GRAW+: A two‐view graph propagation method with word coupling for readability assessment

Abstract

Existing methods for readability assessment usually construct inductive classification models to assess the readability of singular text documents based on extracted features, which have been demonstrated to be effective. However, they rarely make use of the interrelationship among documents on readability, which can help increase the accuracy of readability assessment. In this article, we adopt a graphbased classification method to model and utilize the relationship among documents using the coupled bag-of-words model. We propose a word coupling method to build the coupled bag-of-words model by estimating the correlation between words on reading difficulty. In addition, we propose a two-view graph propagation method to make use of both the coupled bag-of-words model and the linguistic features. Our method employs a graph merging operation to combine graphs built according to different views, and improves the label propagation by incorporating the ordinal relation among reading levels. Experiments were conducted on both English and Chinese data sets, and the results demonstrate both effectiveness and potential of the method.

A network coupling approach to detecting hierarchical linkages between science and technology

“…For instance, Zhao and Strotmann (2010) constructed a hybrid citation network based on bibliographic coupling, which is far more effective than cocitation and direct citation networks in predicting research trends. Jiang et al (2019) used graph merging to combine graphs constructed based on different views and improve label propagation by merging ordered relations between reading layers. Ba and Liang (2021) measured the interaction pattern of S&T network linkages by constructing coupled S&T networks.…”

Trends, Limitations and Open Challenges in Automatic Readability Assessment Research

“…Xia et al (2016) showed that pair wise ranking approach may generalize better compared to classification. Unlike such approaches, Jiang et al (2019) proposed a graph propagation based approach to ARA, which can potentially consider the inter-relationships between documents while modeling readability.…”

InterpretARA: Enhancing Hybrid Automatic Readability Assessment with Linguistic Feature Interpreter and Contrastive Learning

“…In ARA, researchers often utilize various linguistic features to quantify the difficulty level of a text (Flesch 1948;Chall 1948). These features can capture different aspects of the linguistic nature of a text, including vocabulary, syntax, semantics, and discourse structures (Dell'Orletta, Montemagni, and Venturi 2011;Hancke, Vajjala, and Meurers 2012;Sung et al 2015;Denning, Pera, and Ng 2016;Arfé, Mason, and Fajardo 2018;Jiang et al 2019). Due to the richness of linguistic features, they are incorporated in both traditional machine learning and modern deep learning approaches.…”