Search Results (14)

Background/Objectives: This study examined the effects of a computerized naming intervention combined with either cerebellar anodal transcranial direct-current stimulation (A-tDCS) or sham (S-tDCS) on noun and verb naming in Cantonese-speaking persons with chronic stroke-related aphasia (PWA). Methods: A double-blind, randomized, crossover, sham-controlled clinical trial was conducted with six Cantonese-speaking PWA following stroke. Participants received a 60 min computerized naming intervention incorporating audio–visual speech perception cues over five consecutive days, paired with concurrent 20 min of either 2 mA cerebellar A-tDCS or S-tDCS. Generalized linear mixed-effects models (GLMM) and linear mixed-effects models (LME) were used to evaluate naming accuracy and reaction time (RT). Individual variability was further explored through single-case analyses of naming accuracy changes across conditions and grammatical categories. Results: The GLMM showed a significant three-way interaction of condition, grammatical category, and time (p < 0.05). Specifically, the intervention paired with S-tDCS significantly improved verb naming, whereas A-tDCS did not induce significant improvements at the group level, effectively showing significantly smaller gains regarding verb naming compared to S-tDCS. Overall, RT decreased post-treatment across groups, but no significant differences emerged by the tDCS condition. The results support the promising efficacy of the Cantonese computerized audio–visual noun and verb naming therapy. Single-case analyses revealed high inter-individual variability in response to neuromodulation effects on naming and behavioral treatment outcomes. Conclusions: This study contributes to the emerging literature on cerebellar neuromodulation in post-stroke aphasia and underscores the need for larger trials examining grammar-specific (particularly verb-related) effects and polarity-dependent outcomes. It also highlights the value of developing personalized neuromodulation protocols to optimize the efficacy of behavioral language interventions in people with aphasia. Full article

This study investigates the impact of structured input, referential activities, and affective activities on English simple past tense acquisition in a second language (L2). Thirty-three participants from a senior high school were divided into four groups based on the pretest–posttest design: referential only, affective only, a combination of both, and a control group. A self-paced reading (SPR) test was used to measure accuracy and response times to evaluate the effectiveness of these instructional strategies. Structured input and referential tasks enhance grammatical acquisition more rapidly and accurately than affective-only treatments or controls, showing the beneficial effects of structured input on grammar acquisition. The results emphasized the importance of designing instructional strategies that address specific processing challenges in L2 learning by focusing on form–meaning connections. By demonstrating differential impacts of structured input activities on grammatical learning and processing efficiency, the research contributes to the field of second language acquisition. The SPR method was selected for its ability to capture subtle, immediate differences in processing at the word level, its suitability for controlled classroom-based online administration, and its established validity in L2 processing research. Unlike other methods, SPR allows precise measurement of reaction times for specific sentence components, isolating processing effects of the target grammatical form while minimizing the influence of explicit knowledge. Full article

Text classification remains a challenging task in natural language processing (NLP) due to linguistic complexity and data imbalance. This study proposes a hybrid approach that integrates grammar-based feature engineering with deep learning and transformer models to enhance classification performance. A dataset of factoid and non-factoid questions, further categorized into causal, choice, confirmation, hypothetical, and list types, is used to evaluate several models, including CNNs, BiLSTMs, MLPs, BERT, DistilBERT, Electra, and GPT-2. Grammatical and domain-specific features are explicitly extracted and leveraged to improve multi-class classification. To address class imbalance, the SMOTE algorithm is applied, significantly boosting the recall and F1-score for minority classes. Experimental results show that DistilBERT achieves the highest binary classification accuracy, equal to 94%, while BiLSTM and CNN outperform transformers in multi-class settings, reaching up to 92% accuracy. These findings confirm that grammar-based features provide critical syntactic and semantic insights, enhancing model robustness and interpretability beyond conventional embeddings. Full article

