Black-box machine translation systems have proven incredibly useful for a variety of applications yet by design are hard to adapt, tune to a specific domain, or build on top of. In this work, we introduce a method to improve such systems via automatic pre-processing (APP) using sentence simplification. We first propose a method to automatically generate a large in-domain paraphrase corpus through back-translation with a black-box MT system, which is used to train a paraphrase model that “simplifies” the original sentence to be more conducive for translation. The model is used to preprocess source sentences of multiple low-resource language pairs. We show that this preprocessing leads to better translation performance as compared to non-preprocessed source sentences. We further perform side-by-side human evaluation to verify that translations of the simplified sentences are better than the original ones. Finally, we provide some guidance on recommended language pairs for generating the simplification model corpora by investigating the relationship between ease of translation of a language pair (as measured by BLEU) and quality of the resulting simplification model from back-translations of this language pair (as measured by SARI), and tie this into the downstream task of low-resource translation.
It has been reported that retinoic acid (RA) may inhibit the growth of RPE and be used in the treatment of proliferative vitreoretinopathy (PVR). However, previous reports in this field have been conflicting. The main reason for these contradictory findings is that different methods for evaluating the effects of RA on RPE from different species have been used. In human specimens, only RPE from the donor eye (stationary) but not RPE from the PVR membrane (already at active proliferation status) have been tested. This study tested the effects of RA on the growth of RPE using a novel in vitro model: RPE from the PVR membranes, which simulates the in vivo situation of PVR patients better than RPE from the donor eyes. This study also used various methods to solve the conflicting results reported previously. We found that both all transretinoic acid (all-RA) and 13-cis-retinoic acid (cis-RA) can promptly (though not completely) inhibit proliferation of RPE (inhibition rate of 89%-90%) over a very wide range of concentrations (10(-9)-10(-5) M) and various lengths of periods (2-12 days) in a dose-dependent and time-dependent manner and without evident cytotoxic effects. Previously reported disadvantages discovered from the study of RPE from donor eyes, e.g., the absence of inhibitory effects on the early passages of cultured cells and inhibition occurring only after long-term treatment, do not present in RPE cells from the PVR membrane. The proliferation of RPE recovered from the inhibition by RA rapidly after the discontinuation of treatment, indicating that a continuous supply of the drug over a long period, i.e., until the breakdown of the blood-retinal barrier has been repaired, is essential for the success of drug treatment of PVR.
Identifying people in historical photographs is important for preserving material culture, correcting the historical record, and creating economic value, but it is also a complex and challenging task. In this article, we focus on identifying portraits of soldiers who participated in the American Civil War (1861--65), the first widely photographed conflict. Many thousands of these portraits survive, but only 10%--20% are identified. We created Photo Sleuth, a web-based platform that combines crowdsourced human expertise and automated face recognition to support Civil War portrait identification. Our mixed-methods evaluations of Photo Sleuth one month and 11 months after its public launch showed that it helped users successfully identify unknown portraits and provided a sustainable model for volunteer contribution. We also discuss implications for crowd-AI interaction and person identification pipelines.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.