We propose a recurrent neural model that generates natural-language questions from documents, conditioned on answers. We show how to train the model using a combination of supervised and reinforcement learning. After teacher forcing for standard maximum likelihood training, we fine-tune the model using policy gradient techniques to maximize several rewards that measure question quality. Most no...

The depth and timing of source text understanding during translation production is controversial in translation process research. Two theories seem to compete: defenders of the deep/alternating school assume that translators proceed in cycles of comprehension-transferproduction, while other translation scholars suggest that translations may be produced in a fashion of shallow and parallel compr...

This study examined the effects of three types of reading interventions on the science text reading of secondary students with learning disabilities (LD). Twenty 10th-grade students with LD participated in the study. Using a within-subjects design, the relative effects of three different instructional approaches—text reading, vocabulary learning, and text reading plus vocabulary learning—were e...

This study aimed to investigate the influence of background knowledge and cultural familiarity on reading comprehension and vocabulary inference of Turkish 7th grade students in public primary schools in Turkey. For this purpose; two texts, one of which was based on the authentic culture and the other one based on the nativized version, a vocabulary test and parallel reading comprehension tests...

The full meaning of a word takes years to be learned, and we want to find the minimum that needs to be learned in order to efficiently read a text. We used Latent Semantic Analysis (LSA), Neural Networks (NN), and Semantic graphs to model word acquisition and reading comprehension. We modeled word learning using a corpus -Wikipediato train the NN and LSA; in the case of semantic graphs, each ti...

the purpose of this study was to investigate two different techniques for testing reading comprehension questions. differences are measured by two different types of questions including inferential and referential questions. to answer the referential questions, the reader is required to move beyond the level of sentence comprehension and understand the relationship among the sentences. while, i...