This article describes the operation and the experimental evaluation of the pronunciation modeling component of the first Hungarian large vocabulary continuous speech recognition system. The proposed method is based on the implementation of context dependent rewrite rules by weighted finite state transducers (WFSTs). The proposed phonological model decreases the error rate by 8.32% relatively c...

One could assume that the negative morpheme no in (2) is not semantically negative, but this would be dissatisfying in several respects. First of all, in general we of course would prefer not to postulate an ambiguity, on fear of missing a generalization. Second, the question arises of why it is the negative morpheme which takes on the pleonastic function, if the meaning of this item is not imp...

Participants segmented and shifted a sequence of letters from a source word to a target word and then named the product aloud. Morphemic and nonmorphemic letter sequences (e.g., EN) from phonemically matched words such as HARDEN and GARDEN were compared. In 4 experiments, naming latencies were faster for morphemic sequences than their nonmorphemic controls in both English, in which the morphemi...

Event nouns can lexically encode eventive information. Recently these nouns have generated considerable scholarly interest. However, little research has been conducted in their morphological and syntactic structure, qualia modification, event representing feature, and information inheritance characteristics. This study has these main findings. 1) Morphologically, the modifier and the head is ei...

Connecting automatic speech recognizers (ASRs) and language analyzers is difficult since they may be based on differences in part-of-speech (POS) systems; the latter cannot directly analyze the outputs of the former. In addition, in unsegmented languages such as Japanese, the ASR outputs are likely to have different word segmentation from that of the language analyzer inputs because they are in...

We investigate different representational granularities for sub-lexical representation in statistical machine translation work from English to Turkish. We find that (i) representing both Turkish and English at the morpheme-level but with some selective morpheme-grouping on the Turkish side of the training data, (ii) augmenting the training data with “sentences” comprising only the content words...