This study conducts a comprehensive evaluation of large language models (LLMs), including ChatGPT 4, ERNIE Bot 4, and Gemini Advanced, in the context of translating Buddhist texts from classical Chinese to modern English. Focusing on three distinct Buddhist texts encompassing various literary forms and complexities, the analysis examines the models’ capabilities in handling specialized Buddhist terminology, classical Chinese grammar, and the translation of complex, lengthy sentences. The study employs a methodology where selected excerpts from these texts are translated by the LLMs, followed by an in-depth analysis comparing these machine-generated translations to human translations. The evaluation criteria include word translation accuracy, the ability to recognize and correctly interpret specific meanings within both classical and modern contexts, and the completeness of phrases without omitting or unnecessarily adding words. The findings reveal significant variations in the performance of these LLMs, with detailed observations on their strengths and weaknesses in translating specialized terms, managing grammatical structures unique to classical Chinese, and maintaining the integrity of the original texts’ meanings. This paper aims to shed light on the potential and limitations of using LLMs for translating complex literary works from ancient to modern languages, contributing valuable insights into the field of computational linguistics and the ongoing development of translation technologies. Full article

Understanding the structural intricacies of RNA molecules is essential for deciphering numerous biological processes. Traditionally, scientists have relied on experimental methods to gain insights and draw conclusions. However, the recent advent of advanced computational techniques has significantly accelerated and refined the accuracy of research results in several areas. A particularly challenging aspect of RNA analysis is the prediction of its secondary structure, which is crucial for elucidating its functional role in biological systems. This paper deals with the prediction of pseudoknots in RNA, focusing on two types of pseudoknots: K-type and M-type pseudoknots. Pseudoknots are complex RNA formations in which nucleotides in a loop form base pairs with nucleotides outside the loop, and thus contribute to essential biological functions. Accurate prediction of these structures is crucial for understanding RNA dynamics and interactions. Building on our previous work, in which we developed a framework for the recognition of H- and L-type pseudoknots, an extended grammar-based framework tailored to the prediction of K- and M-type pseudoknots is proposed. This approach uses syntactic pattern recognition techniques and provides a systematic method to identify and characterize these complex RNA structures. Our framework uses context-free grammars (CFGs) to model RNA sequences and predict the occurrence of pseudoknots. By formulating specific grammatical rules for type K- and M-type pseudoknots, we enable efficient parsing of RNA sequences to recognize potential pseudoknot configurations. This method ensures an exhaustive exploration of possible pseudoknot structures within a reasonable time frame. In addition, the proposed method incorporates essential concepts of biology, such as base pairing optimization and free energy reduction, to improve the accuracy of pseudoknot prediction. These principles are crucial to ensure that the predicted structures are biologically plausible. By embedding these principles into our grammar-based framework, we aim to predict RNA conformations that are both theoretically sound and biologically relevant. Full article

Semantic generation of network access control policies can help network administrators accurately implement policies to achieve desired security objectives. Current semantic generation research mainly focuses on semantic generation of single grammar and lacks work on automatically generating semantics for different grammatical strategies. Generating semantics for different grammars is a tedious, inefficient, and non-scalable task. Inspired by sequence labeling in the field of natural language processing, this article models automatic semantic generation as a sequence labeling task. We propose a semantic generation algorithm named ADLBiLSTM. The algorithm uses a self-attention mechanism and double-layer BiLSTM to extract the features of security policies from different aspects, so that the algorithm can flexibly adapt to policies of different complexity without frequent modification. Experimental results showed that the algorithm has good performance and can achieve high accuracy in semantic generation of access control list (ACL) and firewall data and can accurately understand and generate the semantics of network access control policies. Full article

Aspect-level sentiment analysis is a research focal point for natural language comprehension. An attention mechanism is a very important approach for aspect-level sentiment analysis, but it only fuses sentences from a semantic perspective and ignores grammatical information in the sentences. Graph convolutional networks (GCNs) are a better method for processing syntactic information; however, they still face problems in effectively combining semantic and syntactic information. This paper presents a sentiment-supported graph convolutional network (SSGCN). This SSGCN first obtains the semantic information of the text through aspect-aware attention and self-attention; then, a grammar mask matrix and a GCN are applied to preliminarily combine semantic information with grammatical information. Afterward, the processing of these information features is divided into three steps. To begin with, features related to the semantics and grammatical features of aspect words are extracted. The second step obtains the enhanced features of the semantic and grammatical information through sentiment support words. Finally, it concatenates the two features, thus enhancing the effectiveness of the attention mechanism formed from the combination of semantic and grammatical information. The experimental results show that compared with benchmark models, the SSGCN had an improved accuracy of 6.33–0.5%. In macro F1 evaluation, its improvement range was 11.68–0.5%. Full article

