Clues In Context Research Articles

Diversified generation of commonsense reasoning questions

This paper studies a new task of generating commonsense reasoning questions (QG) with diversity and controlled difficulty based on the given text. Unlike existing shallow questions, ours are more complex, in-depth, and thought-provoking. The answers need to be inferred from multiple clues in the text context. Some clues are implicit and require to be deduced from external commonsense knowledge. This process involves a series of comprehension skills, such as complex reasoning, hidden knowledge inference, etc. That can support many applications in the field of QA, online education, etc. Existing work mostly learns a binary mapping QG model, which is hard to control diversity and difficulty, leading to shallow results. Also, this model relies heavily on lots of scarce training data, which is unsuited to the low-resource scenario. To address these problems, we propose a new controllable framework to yield diverse in-depth results with low resources. In detail, we first build a graph to capture potential contextual and commonsense clues. To improve controllability, we utilize a reasoning skeleton with a desired difficulty level to retrieve a question-specific subgraph. It embodies all the necessary clues that can guide the asking direction well. The classical asking skeletons and expressive forms are learned from unlabeled data which is easy to obtain. The forms contain useful syntactic structures that help to organize questions well. By multiple sampling of the forms, we can yield diverse results. We optimize the results’ quality by multiple reinforced rewards, including context consistency, reasoning difficulty, and fluency. Moreover, we employ the generated results to facilitate the downstream application of fact-checking. Extensive experiments are conducted on two popular datasets. The results verify the effectiveness of our new approach, with 11% performance improvements on average.

The Epistemic Marker Určit Ě in the Light of Corpus Data

Abstract The paper presents a pilot study for a research project on epistemic modality and/or evidentiality markers in Czech. The study focuses on the expression určitě. Although, this marker is typically considered to signal high certainty, the dictionary of standard Czech (Slovník spisovné češtiny, SSČ) also offers an alternative meaning of probability, indicating a lower degree of certainty. We use parallel data from the InterCorp v15 corpus to determine whether the probability meaning can be identified unequivocally in real language data and whether it correlates with specific translation equivalents, linguistic features, or lexical context. Based on our findings, we propose an alternative method for distinguishing between different shades of meaning based on the communicative functions of the utterances, and we draw conclusions regarding the relevance of individual grammatical and lexical clues in context for future annotations.

Histopathology of intestinal villi in neonatal and paediatric age: main features with clinical correlation - Part I.

SummaryThe neonatal and paediatric spectrum of small bowel disorders encompass a wide variety of conditions, ranging from food allergies to life-threatening surgical emergencies or life-long medical conditions and, as such, it comes with a whole set of diagnostic challenges for the non-paediatric pathologist. Histologic examination is a cornerstone of diagnosis in a large number of diseases and may still provide important diagnostic clues in the appropriate clinical context. In this review, divided in two sections, we aim to provide a comprehensive histopathological summary of paediatric small bowel alteration and their differential diagnoses with a reference to the main clinical aspects required for appropriate interpretation. Specifically, in this first part, we describe congenital and metabolic disorders, intestinal lymphangiectasia, immunodeficiencies, GVHD, and necrotising enterocolitis.

The Influence of Visible Clues on Non-visible Clues in Context of Franchise Family Restaurants

The Influence of Visible Clues on Non-visible Clues in Context of Franchise Family Restaurants

The Influence of Visible Clues on Non-visible Clues in Context of Franchise Family Restaurants

The Influence of Visible Clues on Non-visible Clues in Context of Franchise Family Restaurants

Uma análise funcionalista da indeterminação do sujeito no Português Popular falado em São Paulo

Using the theoretical and methodological tools of Functional Linguistics, this paper presents a quantitative analysis of pronominal and verbal forms expressing subject indeterminacy in Popular Portuguese spoken in São Paulo. Data consist of 23 interviews which are part of the Popular Portuguese Project in São Paulo. This data correspond to recordings of speeches of adults illiterate or that have only few years of schooling, immigrants as well as people who were born in São Paulo, who live in slams located in the capital. The analysis demonstrates that (i) there are many forms that express subject indeterminacy in Popular Portuguese; (ii) considering that indeterminacy is a matter of degree, there are forms that are more indeterminate than others which are accompanied by referential clues in text or context; (iii) these forms can have different functions according to the context in which they appear; (iv) they can refer from one to three grammatical persons when they reach a high degree of generalization.

Vocabulary Expansion in the TESOL Classroom

The precise when of massive vocabulary expansion depends on the learer's maturity. It follows the phase during which the goal is reliable control of basic pronunciation and grammatical habits. Overlapping that phase and becoming dominant in the intermediate stage is the development of skills of comprehension and the build-up of vocabulary resources. At first skills are more important than resources for the learner. The fact that words have multiple meanings, and the huge proportion of low-frequency words in any text, make it impossible to anticipate the specific vocabulary needs of a FL learner. Learners must acquire courage and skill; they must be helped to avoid panic and be willing to guess at the meanings of unfamiliar words. Developing this skill calls for using-ultimately using unconsciously-the clues in a total context, both the grammatical and the pragmatic clues. This includes practice in skimming as a necessary part of an intermediate FL program. All experienced readers skim a great deal of what they read, and almost all listening is with less than 100% attention. To neglect skimming is to leave a major gap in FL instruction. Skimming skill is not only in itself an indispensable part of language competence; pedagogically it is the way to build up habits of guessing at meanings and realistically tolerating a certain amount of vagueness. It is from extensive contact with the FL and skill in intelligent use of clues in context, not from word-matching or laborious pseudo-translation, that vocabulary expands.

Clues in the Context: A reading project on new words

Clues in the Context<i>: A reading project on new words</i>