Metraipolis › archive › dc_ec88bc406516

DecisionTraceBench: Tracking Decision-Critical Information in Multi-Party Spoken Conversations — arXiv:2610.34746

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.34746
License (as archived)
CC0-1.0
Archived
2026-10-27T08:11:38.152Z
SHA-256 of the archived text
c1382d71d7a60d953eaab05820ed3b83b0a082ccc3335792f03bd4fadfd4edaf
Stable URL of this page
https://metraipolis.com/documento/dc_ec88bc406516

How to cite

DecisionTraceBench: Tracking Decision-Critical Information in Multi-Party Spoken Conversations — arXiv:2610.34746 — arXiv (en), https://arxiv.org/abs/2610.34746 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ec88bc406516 (Metraipolis archive dc_ec88bc406516, sha256 c1382d71d7a60d953eaab05820ed3b83b0a082ccc3335792f03bd4fadfd4edaf)
BibTeX
@misc{dcec88bc406516,
  title = {DecisionTraceBench: Tracking Decision-Critical Information in Multi-Party Spoken Conversations — arXiv:2610.34746},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.34746},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ec88bc406516; sha256 c1382d71d7a60d953eaab05820ed3b83b0a082ccc3335792f03bd4fadfd4edaf},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.34746. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

In high-stakes domains such as finance, critical information evolves through multi-party interaction. Recovering these changes and their sources is fundamental to reliable speech understanding. Existing speech benchmarks focus on transcription, local understanding, and turn-level state prediction, leaving joint decision evolution across speakers insufficiently evaluated. We introduce DecisionTraceBench, a speech state-tracking benchmark for Decision-Critical Information State Tracking (DCIST). It jointly evaluates decision event sequences, speaker attribution, scope, and temporal evidence, using trajectory-level F1 to measure complete event-chain recovery. Our dataset comprises Chinese financial dialogues involving two to four speakers, with 100.39 hours of synthetic audio and 8.61 hours of human-performed recordings. On a fixed cohort of human-performed recordings, five cascaded systems and two direct-audio models achieve DCIST scores of 5.73% to 21.35%. Ablations with speaker-labeled reference scripts and no temporal localization requirements yield a maximum score of only 38.72%. Audio perturbation experiments further show that overall transcription error rates do not adequately reflect complete trajectory recovery. These findings reveal limitations in jointly understanding information provenance and state evolution, providing a benchmark for evaluating decision-process understanding and decision traceability in high-stakes speech applications.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.34746 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-27T07:56:00.235Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).