Скачать или смотреть Atharva Kulkarni: SynthDST - Synthetic Data for Few-Shot Dialog State Tracking

#research #Conformal #Prediction

Atharva Kulkarni is a CS PhD student at USC NLP Group and DILL Lab. In this session we are discussing his paper "SynthDST: Synthetic Data is All You Need for Few-Shot Dialog State Tracking".


EPISODE LINKS:
Link to the paper: https://arxiv.org/abs/2402.02285
Atharva's Google scholar: https://scholar.google.com/citations?...
Atharva's LinkedIn:   / athrvkk  


OUTLINE:
00:00 - Intro
03:48 - Templates and schemas
06:28 - Data acquisition
08:53 - SynthDST overview
17:18 - Dialogue structure synthesis
22:23 - Template response generation
29:06 - Slot-value pairs & manual data creation
36:36 - Utterance paraphrasing
40:30 - Evaluation
49:10 - Results
51:21 - Concluding remarks


SOCIAL MEDIA LINKS:
  / forsstrom.  .
https://niklasopf.github.io/