Grammar error correction (GEC) is a crucial task in the field of Natural Language Processing (NLP). Its objective is to automatically detect and rectify grammatical mistakes in sentences, which possesses immense application research value. Currently, mainstream grammar-correction methods primarily rely on sequence labeling and text generation, which are two kinds of end-to-end methods. These methods have shown exemplary performance in areas with low error density but often fail to deliver satisfactory results in high-error density situations where multiple errors exist in a single sentence. Consequently, these methods tend to overcorrect correct words, leading to a high rate of false positives. To address this issue, we researched the specific characteristics of the Chinese grammar error correction (CGEC) task in high-error density situations. We proposed a grammar-correction method based on prompt templates. Firstly, we proposed a strategy for constructing prompt templates suitable for CGEC. This strategy transforms the CGEC task into a masked fill-in-the-blank task compatible with the masked language model BERT. Secondly, we proposed a method for dynamically updating templates, which incorporates already corrected errors into the template through dynamic updates to improve the template quality. Moreover, we used the phonetic and graphical resemblance knowledge from the confusion set as guiding information. By combining this with BERT’s prediction results, the model can more accurately select the correct characters, significantly enhancing the accuracy of the model’s prediction correction results. Our methods were validated through experiments on a public grammar-correction dataset. The results indicate that our method achieves higher correction performance and lower false correction rates in high-error density scenarios. Full article

The problems of machine translation are constantly arising. While the most advanced translation platforms, such as Google and Yandex, allow for high-quality translations of languages with simple grammatical structures, more morphologically rich languages still suffer from the translation of complex sentences, and translation services leave many structural errors. This study focused on designing the rules for the grammatical structures of complex sentences in the Kazakh language, which has a difficult grammar with many rules. First, the types of complex sentences in the Kazakh language were thoroughly observed with the use of templates from the FuzzyWuzzy library. Then, the correction of complex sentences was completed with parallel corpora. The sentences were translated into English and Russian by existing machine translation systems. Therefore, the grammar of both Kazakh–English and Kazakh–Russian language pairs was considered. They both used the rules specifically designed for the post-editing steps. Finally, the performance of the developed algorithm was evaluated for an accuracy score for each pair of languages. This approach was then proposed for use in other corpora generation, post-editing, and analysis systems in future works. Full article

Previous work has provided contrasting evidence on syntax acquisition. Syntax-internal factors, i.e., instinctive knowledge of the universals of grammar (UG) for finite-state grammar (FSG) and phrase-structure grammar (PSG) but also syntax-external factors such as language competence, working memory (WM) and demographic factors may affect syntax acquisition. This study employed an artificial grammar paradigm to identify which factors predicted syntax acquisition. Thirty-seven healthy individuals and forty-nine left-hemispheric stroke patients (fourteen with aphasia) read syllable sequences adhering to or violating FSG and PSG. They performed preference classifications followed by grammatical classifications (after training). Results showed the best classification accuracy for sequences adhering to UG, with performance predicted by syntactic competence and spatial WM. Classification of ungrammatical sequences improved after training and was predicted by verbal WM. Although accuracy on FSG was better than on PSG, generalization was fully possible only for PSG. Education was the best predictor of syntax acquisition, while aphasia and lesion volume were not predictors. This study shows a clear preference for UG, which is influenced by spatial and linguistic knowledge, but not by the presence of aphasia. Verbal WM supported the identification of rule violations. Moreover, the acquisition of FSG and PSG was related to partially different mechanisms, but both depended on education. Full article

