Tag
This paper introduces Holographic Neural PCFG (Hol-PCFG), which recasts probabilistic context-free grammar rule scoring using holographic embeddings, achieving state-of-the-art unsupervised parsing performance across six languages with a 99.94% reduction in parameters.
This paper investigates whether dependency parsing of non-human primate vocalizations or gestures can be evaluated without a gold standard. Using network science, the authors show that the proportion of correct edges retrieved by a parser is necessarily high due to the fast decay of sequence length distributions in non-human primates, making evaluation feasible, unlike for human language.