VLSP 2025 Shared Task: Vietnamese Temporal Question Answering (TemporalQA)

Abstract

In 2025, the eleventh edition of the Vietnamese Language and Speech Processing workshop (VLSP 2025) launched its first shared task on Vietnamese Temporal Question Answering (TemporalQA). This task aims to benchmark systems in their ability to comprehend and perform reasoning
over temporal information expressed in Vietnamese. TemporalQA is composed of two subtasks. The
first, Date Arithmetic (date-arith), evaluates models on questions involving temporal computation,
such as adding or subtracting time intervals from a reference date, requiring precise interpretation
of temporal expressions within a given context. The second, Duration Question Answering (durationQA), assesses systems on estimating the duration of events or activities based on contextual descriptions. By formalizing these two subtasks and releasing curated datasets, TemporalQA provides
the first dedicated benchmark for temporal reasoning in Vietnamese, filling an important resource
gap and supporting future research on temporal understanding and reasoning for Vietnamese NLP.
The benchmark was hosted on the AIHUB platform, where system rankings were determined using
a private test set following a public testing phase. Results show that the leading system for Subtask 1
reached an accuracy of 99%, while the best-performing system in Subtask 2 achieved an F1-score of
81.89% and an Exact Match score of 47.52%.
Keywords: Temporal Reasoning, Question Answering, TemporalQA, VLSP 2025, Vietnamese

02:31:05 dblclick "Temporal Reasoning, Question Answering, TemporalQA, VLSP 2025, Vietnamese" sidepanelOpen=false autoUpdate=true isAdv=true