Grammar provides the framework for understanding and producing language. In aphasia, an acquired language disorder, grammatical deficits are diversified and widespread. However, the few assessments for testing grammar in the German language do not consider current linguistic, psycholinguistic, and functional imaging data, which have been shown to be crucial for effective treatment. This study developed German language versions of the Northwestern Assessment of Verbs and Sentences (NAVS-G) and the Northwestern Anagram Test (NAT-G) to examine comprehension and production of verbs, controlling for the number and optionality of verb arguments, and sentences with increasing syntactic complexity. The NAVS-G and NAT-G were tested in 27 healthy participants, 15 right hemispheric stroke patients without aphasia, and 15 stroke patients with mild to residual aphasia. Participants without aphasia showed near-perfect performance, with the exception of (object) relative sentences, where accuracy was associated with educational level. In each patient with aphasia, deficits in more than one subtest were observed. The within and between population-groups logistic mixed regression analyses identified significant impairments in processing syntactic complexity at the verb and sentence levels. These findings indicate that the NAVS-G and NAT-G have potential for testing grammatical competence in (German) stroke patients. Full article

This study investigates the effects of task-essential training on offline and online processing of verbal morphology and explores how working memory (WM) modulates the effects of training. We compare a no-training control group to two training groups who completed a multisession task-essential training focused on Spanish verbal inflections related to person–number agreement and tense. Effects of training were evaluated using an offline aural interpretation task and an online self-paced reading (SPR) assessment, administered as a pretest, posttest, and delayed posttest. Results showed that training led to more accurate interpretation of both person-number and tense information in the offline interpretation test. While higher WM was associated generally with greater accuracy, higher WM did not lead to greater gains from training. The SPR results showed that training did not increase sensitivity to subject–verb agreement or adverb–verb tense violations. However, among participants who underwent training, WM enhanced sensitivity under some conditions. These results demonstrate a role for individual differences in WM for offline and online processing, and they suggest that while task-essential training has been shown repeatedly to improve offline processing of target forms, its effects on online processing of redundant verbal morphology are more limited. Implications for L2 learning are discussed. Full article

Grammatical inference (GI), i.e., the task of finding a rule that lies behind given words, can be used in the analyses of amyloidogenic sequence fragments, which are essential in studies of neurodegenerative diseases. In this paper, we developed a new method that generates non-circular parsing expression grammars (PEGs) and compares it with other GI algorithms on the sequences from a real dataset. The main contribution of this paper is a genetic programming-based algorithm for the induction of parsing expression grammars from a finite sample. The induction method has been tested on a real bioinformatics dataset and its classification performance has been compared to the achievements of existing grammatical inference methods. The evaluation of the generated PEG on an amyloidogenic dataset revealed its accuracy when predicting amyloid segments. We show that the new grammatical inference algorithm achieves the best ACC (Accuracy), AUC (Area under ROC curve), and MCC (Mathew’s correlation coefficient) scores in comparison to five other automata or grammar learning methods. Full article

This work focuses on designing a grammar detection system that understands both structural and contextual information of sentences for validating whether the English sentences are grammatically correct. Most existing systems model a grammar detector by translating the sentences into sequences of either words appearing in the sentences or syntactic tags holding the grammar knowledge of the sentences. In this paper, we show that both these sequencing approaches have limitations. The former model is over specific, whereas the latter model is over generalized, which in turn affects the performance of the grammar classifier. Therefore, the paper proposes a new sequencing approach that contains both information, linguistic as well as syntactic, of a sentence. We call this sequence a Lex-Pos sequence. The main objective of the paper is to demonstrate that the proposed Lex-Pos sequence has the potential to imbibe the specific nature of the linguistic words (i.e., lexicals) and generic structural characteristics of a sentence via Part-Of-Speech (POS) tags, and so, can lead to a significant improvement in detecting grammar errors. Furthermore, the paper proposes a new vector representation technique, Word Embedding One-Hot Encoding ($W_E{O}_E$) to transform this Lex-Pos into mathematical values. The paper also introduces a new error induction technique to artificially generate the POS tag specific incorrect sentences for training. The classifier is trained using two corpora of incorrect sentences, one with general errors and another with POS tag specific errors. Long Short-Term Memory (LSTM) neural network architecture has been employed to build the grammar classifier. The study conducts nine experiments to validate the strength of the Lex-Pos sequences. The Lex-Pos -based models are observed as superior in two ways: (1) they give more accurate predictions; and (2) they are more stable as lesser accuracy drops have been recorded from training to testing. To further prove the potential of the proposed Lex-Pos -based model, we compare it with some well known existing studies. Full article