contact Theory · 2026 · DOI

  • Simulation-only study, without real-world deployment. Limited to a specific platform (Y Social) and agent design (stateless Dolphin Mistral 24B). Toxicity is misallocated across content layers.

    Towards operational validation of LLM-agent social simulations: a replicated study of a Reddit-like technology forum · 2026 · DOI

  • Improving the design of LLM-agent social simulations to reduce toxicity and increase realism. Developing more effective methods for validating agent-based models. Exploring the applications of LLM-agent social simulations in different domains.

    Towards operational validation of LLM-agent social simulations: a replicated study of a Reddit-like technology forum · 2026 · DOI

  • Future research should explore the use of LLMs for persona generation in other domains and contexts. Future research should investigate the effectiveness of participatory methods in capturing the heterogeneity of lived experiences.

    Understanding Down Syndrome Stereotypes in LLM-Based Personas · 2026 · DOI

  • The study identifies a gap in the existing literature on the use of LLMs for persona generation and the challenges of defining and mitigating stereotypes. The study highlights the need for participatory methods that capture the heterogeneity of lived experiences of people with Down syndrome.

    Understanding Down Syndrome Stereotypes in LLM-Based Personas · 2026 · DOI

Most-cited papers in Persona Design and Applications

Most recent work

Find a gap in your own Persona Design and Applications sub-topic

This page shows what the Persona Design and Applications literature already flags as unresolved. To narrow it to your specific question, run the guided finder — it searches the gap library on demand and checks candidates against 250M+ OpenAlex works.

Open the Research Gap Finder →

Related topics in Computer Science

Online Learning and Analytics
851 open questions
Digital literacy in education
518 open questions
AI in Service Interactions
517 open questions
Music Technology and Sound Studies
423 open questions
Teaching and Learning Programming
384 open questions
Quantum Computing Algorithms and Architecture
368 open questions
Virtual Reality Applications and Impacts
317 open questions
Intelligent Tutoring Systems and Adaptive Learning
298 open questions
51 open questions have been extracted from the limitations and future-work passages of 226 Persona Design and Applications papers in our library. Each one below links back to the study that raised it, so you can read the original claim in context.

Tools for your next paper

Gaps & Literature SearchFind open questions in any field — 120,000+ open gaps mined from our own library.Pre-CheckIs your paper ready? A Tier 1–5 acceptance probability in seconds.Journal FinderA ranked shortlist of target journals from a 17,500-venue index.Duplicate Publication CheckerCheck prior-publication & salami-slicing risk across 8 sources.Citation GeneratorFreeA DOI or title → APA, MLA, Vancouver, Chicago, BibTeX, RIS. No signup.Graphical Abstract MakerFreeTurn your findings into a colorblind-safe graphical abstract.

Compare the category — Honest roundups of the AI research tools, ours listed alongside the alternatives.

Command palette

Jump anywhere, run any action.