Explore the vast range of titles available.

We propose two models for verbalizing numbers, a key component in speech recognition and synthesis systems. The first model uses an end-to-end recurrent neural network. The second model, drawing inspiration from the linguistics literature, uses finite-state transducers constructed with a minimal amount of training data. While both models achieve near-perfect performance, the latter model can be trained using several orders of magnitude less data than the former, making it particularly useful for low-resource languages.

Deficits in social communication, particularly pragmatic language, are characteristic of individuals with autism spectrum disorder (ASD). Speech disfluencies may serve pragmatic functions such as cueing speaking problems. Previous studies have found that speakers with ASD differ from typically developing (TD) speakers in the types and patterns of disfluencies they produce, but fail to provide sufficiently detailed characterizations of the methods used to categorize and quantify disfluency, making cross-study comparison difficult. In this study we propose a simple schema for classifying major disfluency types, and use this schema in an exploratory analysis of differences in disfluency rates and patterns among children with ASD compared to TD and language impaired (SLI) groups. 115 children ages 4-8 participated in the study (ASD = 51; SLI = 20; TD = 44), completing a battery of experimental tasks and assessments. Measures of morphological and syntactic complexity, as well as word and disfluency counts, were derived from transcripts of the Autism Diagnostic Observation Schedule (ADOS). High inter-annotator agreement was obtained with the use of the proposed schema. Analyses showed ASD children produced a higher ratio of content to filler disfluencies than TD children. Relative frequencies of repetitions, revisions, and false starts did not differ significantly between groups. TD children also produced more cued disfluencies than ASD children.

Background It remains unclear whether the opioid-sparing effects of dexmedetomidine seen in patients undergoing general anesthesia are reproducible in patients undergoing spinal anesthesia. We hypothesized that the administration of intravenous dexmedetomidine for sedation during total knee arthroplasty under spinal anesthesia would decrease postoperative morphine consumption in the first 24 hr following surgery. Methods We conducted this prospective double-blind randomized-controlled trial in 40 patients (American Society of Anesthesiologists physical status I-III) undergoing total knee arthroplasty with a standardized spinal anesthetic. Patients were randomized to receive either a dexmedetomidine loading dose of 0.5 µg·kg −1 over ten minutes, followed by an infusion of 0.5 µg·kg·hr −1 for the duration of the surgery, or a normal saline loading dose and an infusion of an equivalent volume. The primary outcome was the consumption of morphine delivered via patient-controlled analgesia in the first 24 hr following surgery. Results The mean (SD) cumulative morphine at 24 hr in the dexmedetomidine group was 29.2 (11.2) mg compared with 61.2 (17.2) mg in the placebo group (mean difference, 32.0 mg; 95% confidence interval, 22.7 to 41.2; P < 0.001). In the dexmedetomidine group, there was a delay in the time to first analgesic request ( P = 0.003) and a reduction in the mean morphine use at six and 12 hr following surgery (both P < 0.001). Conclusions Dexmedetomidine was associated with a significant decrease in morphine use in the first 24 hr following total knee arthroplasty. Our study shows that an intraoperative infusion of dexmedetomidine for sedation in patients receiving spinal anesthesia can produce postoperative analgesic effects. This offers another potential adjunct in the multimodal pain management of these patients. This trial was registered at ClinicalTrials.gov (identifier NCT02026141).

This study was intended to evaluate a series of algorithms developed to perform automatic classification of paraphasic errors (formal, semantic, mixed, neologistic, and unrelated errors). We analyzed 7,111 paraphasias from the Moss Aphasia Psycholinguistics Project Database (Mirman et al., 2010) and evaluated the classification accuracy of 3 automated tools. First, we used frequency norms from the SUBTLEXus database (Brysbaert & New, 2009) to differentiate nonword errors and real-word productions. Then we implemented a phonological-similarity algorithm to identify phonologically related real-word errors. Last, we assessed the performance of a semantic-similarity criterion that was based on word2vec (Mikolov, Yih, & Zweig, 2013). Overall, the algorithmic classification replicated human scoring for the major categories of paraphasias studied with high accuracy. The tool that was based on the SUBTLEXus frequency norms was more than 97% accurate in making lexicality judgments. The phonological-similarity criterion was approximately 91% accurate, and the overall classification accuracy of the semantic classifier ranged from 86% to 90%. Overall, the results highlight the potential of tools from the field of natural language processing for the development of highly reliable, cost-effective diagnostic tools suitable for collecting high-quality measurement data for research and clinical purposes.

