Syn-TurnTurk: A Synthetic Dataset for Turn-Taking Prediction in Turkish Dialogues

arXiv cs.CL / 4/16/2026

📰 NewsSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper addresses turn-taking timing in Turkish voice chatbots, noting that relying on silence detection often leads to bot interruptions due to irregular human pauses.
It introduces Syn-TurnTurk, a synthetic Turkish dialogue dataset generated with multiple Qwen LLMs to better reflect real exchanges, including overlaps and deliberate silences.
The study evaluates both traditional and deep learning approaches for turn-taking prediction, reporting strong performance with BI-LSTM and an Ensemble (LR+RF) setup (accuracy 0.839, AUC 0.910).
The authors argue the dataset can improve models’ ability to detect linguistic cues, which may lead to more natural human-machine interaction in Turkish.
The work highlights a data-quality gap for Turkish turn-taking prediction and positions synthetic data as a practical route to address it for future research and model development.

Abstract

Managing natural dialogue timing is a significant challenge for voice-based chatbots. Most current systems usually rely on simple silence detection, which often fails because human speech patterns involve irregular pauses. This causes bots to interrupt users, breaking the conversational flow. This problem is even more severe for languages like Turkish, which lack high-quality datasets for turn-taking prediction. This paper introduces Syn-TurnTurk, a synthetic Turkish dialogue dataset generated using various Qwen Large Language Models (LLMs) to mirror real-life verbal exchanges, including overlaps and strategic silences. We evaluated the dataset using several traditional and deep learning architectures. The results show that advanced models, particularly BI-LSTM and Ensemble (LR+RF) methods, achieve high accuracy (0.839) and AUC scores (0.910). These findings demonstrate that our synthetic dataset can have a positive affect for models understand linguistic cues, allowing for more natural human-machine interaction in Turkish.