README.md

---
license: cc-by-4.0
tags:
- Dataset
- Fillers
- Conversation-Analysis
---
This dataset documents filler word usage among non‑native English‑speaking university students in Hong Kong during career‑oriented interviews. 
It consists of one hour of audio recordings, transcriptions, and lexical‑level annotations highlighting sound, conversational, transitional, and emphasis fillers.
Data were collected through stratified sampling across diverse academic levels, processed with noise reduction, and annotated using a dual AI‑assisted and manual verification approach. 
Stored in WAV/MP3 and CSV formats, the dataset supports both linguistic and computational analysis. 
Its reuse potential spans discourse studies, NLP benchmarking, and pedagogical applications, offering a valuable resource for examining disfluency in professional communication.