JSAI2020
Jun 9 - Jun 12, 2020Virtual Meetings
JSAI2020
Jun 9 - Jun 12, 2020Virtual Meetings
[1D5-GS-9-04]Unsupervised Double Articulation and Meaning Acquisition of Natural Speech in a Video Game Environment
〇Kotaro Yamaguchi1, Natsuki Oka1, Tadahiro Taniguchi2, Ryo Ozaki2(1. Kyoto Institute of Technology, 2. Ritsumeikan University)
Keywords:
Unsupervised double articulation,Language acquisition,Reinforcement learning
Infants need to segment their native language into phonemes and words at the same time without supervision. Taniguchi, Nagasaka, & Nakashima (2016) showed that Nonparametric Bayesian Double Articulation Analyzer could analyze latent double articulation structure, i.e., hierarchically organized latent words and phonemes, of utterance data consisting of a limited vocabulary in an unsupervised manner by assuming hierarchical Dirichlet process hidden language model (HDP-HLM). In this study, we attempted unsupervised double articulation analysis of natural speech in a video game environment and tried to give meaning to the segmented words. The result of an experiment demonstrated that the utterances were roughly correctly segmented, and the meanings of up, down, left and right were almost correctly learned.