Abstract Pre-trained language models achieve high performance on machine reading comprehension task, but these lack robustness and are vulnerable to adversarial samples. Most of the current methods for improving model based data enrichment. However, do not solve problem poor context representation model. We find that plays a key role in model, dense space results robustness. To deal with this, ...

We bring together knowledge from two different types of language learning data, texts learners read and texts they write, to improve linguistic complexity classification in the latter. Linguistic complexity in the foreign and second language learning context can be expressed in terms of proficiency levels. We show that incorporating features capturing lexical complexity information from reading...

Recently proposed Story Cloze Test [Mostafazadeh et al., 2016] is a commonsense machine comprehension application to deal with natural language understanding problem. This dataset contains a lot of story tests which require commonsense inference ability. Unfortunately, the training data is almost unsupervised where each context document followed with only one positive sentence that can be infer...

This study aimed to analyze the relationship between reading anxiety and comprehension of ESP students in Universitas Muhammadiyah Malang. The participants this were 50 first-year Governmental Science who took intensive I. used correlational design two variables (Reading Reading comprehension). In collecting data, English Foreign Language Anxiety Inventory (EFLRAI) questionnaire was investigate...

Reading time is commonly used as a basis for the study of reading. It has been observed in previous studies that unexpected sentences initiating a new context in a body of text considerably increases its reading time. There are two general accounts of the mechanism behind this slowdown. The first views reading, both slow and fast, to be a single, continuous process. The second views reading as ...

The restoration of wooden artworks often requires high resolution 3D close-range elevation models. This work shows in that context some experience developed for the survey of Donatello’s wooden statue “La Maddalena”. A conventional close-range photogrammetric survey was carried out for the post elaboration of 3D high resolution cloud points data obtained by a fringe interferometric system devic...