Mengyang Qiu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Mengyang Qiu has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Jungyeul Park
The University of British Columbia Canada
2 shared publications - Jayoung Song
Department of Asian Studies, The Pennsylvania State University, University Park, USA
1 shared publication Jiexin Kuang
Department of Linguistics The University of British Columbia Canada
1 shared publicationMin Zeng
Department of Linguistics The University of British Columbia Canada
1 shared publication- Yu Li
School of Chinese Language and Literature , 12390 Wuhan University , Wuhan , China
1 shared publication
Latest Publications
- Open Access
- 23 May 2025
- Published by Walter de Gruyter GmbH in Poznan Studies in Contemporary Linguistics
- Vol. 61 (2) , 121-172
A network analysis of the semantic evolution of ‘fruit’ and ‘stone’ in Tibeto-Burman languages
The lexemes ‘fruit’ and ‘stone’ are known as the origins of the numeral classifiers for small round objects in many Tibeto-Burman languages. This paper employs a correlation-based network construction method to investigate the colexification networks of the two concepts in 58 + 68 Tibeto-Burman languages. A total of 104 concepts colexified with ‘fruit’ and 99 concepts colexified with ‘stone’ are organized into macro semantic classes. Semantic networks on the basis of the similarities in colexification patterns of concepts, as well as languages networks on the basis of the similarities in colexification patterns of languages, are constructed for ‘fruit’ and ‘stone’, respectively. The results indicate that classifiers for small round objects evolved from either ‘fruit’ or ‘stone’ are directly colexified with class terms in compound nouns denoting varieties of fruits/stones and the shape class of small round objects, indicating that they are diachronically related. However, ‘fruit’ and ‘stone’ differ significantly in their modes of deriving a classifier. Moreover, languages that have developed classifiers from ‘fruit’ are mostly from the Ngwi subgroup, whereas languages whose classifiers are colexified with ‘stone’ evolved independently.
- 23 May 2025
- Published by Walter de Gruyter GmbH in Poznan Studies in Contemporary Linguistics
- Vol. 61 (2) , 121-172
Open AccessA network analysis of the semantic evolution of ‘fruit’ and ‘stone’ in Tibeto-Burman languages
The lexemes ‘fruit’ and ‘stone’ are known as the origins of the numeral classifiers for small round objects in many Tibeto-Burman languages. This paper employs a correlation-based network construction method to investigate the colexification networks of the two concepts in 58 + 68 Tibeto-Burman languages. A total of 104 concepts colexified with ‘fruit’ and 99 concepts colexified with ‘stone’ are organized into macro semantic classes. Semantic networks on the basis of the similarities in colexification patterns of concepts, as well as languages networks on the basis of the similarities in colexification patterns of languages, are constructed for ‘fruit’ and ‘stone’, respectively. The results indicate that classifiers for small round objects evolved from either ‘fruit’ or ‘stone’ are directly colexified with class terms in compound nouns denoting varieties of fruits/stones and the shape class of small round objects, indicating that they are diachronically related. However, ‘fruit’ and ‘stone’ differ significantly in their modes of deriving a classifier. Moreover, languages that have developed classifiers from ‘fruit’ are mostly from the Ngwi subgroup, whereas languages whose classifiers are colexified with ‘stone’ evolved independently.
- Open Access
Frustratingly Simple Prompting-based Text Denoising
This paper introduces a novel perspective on the automated essay scoring (AES) task, challenging the conventional view of the ASAP dataset as a static entity. Employing simple text denoising techniques using prompting, we explore the dynamic potential within the dataset. While acknowledging the previous emphasis on building regression systems, our paper underscores how making minor changes to a dataset through text denoising can enhance the final results.
Open AccessFrustratingly Simple Prompting-based Text Denoising
This paper introduces a novel perspective on the automated essay scoring (AES) task, challenging the conventional view of the ASAP dataset as a static entity. Employing simple text denoising techniques using prompting, we explore the dynamic potential within the dataset. While acknowledging the previous emphasis on building regression systems, our paper underscores how making minor changes to a dataset through text denoising can enhance the final results.
- Open Access
Evaluating Prompting Strategies for Grammatical Error Correction Based on Language Proficiency
The writing examples of English language learners may be different from those of native speakers. Given that there is a significant differences in second language (L2) learners' error types by their proficiency levels, this paper attempts to reduce overcorrection by examining the interaction between LLM's performance and L2 language proficiency. Our method focuses on zero-shot and few-shot prompting and fine-tuning models for GEC for learners of English as a foreign language based on the different proficiency. We investigate GEC results and find that overcorrection happens primarily in advanced language learners' writing (proficiency C) rather than proficiency A (a beginner level) and proficiency B (an intermediate level). Fine-tuned LLMs, and even few-shot prompting with writing examples of English learners, actually tend to exhibit decreased recall measures. To make our claim concrete, we conduct a comprehensive examination of GEC outcomes and their evaluation results based on language proficiency.
Open AccessEvaluating Prompting Strategies for Grammatical Error Correction Based on Language Proficiency
The writing examples of English language learners may be different from those of native speakers. Given that there is a significant differences in second language (L2) learners' error types by their proficiency levels, this paper attempts to reduce overcorrection by examining the interaction between LLM's performance and L2 language proficiency. Our method focuses on zero-shot and few-shot prompting and fine-tuning models for GEC for learners of English as a foreign language based on the different proficiency. We investigate GEC results and find that overcorrection happens primarily in advanced language learners' writing (proficiency C) rather than proficiency A (a beginner level) and proficiency B (an intermediate level). Fine-tuned LLMs, and even few-shot prompting with writing examples of English learners, actually tend to exhibit decreased recall measures. To make our claim concrete, we conduct a comprehensive examination of GEC outcomes and their evaluation results based on language proficiency.