Search

AbstractIdiomatic expressions (IEs), characterized by their non-compositionality, are an important part of natural language. They have been a classical challenge to NLP, including pre-trained language models that drive today’s state-of-the-art. Prior work has identified deficiencies in their contextualized representation stemming from the underlying compositional paradigm of representation. In this work, we take a first-principles approach to build idiomaticity into BART using an adapter as a lightweight non-compositional language expert trained on idiomatic sentences. The improved capability over baselines (e.g., BART) is seen via intrinsic and extrinsic methods, where idiom embeddings score 0.19 points higher in homogeneity score for embedding clustering, and up to 25% higher sequence accuracy on the idiom processing tasks of IE sense disambiguation and span detection.

Porosity in coatings is a vital parameter for many engineering applications. Having a porous coating is good with respect to few aspects and bad with respect to others. One of the main applications of porous metal coating is in the field of membrane filters for oil water separation to deal with industrial waste water and off-shore oil spill. It is also used for electrolytic cleaning (purification) of water. Studying the nature of formation and development of pores is being carried out by many researchers to determine the parameters affecting the porosity of coatings. Optimizing the parameters to obtain a favourable coating with least defect is the idea behind these research topics. Nickel and chromium coatings are widely used for many industrial applications because of specific properties. Nickel-chromium alloy coatings offer better hardness, wear and corrosion resistance compared to individual nickel or chromium coatings. The alloy coatings are observed to be porous in nature compared to individual nickel and chromium coatings. Controlling the porosity of the deposition by varying the process parameters is a very interesting topic of research as it greatly influences the properties. In the present work, the coating microstructures of nickel-chromium alloys developed by electrodeposition techniques are analysed and evaluated using Field Emission Scanning Electron Microscopy (FESEM) for varying current densities and plating time resulting a good uniform coating with varying porosity

AbstractIdiomatic expressions are an integral part of natural language and constantly being added to a language. Owing to their non-compositionality and their ability to take on a figurative or literal meaning depending on the sentential context, they have been a classical challenge for NLP systems. To address this challenge, we study the task of detecting whether a sentence has an idiomatic expression and localizing it when it occurs in a figurative sense. Prior research for this task has studied specific classes of idiomatic expressions offering limited views of their generalizability to new idioms. We propose a multi-stage neural architecture with attention flow as a solution. The network effectively fuses contextual and lexical information at different levels using word and sub-word representations. Empirical evaluations on three of the largest benchmark datasets with idiomatic expressions of varied syntactic patterns and degrees of non-compositionality show that our proposed model achieves new state-of-the-art results. A salient feature of the model is its ability to identify idioms unseen during training with gains from 1.4% to 30.8% over competitive baselines on the largest dataset.

Infantile myofibromatosis, the most common fibrous tumor of infancy, is classified in 2 forms; as a solitary nodule or as numerous, widely-distributed multicentric lesions with or without visceral involvement. Although benign, multicentric myofibromas are still associated with a high incidence of morbidity and mortality due to the infiltration of critical structures. Herein, we present a case of an infant with aggressive PDGFRB and NOTCH3 mutation-negative myofibromas distributed throughout the vascular, respiratory, and gastrointestinal systems. The extensive disease resulted in pulmonary hypertension, respiratory failure and gastrointestinal obstruction refractory to chemotherapy and unamenable to surgical resection. Despite the presence of numerous highly invasive myofibromas, multiple imaging modalities largely underestimated, or even missed, tumors found at autopsy. This case demonstrates the limitations of radiographic imaging to assess disease burden in multicentric infantile myofibromatosis. The postmortem findings of extensive disease far exceeding what was demonstrated by multiple imaging modalities suggests that pediatricians should have a high index of suspicion for undetected tumors if clinical deterioration is otherwise unexplained.

Discussion forums are important components of online courses because of the collaboration and community they foster, and the language used within the discussion forums may be influential in this development. In particular, studying the gendered language patterns of discussion forums can help gain insight into students' state of mind and propensity to form a community and thus ultimately may help explain men and women's differential success in online courses. This study describes the state of gendered language use in two online science courses. The results reveal that women and men do not differ in their language use along traditionally gendered lines, which holds promise for women in online courses. This means that the language that women use does not overtly mark them as female, which has the potential to help subvert the typical result of the negative outcomes associated with the female marker in science courses.

A direct and simple numerical method is presented for calculating the transfer function matrix of a linear time invariant multivariable system (A, B, C). The method is based on the matrix-determinant identity, and it involves operations with an auxiliary vector on the matrices. The method is computationally faster compared to Liverrier and Danilevsky methods.