Critiquing for Music Exploration in Conversational Recommender Systems
summary
A music chatbot with three critiquing variants is implemented and evaluated online (between-subject N=107) to test how user-initiated critiquing and two system-suggested strategies work for exploration. The system logs playlists of 20 songs and top‑5 choices; results show Cascading SC yields higher perceived diversity, Progressive SC yields higher serendipity, both UC and SC increase interactions, and Cascading SC uniquely turns more listening and turns into greater perceived helpfulness and serendipity.
AI-written summary · about
chapters (6)
related talks
22mPersonalized explanations for hybrid recommender systems
35mNatural and conversational shopping experiences
16mWhere can my career take me?: harnessing dialogue for interactive career goal recommendations
12mExplaining recommendations in an interactive hybrid social recommender
12mSynthia: Visually Interpreting and Synthesizing Feedback for Writing Revision
44mImagined Dragons: Building an Imagination-Powered Music Recommendation Platform
06m20 questions and AI chat bots
13mInterQuest: A Mixed-Initiative Framework for Dynamic User Interest Modeling in Conversational Search