Search

This work presents an information-theoretic operationalisation of cross-linguistic non-arbitrariness. It is not a new idea that there are small, cross-linguistic associations between the forms and meanings of words. For instance, it has been claimed (Blasi et al., 2016) that the word for “tongue” is more likely than chance to contain the phone [l]. By controlling for the influence of language family and geographic proximity within a very large concept-aligned, cross-lingual lexicon, we extend methods previously used to detect within language non-arbitrariness (Pimentel et al., 2019) to measure cross-linguistic associations. We find that there is a significant effect of non-arbitrariness, but it is unsurprisingly small (less than 0.5% on average according to our information-theoretic estimate). We also provide a concept-level analysis which shows that a quarter of the concepts considered in our work exhibit a significant level of cross-linguistic non-arbitrariness. In sum, the paper provides new methods to detect cross-linguistic associations at scale, and confirms their effects are minor., Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, ISBN:978-1-954085-46-6

Multilingual Automated Speech Recognition (ASR) systems allow for the joint training of data-rich and data-scarce languages in a single model. This enables data and parameter sharing across languages, which is especially beneficial for the data-scarce languages. However, most state-of-the-art multilingual models require the encoding of language information and therefore are not as flexible or scalable when expanding to newer languages. Language-independent multilingual models help to address this issue, and are also better suited for multicultural societies where several languages are frequently used together (but often rendered with different writing systems). In this paper, we propose a new approach to building a language-agnostic multilingual ASR system which transforms all languages to one writing system through a many-to-one transliteration transducer. Thus, similar sounding acoustics are mapped to a single, canonical target sequence of graphemes, effectively separating the modeling and rendering problems. We show with four Indic languages, namely, Hindi, Bengali, Tamil and Kannada, that the language-agnostic multilingual model achieves up to 10% relative reduction in Word Error Rate (WER) over a language-dependent multilingual model.

We present methods for calculating a measure of phonotactic complexity—bits per phoneme— that permits a straightforward cross-linguistic comparison. When given a word, represented as a sequence of phonemic segments such as symbols in the international phonetic alphabet, and a statistical model trained on a sample of word types from the language, we can approximately measure bits per phoneme using the negative log-probability of that word under the model. This simple measure allows us to compare the entropy across languages, giving insight into how complex a language’s phonotactics is. Using a collection of 1016 basic concept words across 106 languages, we demonstrate a very strong negative correlation of − 0.74 between bits per phoneme and the average length of words., Transactions of the Association for Computational Linguistics, 8, ISSN:2307-387X

A longstanding debate in semiotics centers on the relationship between linguistic signs and their corresponding semantics: is there an arbitrary relationship between a word form and its meaning, or does some systematic phenomenon pervade? For instance, does the character bigram \textit{gl} have any systematic relationship to the meaning of words like \textit{glisten}, \textit{gleam} and \textit{glow}? In this work, we offer a holistic quantification of the systematicity of the sign using mutual information and recurrent neural networks. We employ these in a data-driven and massively multilingual approach to the question, examining 106 languages. We find a statistically significant reduction in entropy when modeling a word form conditioned on its semantic representation. Encouragingly, we also recover well-attested English examples of systematic affixes. We conclude with the meta-point: Our approximate effect size (measured in bits) is quite small---despite some amount of systematicity between form and meaning, an arbitrary relationship and its resulting benefits dominate human language., Accepted for publication at ACL 2019

A simulation framework could decrease the burden of attending long and tiring experimental sessions on the potential users of brain-computer interface (BCI) systems. Specifically during the initial design of a BCI, a simulation framework that could replicate the operational performance of the system would be a useful tool for designers to make design choices. In this manuscript, we develop a Monte Carlo-based probabilistic simulation framework for electroencephalography (EEG) based BCI design. We employ one event-related potential (ERP) based typing and one steady-state evoked potential (SSVEP) based control interface as testbeds. We compare the results of simulations with real-time experiments. Even though over- and underestimation of the performance is possible, the statistical results over the Monte Carlo simulations show that the developed framework generally provides a good approximation of the real-time system performance.

Current scanning access methods for text generation in AAC devices are limited to relatively few options, most notably row/column variations within a matrix. We present Huffman scanning, a new method for applying statistical language models to binary-switch, static-grid typing AAC interfaces, and compare it to other scanning options under a variety of conditions. We present results for 16 adults without disabilities and one 36-year-old man with locked-in syndrome who presents with complex communication needs and uses AAC scanning devices for writing. Huffman scanning with a statistical language model yielded signifi cant typing speedups for the 16 participants without disabilities versus any of the other methods tested, including two row/column scanning methods. A similar pattern of results was found with the individual with locked-in syndrome. Interestingly, faster typing speeds were obtained with Huffman scanning using a more leisurely scan rate than relatively fast individually calibrated scan rates. Overall, the results reported here demonstrate great promise for the usability of Huffman scanning as a faster alternative to row/column scanning.

Discriminative language modeling is a structured classification problem. Log-linear models have been previously used to address this problem. In this paper, the standard dot-product feature representation used in log-linear models is replaced by a non-linear function parameterized by a neural network. Embeddings are learned for each word and features are extracted automatically through the use of convolutional layers. Experimental results show that as a stand-alone model the continuous space model yields significantly lower word error rate (1% absolute), while having a much more compact parameterization (60%-90% smaller). If the baseline scores are combined, our approach performs equally well.

This paper examines efficient predictive broad-coverage parsing without dynamic programming. In contrast to bottom-up methods, depth-first top-down parsing produces partial parses that are fully connected trees spanning the entire left context, from which any kind of non-local dependency or partial semantic interpretation can in principle be read. We contrast two predictive parsing approaches, top-down and left-corner parsing, and find both to be viable. In addition, we find that enhancement with non-local information not only improves parser accuracy, but also substantially improves the search efficiency., 8 pages, 3 tables, 3 figures