Reasoning about RoboCup Soccer Narratives
This addresses the challenge of narrative understanding in dynamic systems like RoboCup soccer, offering an unsupervised method that could reduce reliance on labeled data.
The paper tackles the problem of translating narratives into event sequences without labeled data, using domain knowledge of event preconditions and effects, and demonstrates that it outperforms state-of-the-art supervised systems in reconstructing RoboCup soccer games from commentaries.
This paper presents an approach for learning to translate simple narratives, i.e., texts (sequences of sentences) describing dynamic systems, into coherent sequences of events without the need for labeled training data. Our approach incorporates domain knowledge in the form of preconditions and effects of events, and we show that it outperforms state-of-the-art supervised learning systems on the task of reconstructing RoboCup soccer games from their commentaries.