Paper Category: User Experience (UX) of CRS
Navigating User Experience of ChatGPT-based Conversational Recommender Systems: The Effects of Prompt Guidance and Recommendation Domain
Conversational recommender systems (CRS) enable users to articulate preferences through natural language. With large language models (LLMs) like ChatGPT, enhancing user engagement and recommendation quality is a key focus. This study investigates how prompt guidance (PG) and recommendation domain (RD) impact the overall user experience (UX) of ChatGPT-based CRS. An online empirical study (N=100) revealed that PG significantly boosts system explainability, adaptability, perceived ease of use, and transparency. Users showed greater novelty perception and willingness to try recommendations in book domains compared to job domains. Interaction effects suggest PG's influence is modulated by the recommendation domain. This work offers crucial user-centered evaluation and practical design guidance for ChatGPT-based CRS.
Executive Impact at a Glance
Key findings reveal how strategic prompt guidance and domain understanding can significantly elevate user engagement and system performance in AI-driven recommendation platforms.
0
Participants Evaluated
0
UX Improvement with Prompt Guidance
0
Higher Novelty in Book Recs
0
Increased Intent to Try (Low-Stake)
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
Enhanced UX with Prompt Guidance
15% Improvement in Explainability, Adaptability, Ease of Use, and TransparencyPrompt guidance significantly enhances key UX dimensions such as explainability, adaptability, perceived ease of use, and transparency. Users in the withPG group reported higher mean scores across these metrics, indicating a clearer understanding and easier interaction with the ChatGPT-based CRS.
Decision-Making Flow in CRS
User Input (PG/No PG)
→
ChatGPT Processing
→
Recommendation Generation
→
User Feedback/Refinement
→
Decision/Trial (RD influence)
The interaction flow within a ChatGPT-based CRS is dynamic. It begins with user input, guided or unguided, leading to ChatGPT processing and recommendation generation. User feedback then refines the recommendations, with the ultimate decision-making and trial influenced by the recommendation domain (high vs. low stakes).
| Feature | Book Recommendations (Low Stake) | Job Recommendations (High Stake) |
|---|---|---|
| Perceived Novelty |
|
|
| Intention to Try |
|
|
| Risk Perception |
|
|
| Decision Speed |
|
|
User engagement and perceived novelty vary significantly between recommendation domains, indicating a need for domain-specific design adaptations.
Tailoring Guidance for User Expertise
Scenario: A key finding is that the benefits of prompt guidance are particularly pronounced for users with less prior experience with recommender systems (RS novices). For these users, PG significantly improved perceived explainability, ease of use, and transparency of the ChatGPT-based CRS.
Impact: This suggests that effective CRS design should adapt its guidance level based on user familiarity. Novices benefit from structured prompts, while experienced users may prefer less restrictive interactions, fostering greater autonomy.
Outcome: Designing adaptive prompt guidance based on user experience levels can optimize system utility and satisfaction across diverse user groups, enhancing overall UX and system adoption.
Calculate Your Potential AI ROI
Estimate the efficiency gains and cost savings your enterprise could achieve by implementing intelligent AI solutions.
Estimated Annual Savings
$0
Annual Hours Reclaimed
0
Your AI Implementation Roadmap
A phased approach to integrate conversational AI seamlessly into your operations.
Phase 1: Discovery & Strategy
Conduct a thorough assessment of your current systems and identify key areas where ChatGPT-based CRS can drive the most value. Define clear objectives and success metrics for your AI initiatives.
Phase 2: Pilot Development & Testing
Develop a pilot ChatGPT-based CRS tailored to a specific domain (e.g., internal knowledge base, customer support). Implement prompt guidance strategies and rigorously test user experience with a controlled group.
Phase 3: Iterative Refinement & Expansion
Based on pilot feedback, refine the AI model, prompt engineering, and user interface. Gradually expand the CRS to additional domains, continuously monitoring performance and user satisfaction.
Phase 4: Full-Scale Deployment & Optimization
Roll out the refined ChatGPT-based CRS across your enterprise. Establish ongoing monitoring, maintenance, and optimization processes to ensure long-term effectiveness and adapt to evolving user needs and AI capabilities.
Ready to Transform Your Enterprise with AI?
Unlock the full potential of conversational AI for enhanced user experience and operational efficiency. Let's build a smarter future, together.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat