![]() ![]() ![]() Other models have argued that differences in first and second language processing result largely from capacity differences, differences in susceptibility to interference, or lack of predictive ability. Models grounded in neurocognitive approaches to memory hold that late bilinguals recruit different memory systems compared to native speakers of the target language. Proposals rooted in generative approaches to language acquisition argue that adult second language (L2) learners lack access to the universal principles or the ability to reset parameters that guide language acquisition and language processing in their L2. The debate about bilingual sentence processing has, instead, focused on whether bilingual speakers process their second language in a manner similar to monolingual speakers of the target language. The theoretical divide in the field has been about whether the architecture and mechanisms of the human sentence processor are modular-and computations are carried out serially-or whether it is interactive and computations are carried out in parallel. Often overlooked, robust automatic evaluation methodology is necessary for improving systems, and this work presents new metrics and outlines important considerations for reliably measuring the quality of the generated text.The main goal of monolingual models of sentence processing is to explain how the syntactic processor (or parser) assigns structure to an incoming string of words. We analyze common evaluation practices and propose better methods that more accurately measure the quality of output. Sentential paraphrases must fulfill a variety of requirements: preserve the meaning of the original sentence, be grammatical, and meet any stylistic or task-specific constraints. ![]() Once candidate sentences are generated, it is crucial to have reliable evaluation methods. Our method generates more meaning-preserving and grammatical sentences than earlier approaches and requires less task-specific data. We modify the statistical machine translation pipeline to harness monolingual resources and insights into task constraints in order to drastically diminish the amount of annotated data necessary to train a robust system. Parallel bilingual data naturally occurs between common language pairs (such as English and French), but for monolingual sentence rewriting, there is little existing parallel data and annotation is costly. In machine translation, a large quantity of parallel data is necessary to model the transformations from input to output text. Monolingual rewriting can be thought of as translating between two types of English (such as from complex to simple), and therefore our approach is inspired by statistical machine translation. We also perform a detailed analysis of the evaluation methodologies for each task, identify bias in common evaluation techniques, and propose more reliable practices. We introduce a unified framework for monolingual sentence rewriting, and apply it to three representative tasks: sentence compression, text simplification, and grammatical error correction. In this thesis, we investigate approaches to paraphrasing entire sentences within the constraints of a given task, which we call monolingual sentence rewriting. ![]()
0 Comments
Leave a Reply. |
Details
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |