Search Results (19)

Orthographic knowledge is a critical component of skilled language use, yet its large-scale behavioral signatures remain understudied in Spanish. To address this gap, we developed OrthoKnow-SP, a megastudy that captures spelling decisions from 27,185 native Spanish-speaking adults who completed an 80-item forced-choice task. Each trial required selecting the correctly spelled word from a pair comprising a real word and a pseudohomophone foil that preserved pronunciation while violating the correct graphemic representation. The stimuli targeted six high-confusability contrasts in Spanish orthography. We recorded response accuracy and reaction times for over 2.17 million trials, alongside demographic and device metadata. Results show robust variability across items and individuals, with item-level metrics closely aligned with independent norms of word prevalence. A composite difficulty index integrating speed and accuracy further allowed fine-grained item ranking. The dataset provides the first population-scale norms of Spanish spelling difficulty, capturing regional and generational diversity absent from traditional lab-based studies. Public release of OrthoKnow-SP enables new research on the cognitive and demographic factors shaping orthographic decisions, and provides educators, clinicians, and developers with a valuable benchmark for assessing spelling competence and modeling written language processing. Full article

Named Entity Recognition (NER) is a fundamental task in Natural Language Processing (NLP) that supports applications such as information retrieval, sentiment analysis, and text summarization. While substantial progress has been made in NER for widely studied languages like English, Arabic presents unique challenges due to its morphological richness, orthographic ambiguity, and the frequent occurrence of nested and overlapping entities. This paper introduces a novel Arabic NER framework that addresses these complexities through architectural innovations. The proposed model incorporates a Hybrid Feature Fusion Layer, which integrates external lexical features using a cross-attention mechanism and a Gated Lexical Unit (GLU) to filter noise, while a Compound Span Representation Layer employs Rotary Positional Encoding (RoPE) and Bidirectional GRUs to enhance the detection of complex entity structures. Additionally, an Enhanced Multi-Label Classification Layer improves the disambiguation of overlapping spans and assigns multiple entity types where applicable. The model is evaluated on three benchmark datasets—ANERcorp, ACE 2005, and a custom biomedical dataset—achieving an F1-score of 93.0% on ANERcorp and 89.6% on ACE 2005, significantly outperforming state-of-the-art methods. A case study further highlights the model’s real-world applicability in handling compound and nested entities with high confidence. By establishing a new benchmark for Arabic NER, this work provides a robust foundation for advancing NLP research in morphologically rich languages. Full article

In reading, rapid and reliable word recognition relies on high-quality representations at both the lexical and sublexical levels, with stable and flexible connections between form, sound, and meaning. Earlier studies suggested that meaning knowledge affects the formation and quality of orthographic representation in language learning, but the impact of morphemic meaning frequency on learners’ word recognition was not explored. This research examined second language (L2) Chinese readers’ recognition of compound words containing ambiguous morphemes. Using lexical decision tasks in a priming paradigm, we found that dominant primes (i.e., primes with morphemes encoding dominant meanings) facilitated L2 readers’ recognition of subordinate targets. We suggested that dominant meanings are associated with higher-quality orthographic representations in learners and dominant primes; thus, they facilitate readers’ recognition of orthographically and morphologically related subordinate targets. This study confirmed the role of sublexical constituents’ meaning variables in word recognition in language learning. Full article

Background/Objectives: The present study examines the role of morphemic units in the initial word recognition stage among beginning readers. We assess whether and to what extent sublexical units, such as morphemes, are used in processing French words and how their use varies with reading proficiency. Methods: Two experiments were conducted to investigate the perceptual and morphological effects on the recognition of words presented in central vision, using a variable-viewing-position technique. To explore changes during elementary school years, we tested children from the second and fourth grades, as well as adult readers. Results: The percentage of correct word identification was highest near the center of the word, indicating an optimal viewing position for all three participant groups. Viewing position effects were modulated by age and the properties of the stimuli (length and morphological structure). Experiment 1 demonstrated that lexical decisions are influenced by morphological structure to a decreasing extent as reading skill develops. Experiment 2 revealed that morphological processing in children primarily relies on the orthographic information provided by morphemes (surface morphology), whereas proficient readers process morphological information at a more abstract level, exhibiting a genuine morphological-facilitation effect. Conclusions: Overall, our study strongly indicates that morphemic units play a crucial role in the initial stage of word identification in early reading development. This conclusion aligns with the “word and affix” model, which posits that morphological representations become increasingly independent of orthography as reading ability and word exposure improve. Full article

Decoding semantic concepts for imagination and perception tasks (SCIP) is important for rehabilitation medicine as well as cognitive neuroscience. Electroencephalogram (EEG) is commonly used in the relevant fields, because it is a low-cost noninvasive technique with high temporal resolution. However, as EEG signals contain a high noise level resulting in a low signal-to-noise ratio, it makes decoding EEG-based semantic concepts for imagination and perception tasks (SCIP-EEG) challenging. Currently, neural network algorithms such as CNN, RNN, and LSTM have almost reached their limits in EEG signal decoding due to their own short-comings. The emergence of transformer methods has improved the classification performance of neural networks for EEG signals. However, the transformer model has a large parameter set and high complexity, which is not conducive to the application of BCI. EEG signals have high spatial correlation. The relationship between signals from different electrodes is more complex. Capsule neural networks can effectively model the spatial relationship between electrodes through vector representation and a dynamic routing mechanism. Therefore, it achieves more accurate feature extraction and classification. This paper proposes a spatio-temporal capsule network with a self-correlation routing mechaninsm for the classification of semantic conceptual EEG signals. By improving the feature extraction and routing mechanism, the model is able to more effectively capture the highly variable spatio-temporal features from EEG signals and establish connections between capsules, thereby enhancing classification accuracy and model efficiency. The performance of the proposed model was validated using the publicly accessible semantic concept dataset for imagined and perceived tasks from Bath University. Our model achieved average accuracies of 94.9%, 93.3%, and 78.4% in the three sensory modalities (pictorial, orthographic, and audio), respectively. The overall average accuracy across the three sensory modalities is 88.9%. Compared to existing advanced algorithms, the proposed model achieved state-of-the-art performance, significantly improving classification accuracy. Additionally, the proposed model is more stable and efficient, making it a better decoding solution for SCIP-EEG decoding. Full article

Similarity measures play a pivotal role in automatic techniques designed to analyse large volumes of textual data. Conventional approaches, treating texts as paradigmatic examples of unstructured data, tend to overlook their structural nuances, leading to a loss of valuable information. In this paper, we propose a novel orthographic similarity measure tailored for the semi-structured analysis of texts. We explore a graph-based representation for texts, where the graph’s structure is shaped by a hierarchical decomposition of textual discourse units. Employing the concept of edit distances, our orthographic similarity measure is computed hierarchically across all components in this textual graph, integrating precomputed similarity values among lower-level nodes. The relevance and applicability of the presented approach are illustrated by a real-world example, featuring texts that exhibit intricate interconnections among their components. The resulting similarity scores, between all different structural levels of the graph, allow for a deeper understanding of the (structural) interconnections among texts and enhances the explainability of similarity measures as well as the tools using them. Full article

Psycholinguistic properties such as concept familiarity and concreteness have been investigated in relation to technological innovations in teaching and learning. Due to ongoing advances in semantic representation and machine learning technologies, the automatic extrapolation of lexical psycholinguistic properties has received increased attention across a number of disciplines in recent years. However, little attention has been paid to the reliable and interpretable assessment of familiarity ratings for domain concepts. To address this gap, we present a regression model grounded in advanced natural language processing and interpretable machine learning techniques that can predict domain concepts’ familiarity ratings based on their lexical features. Each domain concept is represented at both the orthographic–phonological level and semantic level by means of pretrained word embedding models. Then, we compare the performance of six tree-based regression models (adaptive boosting, gradient boosting, extreme gradient boosting, a light gradient boosting machine, categorical boosting, and a random forest) on domain concepts’ familiarity rating prediction. Experimental results show that categorical boosting with the lowest MAPE (0.09) and the highest R² value (0.02) is best suited to predicting domain concepts’ familiarity. Experimental results also revealed the prospect of integrating tree-based regression models and interpretable machine learning techniques to expand psycholinguistic resources. Specifically, findings showed that the semantic information of raw words and parts of speech in domain concepts are reliable indicators when predicting familiarity ratings. Our study underlines the importance of leveraging domain concepts’ familiarity ratings; future research should aim to improve familiarity extrapolation methods. Scholars should also investigate the correlation between students’ engagement in online discussions and their familiarity with domain concepts. Full article