Regionalized trauma care has been widely implemented in the United States, with field triage by emergency medical services (EMS) playing an important role in identifying seriously injured patients for transport to major trauma centers. In this study we estimated hospital-level differences in the adjusted cost of acute care for injured patients transported by 94 EMS agencies to 122 hospitals in 7 regions, overall and by injury severity. Among 301,214 patients, the average adjusted per episode cost of care was $5,590 higher in a level 1 trauma center than in a nontrauma hospital. We found hospital-level differences in cost among patients with minor, moderate, and serious injuries. Of the 248,342 low-risk patients-those who did not meet field triage guidelines for transport to trauma centers-85,155 (34.3 percent) were still transported to major trauma centers, accounting for up to 40 percent of acute injury costs. Adhering to field triage guidelines that minimize the overtriage of low-risk injured patients to major trauma centers could save up to $136.7 million annually in the seven regions we studied. [PUBLICATION ABSTRACT]

Background A subgroup of young children with autism spectrum disorders (ASD) have significant language impairments (phonology, grammar, vocabulary), although such impairments are not considered to be core symptoms of and are not unique to ASD. Children with specific language impairment (SLI) display similar impairments in language. Given evidence for phenotypic and possibly etiologic overlap between SLI and ASD, it has been suggested that language-impaired children with ASD (ASD + language impairment, ALI) may be characterized as having both ASD and SLI. However, the extent to which the language phenotypes in SLI and ALI can be viewed as similar or different depends in part upon the age of the individuals studied. The purpose of the current study is to examine differences in memory abilities, specifically those that are key “markers” of heritable SLI, among young school-age children with SLI, ALI, and ALN (ASD + language normal). Methods In this cross-sectional study, three groups of children between ages 5 and 8 years participated: SLI ( n =  18), ALI ( n =  22), and ALN ( n =  20). A battery of cognitive, language, and ASD assessments was administered as well as a nonword repetition (NWR) test and measures of verbal memory, visual memory, and processing speed. Results NWR difficulties were more severe in SLI than in ALI, with the largest effect sizes in response to nonwords with the shortest syllable lengths. Among children with ASD, NWR difficulties were not associated with the presence of impairments in multiple ASD domains, as reported previously. Verbal memory difficulties were present in both SLI and ALI groups relative to children with ALN. Performance on measures related to verbal but not visual memory or processing speed were significantly associated with the relative degree of language impairment in children with ASD, supporting the role of verbal memory difficulties in language impairments among early school-age children with ASD. Conclusions The primary difference between children with SLI and ALI was in NWR performance, particularly in repeating two- and three-syllable nonwords, suggesting that shared difficulties in early language learning found in previous studies do not necessarily reflect the same underlying mechanisms.

This dissertation outlines a program for the theory of phonotactics—the theory of speakers' knowledge of possible and impossible (or likely and unlikely) words—and argues that the alternative view of phonotactics as stochastic, and of phonotactic learning as probabilistic inference, is not capable of accounting for the facts of this domain. Chapter 1 outlines the proposal, precursors, and predictions. Chapter 2 considers evidence from wordlikeness rating tasks. It is argued that intermediate well-formedness ratings are obtained whether or not the categories in question are graded. A primitive categorical model of wordlikeness using prosodic representations is outlined, and shown to predict English speakers' wordlikeness judgements as accurately as state-of-the-art gradient wellformedness models. Once categorical effects are controlled for, gradient models are uncorrelated with well-formedness ratings. Chapter 3 considers the relationship between lexical generalizations, phonological alternations, and speakers' nonce word judgements with a focus on Turkish vowel patterns. It is shown that even exception-filled phonological generalizations have a robust effect on wellformedness judgements, but that statistically reliable phonotactic generalizations may go unlearned when they are not derived from phonological alternations. Chapter 4 investigates the role of phonological alternations in constraining lexical entries, focusing specifically on medial consonant clusters in English. Static phonotactic constraints previously proposed to describe gaps in the inventory of medial clusters are shown to be statistically unsound, whereas phonological alternations impose robust restrictions on the cluster inventory. The remaining gaps in the cluster inventory are attributed to the sparse nature of the lexicon, not static phonotactic restrictions. Chapter 5 summarizes the findings, considers their relation to order of acquisition, and proposes directions for future research.

The common practice of preprocessing text before feeding it into NLP models introduces many decision points which have unintended consequences on model performance. In this opinion piece, we focus on the handling of diacritics in texts originating in many languages and scripts. We demonstrate, through several case studies, the adverse effects of inconsistent encoding of diacritized characters and of removing diacritics altogether. We call on the community to adopt simple but necessary steps across all models and toolkits in order to improve handling of diacritized text and, by extension, increase equity in multilingual NLP.

The single shortest path algorithm is undefined for weighted finite-state automata over non-idempotent semirings because such semirings do not guarantee the existence of a shortest path. However, in non-idempotent semirings admitting an order satisfying a monotonicity condition (such as the plus-times or log semirings), the notion of shortest string is well-defined. We describe an algorithm which finds the shortest string for a weighted non-deterministic automaton over such semirings using the backwards shortest distance of an equivalent deterministic automaton (DFA) as a heuristic for A* search performed over a companion idempotent semiring, which is proven to return the shortest string. While there may be exponentially more states in the DFA, this algorithm needs to visit only a small fraction of them if determinization is performed \"on the fly\".

Student reviews often make reference to professors' physical appearances. Until recently RateMyProfessors.com, the website of this study's focus, used a design feature to encourage a \"hot or not\" rating of college professors. In the wake of recent #MeToo and #TimesUp movements, social awareness of the inappropriateness of these reviews has grown; however, objectifying comments remain and continue to be posted in this online context. We describe two supervised text classifiers for detecting objectifying commentary in professor reviews. We then ensemble these classifiers and use the resulting model to track objectifying commentary at scale. We measure correlations between objectifying commentary, changes to the review website interface, and teacher gender across a ten-year period.