Summary of Predicting User Intents and Musical Attributes From Music Discovery Conversations, by Daeyong Kwon et al.
Predicting User Intents and Musical Attributes from Music Discovery Conversations
by Daeyong Kwon, SeungHeon Doh, Juhan Nam
First submitted to arxiv on: 19 Nov 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Machine Learning (cs.LG); Sound (cs.SD); Audio and Speech Processing (eess.AS)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper investigates intent classification models for music discovery conversations, focusing on pre-trained language models. It proposes a method to concatenate previous chat history with single-turn user queries to improve the understanding of conversation context. The authors also introduce musical attribute classification as a task in addition to predicting functional needs. They achieve significant improvements in F1 scores for both tasks and surpass the performance of the pre-trained Llama 3 model. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper studies how computers can understand what people want when they ask for music recommendations. It looks at special language models that are already trained on a lot of text data. The authors make these models better by adding information about previous conversations, which helps them understand the bigger context. They also test the models’ ability to classify not just what people want but also what kind of music they like. The results show that their approach works well and is even better than some other already-trained models. |
Keywords
» Artificial intelligence » Classification » Llama