In recent studies with the masked priming lexical decision task, matched-case identity-priming effects occur for nonwords but not for words (e.g., nonwords: ERTAR-ERTAR faster than ertar-ERTAR; words: ALTAR-ALTAR produces similar response times as altar-ALTAR). This dissociation is thought to result from lexical feedback influencing orthographic representations in word processing. As nonwords do not receive this feedback, bottom-up processing of prime–target integration leads to matched-case effects. However, the underlying mechanism of this effect in nonwords remains unclear. In this study, we added a color congruency manipulation across the prime and target in the matched-case identity-priming design. We aimed to determine whether the case effects originate at the early stages of prime–target perceptual integration or due to bottom-up activation of case-specific letter detectors. Results replicated the previous dissociation between words and nonwords regarding the matched-case identity effect. Additionally, we did not find any modulation of these effects by prime–target color congruency. These findings suggest that the locus of the matched-case identity effect is at an orthographic level of representation that encodes case information. Full article

Fingerspelling is a critical component of many sign languages. This manual representation of orthographic code is one key way in which signers engage in translanguaging, drawing from all of their linguistic and semiotic resources to support communication. Translanguaging in bimodal bilinguals is unique because it involves drawing from languages in different modalities, namely a signed language like American Sign Language and a spoken language like English (or its written form). Fingerspelling can be seen as a unique product of the unified linguistic system that translanguaging theories purport, as it blends features of both sign and print. The goals of this paper are twofold: to integrate existing research on fingerspelling in order to characterize it as a cognitive-linguistic phenomenon and to discuss the role of fingerspelling in translanguaging and communication. We will first review and synthesize research from linguistics and cognitive neuroscience to summarize our current understanding of fingerspelling, its production, comprehension, and acquisition. We will then discuss how fingerspelling relates to translanguaging theories and how it can be incorporated into translanguaging practices to support literacy and other communication goals. Full article

Reading continues to be a challenging task for most deaf children. Bimodal bilingual education creates a supportive environment that stimulates deaf children’s learning through the use of sign language. However, it is still unclear how exposure to sign language might contribute to improving reading ability. Here, we investigate the relative contribution of several cognitive and linguistic variables to the development of word and text reading fluency in deaf children in bimodal bilingual education programmes. The participants of this study were 62 school-aged (8 to 10 years old at the start of the 3-year study) deaf children who took part in bilingual education (using Dutch and Sign Language of The Netherlands) and 40 age-matched hearing children. We assessed vocabulary knowledge in speech and sign, phonological awareness in speech and sign, receptive fingerspelling ability, and short-term memory at time 1 (T1). At times 2 (T2) and 3 (T3), we assessed word and text reading fluency. We found that (1) speech-based vocabulary strongly predicted word and text reading at T2 and T3, (2) fingerspelling ability was a strong predictor of word and text reading fluency at T2 and T3, (3) speech-based phonological awareness predicted word reading accuracy at T2 and T3 but did not predict text reading fluency, and (4) fingerspelling and STM predicted word reading latency at T2 while sign-based phonological awareness predicted this outcome measure at T3. These results suggest that fingerspelling may have an important function in facilitating the construction of orthographical/phonological representations of printed words for deaf children and strengthening word decoding and recognition abilities. Full article

Background: The present study aims to investigate the effect of the first language (L1) orthography on the perception of the second language (L2) vowel contrasts and whether orthographic effects occur at the sublexical level. Methods: Fourteen adult Greek learners of English participated in two AXB discrimination tests: one auditory and one orthography test. In the auditory test, participants listened to triads of auditory stimuli that targeted specific English vowel contrasts embedded in nonsense words and were asked to decide if the middle vowel was the same as the first or the third vowel by clicking on the corresponding labels. The orthography test followed the same procedure as the auditory test, but instead, the two labels contained grapheme representations of the target vowel contrasts. Results: All but one vowel contrast could be more accurately discriminated in the auditory than in the orthography test. The use of nonsense words in the elicitation task eradicated the possibility of a lexical effect of orthography on auditory processing, leaving space for the interpretation of this effect on a sublexical basis, primarily prelexical and secondarily postlexical. Conclusions: L2 auditory processing is subject to L1 orthography influence. Speakers of languages with transparent orthographies such as Greek may rely on the grapheme–phoneme correspondence to decode orthographic representations of sounds coming from languages with an opaque orthographic system such as English. Full article

