Verb Classification Across Languages

نویسندگان

چکیده

Recent developments in language modeling have enabled large text encoders to derive a wealth of linguistic information from raw corpora without supervision. Their success across natural processing (NLP) tasks has called into question the role man-made computational resources, such as verb lexicons, supporting modern NLP. Still, probing analyses concurrently exposed limitations knowledge possessed by neural architectures, revealing them be clever task solvers rather than self-taught linguists. Can human-designed lexical resources still help fill their gaps? Focusing on classification, we discuss approaches generating classes multilingually and weigh relative benefits undertaking expensive lexicographic work outsourcing untrained native speakers. Then, consider evidence for utility augmenting pretrained models with external ponder ways which human expertise can continue benefit multilingual

برای دانلود باید عضویت طلایی داشته باشید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Verb classes within and across languages

Any attempt to characterize valency patterns within and across languages confronts the phenomenon of verb or valency classes: sets of semantically-related verbs sharing a range of linguistic properties, such as the possible realizations of their arguments and the particular interpretation associated with each possible argument realization. The Leipzig Valency Classes Project provides an excelle...

متن کامل

Events and verb classification*

The prevailing approach to understanding the syntactic mapping of arguments and argument alternations is to assume that verbs are organized into semantically distinct verb classes and that class membership determines the grammaticality of argument alternations. This paper presents syntactic evidence that verb-class membership cannot explain the causative and locative alternations, and that alte...

متن کامل

Why verb - initial languages are not frequent

In our simulations with simple recurrent networks we demonstrate that small artificial languages are learnt worse or better depending on their basic word order. We show that verb-initial languages are difficult to learn, reflecting the lower frequency of verb-initial natural languages. We try to go beyond mere simulations proposing two objective mathematical measures to explain our results.

متن کامل

Automatic Verb Classification Using a

Automatic Verb Classification Using a General Feature Space Eric Joanis Master of Science Graduate Department of Computer Science University of Toronto 2002 We develop a general feature space that can be used for the semantic classification of English verbs. We design a technique to extract these features from a large corpus of English, while trying to maintain portability to other languages—th...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Annual review of linguistics

سال: 2023

ISSN: ['2333-9683', '2333-9691']

DOI: https://doi.org/10.1146/annurev-linguistics-030521-043632