Understanding idioms is important in NLP. In this paper, we study to what extent pre-trained BERT model can encode the meaning of a potentially idiomatic expression (PIE) in a certain context. We make use of a few existing datasets and perform two pr
obing tasks: PIE usage classification and idiom paraphrase identification. Our experiment results suggest that BERT indeed can separate the literal and idiomatic usages of a PIE with high accuracy. It is also able to encode the idiomatic meaning of a PIE to some extent.
The automatic recognition of idioms poses a challenging problem for NLP applications. Whereas native speakers can intuitively handle multiword expressions whose compositional meanings are hard to trace back to individual word semantics, there is stil
l ample scope for improvement regarding computational approaches. We assume that idiomatic constructions can be characterized by gradual intensities of semantic non-compositionality, formal fixedness, and unusual usage context, and introduce a number of measures for these characteristics, comprising count-based and predictive collocation measures together with measures of context (un)similarity. We evaluate our approach on a manually labelled gold standard, derived from a corpus of German pop lyrics. To this end, we apply a Random Forest classifier to analyze the individual contribution of features for automatically detecting idioms, and study the trade-off between recall and precision. Finally, we evaluate the classifier on an independent dataset of idioms extracted from a list of Wikipedia idioms, achieving state-of-the art accuracy.
The Master's stage has its importance in development and intellectual building.
However, in the process of applying for Master's studies, applicants must pass the English
Admission Test. When doing this test, master's students encounter difficultie
s in the area of
everyday English.
This research deals with the difficulties that master's students may face in the area of
everyday or spoken English. It is from the perspective of a sample of students doing a
language course that can help them do the final English admission test at the Higher
Institute of languages, Tishreen University. In order to achieve this goal, a questionnaire
has been designed to investigate into difficulties faced by those students in the area of
everyday English. After conducting the study and analysis, the research arrived at the
following results:
1- Master's students doing the English admission test suffer from difficulties in
idioms, the way they are written, and their use. In addition to that, there are anxieties when
dealing with phrasal verbs and sayings that reflect the cultural aspects of the English
society. Furthermore, there are problems with understanding English conversations.
2- There are no varieties with statistical significance in the area of everyday
English according to the sample of students in regard to the variables of gender and age.
3- Among the views of the sample of students, there exists no relation at a
significance in everyday English that can be attributed to the variable of the stage of
English acquisition.
This research paper aimed to study the idioms lexicalization in medium sized
monolingual dictionaries as an attempt to combine theorization and practice of
contemporary lexicography. Theorization, or the theoretical part of the research, is
dedica
ted to study idioms definition, aspects and types.
To achieve this target, the paper focuses on the theoretical aspect of idioms regarding
institutional lexicons (collaborative work) as exemplified in al-Muʿğam al-Wasīṭ of Cairo
Academy of the Arabic Language. The paper depends on studying individually collected
lexicons of Al- Munğid by Louis Maʿlūf as applied examples on the practical level.
This study investigates the problems and difficulties that
Syrian EFL undergraduate students face in learning English idioms.
The study aims to examine the students' competence in English
idiomatic expressions and their ability to predict their me
anings in
different contexts based on their first language transfer, and
whether there is a significant difference between the participants'
productive and receptive knowledge of idioms.