English serves as today’s lingua franca, a role not eased by the inconsistency of its orthography. Indeed, monolingual readers of more consistent orthographies such as Italian or German learn to read more quickly than monolingual English readers. Here, we assessed whether long-lasting bilingualism would mitigate orthography-specific differences in reading speed and whether the order in which orthographies with a different regularity are learned matters. We studied high-proficiency Italian-English and English-Italian bilinguals, with at least 20 years of intensive daily exposure to the second language and its orthography and we simulated sequential learning of the two orthographies with the CDP++ connectionist model of reading. We found that group differences in reading speed were comparatively bigger with Italian stimuli than with English stimuli. Furthermore, only Italian bilinguals took advantage of a blocked presentation of Italian stimuli compared to when stimuli from both languages were presented in mixed order, suggesting a greater ability to keep language-specific orthographic representations segregated. These findings demonstrate orthographic constraints on bilingual reading, whereby the level of consistency of the first learned orthography affects later learning and performance on a second orthography. The computer simulations were consistent with these conclusions. Full article

The acquisition of new vocabulary is usually mediated by previous experience with language. In the visual domain, the representation of orthographically unfamiliar words at the phonological or conceptual levels may facilitate their orthographic learning. The neural correlates of this advantage were investigated by recording EEG activity during reading novel and familiar words across three different experiments (n = 22 each), manipulating the availability of previous knowledge on the novel written words. A different pattern of event-related potential (ERP) responses was found depending on the previous training, resembling cross-level top-down interactive effects during vocabulary acquisition. Thus, whereas previous phonological experience caused a modulation at the post-lexical stages of the visual recognition of novel written words (~520 ms), additional semantic training influenced their processing at a lexico-semantic stage (~320 ms). Moreover, early lexical differences (~180 ms) elicited in the absence of previous training did not emerge after both phonological and semantic training, reflecting similar orthographic processing and word-form access. Full article

Collecting 3D point cloud data of buildings is important for many applications such as urban mapping, renovation, preservation, and energy simulation. However, laser-scanned point clouds are often difficult to analyze, visualize, and interpret due to incompletely scanned building facades caused by numerous sources of defects such as noise, occlusions, and moving objects. Several point cloud scene completion algorithms have been proposed in the literature, but they have been mostly applied to individual objects or small-scale indoor environments and not on large-scale scans of building facades. This paper introduces a method of performing point cloud scene completion of building facades using orthographic projection and generative adversarial inpainting methods. The point cloud is first converted into the 2D structured representation of depth and color images using an orthographic projection approach. Then, a data-driven 2D inpainting approach is used to predict the complete version of the scene, given the incomplete scene in the image domain. The 2D inpainting process is fully automated and uses a customized generative-adversarial network based on Pix2Pix that is trainable end-to-end. The inpainted 2D image is finally converted back into a 3D point cloud using depth remapping. The proposed method is compared against several baseline methods, including geometric methods such as Poisson reconstruction and hole-filling, as well as learning-based methods such as the point completion network (PCN) and TopNet. Performance evaluation is carried out based on the task of reconstructing real-world building facades from partial laser-scanned point clouds. Experimental results using the performance metrics of voxel precision, voxel recall, position error, and color error showed that the proposed method has the best performance overall. Full article

Event-related potentials (ERPs) were used to investigate co-activation of English words during recognition of American Sign Language (ASL) signs. Deaf and hearing signers viewed pairs of ASL signs and judged their semantic relatedness. Half of the semantically unrelated signs had English translations that shared an orthographic and phonological rime (e.g., BAR–STAR) and half did not (e.g., NURSE–STAR). Classic N400 and behavioral semantic priming effects were observed in both groups. For hearing signers, targets in sign pairs with English rime translations elicited a smaller N400 compared to targets in pairs with unrelated English translations. In contrast, a reversed N400 effect was observed for deaf signers: target signs in English rime translation pairs elicited a larger N400 compared to targets in pairs with unrelated English translations. This reversed effect was overtaken by a later, more typical ERP priming effect for deaf signers who were aware of the manipulation. These findings provide evidence that implicit language co-activation in bimodal bilinguals is bidirectional. However, the distinct pattern of effects in deaf and hearing signers suggests that it may be modulated by differences in language proficiency and dominance as well as by asymmetric reliance on orthographic versus phonological representations. Full article