Text Genres and Registers: The Computation of Linguistic Features

Fang, Alex Chengyu; Cao, Jing

doi:10.1007/978-3-662-45100-7

Cited by 6 publications

(6 citation statements)

References 77 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, one of the limitations of the n-gram model is the fact that the method is ignorant of the grammatical nature of the text. Recently, a Part of Speech (PoS) n-gram model was used extensively for text classification [49,50]. A recent study by Liu et al [44] stated that 'further writing analytics development is needed, particularly for the feelings factor, which does not appear to be well covered by the variables we have considered here -what writing analytic features might prove to be more reliable indicators of feelings if they are denoted as important to the quality of a reflection?'(p.12).…”

Section: Proposed Frameworkmentioning

confidence: 99%

Evaluating an Automated Analysis Using Machine Learning and Natural Language Processing Approaches to Classify Computer Science Students’ Reflective Writing

Alrashidi¹,

Almujally²,

Kadhum³

et al. 2022

Pervasive Computing and Social Networking

View full text Add to dashboard Cite

Reflection writing is a common practice in higher education. However, manual analysis of written reflections is time-consuming. This study presents an automated analysis of reflective writing to analyze reflective writing in CS education based on conceptual Reflective Writing Framework (RWF) and application of natural language processing and machine learning algorithm. This paper investigates two groups of features extraction (n-grams and PoS n-grams) and random forest (RF) algorithm that utilize such features to detect the presence or absence of the seven indicators (description of an experience, understandings, feelings, reasoning, perspective, new learning, and future action). The automated analysis of reflective writing is evaluated based on 74 CS student essays (1113 sentences) that are from the final year project reports in CS's students. Results showed the seven indicators can be reliably distinguished by their features and these indicators can be used in an automated reflective writing analysis for determining the level of students' reflective writing. Finally, we consider the implications of how the conceptualization of refection quality and providing individualized learning support to students in order to help them develop reflective skills.

show abstract

Section: Proposed Frameworkmentioning

confidence: 99%

Evaluating an Automated Analysis Using Machine Learning and Natural Language Processing Approaches to Classify Computer Science Students’ Reflective Writing

Alrashidi¹,

Almujally²,

Kadhum³

et al. 2022

Pervasive Computing and Social Networking

View full text Add to dashboard Cite

show abstract

“…Halliday defines register as a configuration of meanings that are typically associated with a particular situational configuration of field, mode and tenor (Halliday 1985). However, register also has a bearing on genre, in that different genres are observed to have different degree of formality (Fang, Cao 2015). In some cases, the register used in the TT features higher levels of formality than the one chosen for the ST.…”

Section: Register Variationmentioning

confidence: 99%

“Dear Emma…”. Genre Overlapping and Register Variation in the English and Italian Version of Sergio Marchionne’s Letter to Confindustria

Manzella¹

2020

Comparative Legilinguistics

View full text Add to dashboard Cite

The aim of this paper is to examine whether genre overlapping and register variation have implications in terms of translation outcomes. To this end, the English and the Italian version of the letter sent by Sergio Marchionne to the President of Confindustria – Italy’s most important employers’ organisation – will be analysed. The decision to investigate Marchionne’s statement is based on the fact that this is a letter disseminated as a press release, which also contains highly-technical terminology pertaining to the discourse of Employment Relations (ER). It is stressed that a number of issues arise when deciding which genre and register to use when translating specialised texts. It is argued that genre overlapping and register variation affect the texts in English and Italian in important respects, leading to different perceptions on the part of the target audience.

show abstract

“…Studies employing bottom-up clustering methods that support prevalent register classifications notwithstanding (e.g. Gries et al 2011, Fang & Cao 2015, my own studies highlight that it would be naive to rely on these opaque register labels. This is by no means meant as criticism of the respective ICE components.…”

Section: Interim Summary and Critical Reflectionmentioning

confidence: 84%

“…spoken vs. written): in this case, both POS tagsets outperform the word-based model. The experiments ofFang & Cao (2015) further show a substantial drop in the classification accuracy on finer levels of granularity: they achieve accuracy rates of 99.8% when classifying into spoken/written texts, 74.7% on a genre and a mere 58.2% on…”

mentioning

confidence: 97%

“…In contrast, studies interested in the discriminatory power of certain feature sets use prelabelled texts as training data for machine learning algorithms and evaluate the accuracy of different feature sets. Note that while his study provided invaluable insights into register variation, the majority of studies which apply methods for unsupervised clustering rely on much simpler features and still achieve very good results Fang & Cao (2015). compare the classification accuracies for three different representations of the texts in ICE-GB: monograms of a fine-grained POS tagset(AUTASYS tagset, Fang 1996), monograms based on an impoverished version of it (similar to the Oxford Simplified Tagset used in the BNC) and word monograms.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Issues of corpus comparability and register variation in the International Corpus of English: Theories and computer applications

Vetter¹

View full text Add to dashboard Cite

This study offers an account of the issue of corpus comparability of components of the International Corpus of English (ICE). By employing quantitative and qualitative methods, it contributes to corpus-based studies of varieties of English, and corpus linguistics as a linguistic discipline more generally. Specifically, it (i) exemplifies how discrepancies in sampling strategies can decrease the comparability of components of comparable corpus families such as ICE, (ii) presents methods to detect such discrepancies, (iii) develops and releases a user-friendly computer program (ICEtree) that allows the application of these methods to components of ICE that are not investigated in this study and (iv) illustrates how a register-based annotation framework could help mitigate some of the conflicting priorities in the use and compilation of comparable corpora.

show abstract

Text Genres and Registers: The Computation of Linguistic Features

Cited by 6 publications

References 77 publications

Evaluating an Automated Analysis Using Machine Learning and Natural Language Processing Approaches to Classify Computer Science Students’ Reflective Writing

Evaluating an Automated Analysis Using Machine Learning and Natural Language Processing Approaches to Classify Computer Science Students’ Reflective Writing

“Dear Emma…”. Genre Overlapping and Register Variation in the English and Italian Version of Sergio Marchionne’s Letter to Confindustria

Issues of corpus comparability and register variation in the International Corpus of English: Theories and computer applications

Contact Info

Product

Resources

About