Semantic priming typically facilitates lexical access; however, this facilitation may reverse into inhibition under certain interferences, such as embodied conflicts in Second Language (L2) processing. For adult L2 learners, automatic literal sensorimotor simulations may disrupt metaphorical integration, potentially inducing inhibitory reversal in conventional action metaphors [e.g., Chinese “吃亏 “(chī kuī, literally “eat loss,” figuratively “suffer the loss”)]—a key gap in bilingual cognition. This study examines this reversal in L2 Chinese metaphor processing versus L1. Forty-three Vietnamese-speaking L2 learners of Chinese (HSK 5–6) and forty-seven first-language (L1) Chinese speakers completed a delayed-response semantic plausibility judgment task with morpheme-mediated semantic priming. Targets—literal, conventional metaphorical, and unrelated verb-object (VO) constructions—were each preceded by their identical verb morpheme (e.g., “吃” primes “吃亏”). Mixed-effects models revealed an opposite directional reversal in L2 learners: facilitation in literal versus unrelated baseline (shorter reaction times [RTs]; reduced errors) but inhibition in metaphorical versus unrelated baseline (elevated errors; nonsignificant RTs). In contrast to L2’s reversal pattern, L1 Chinese speakers exhibited uniform dual inhibition across literal and metaphorical conditions (elevated errors; nonsignificant RTs), with a significant Group × Condition interaction. This study reveals an L2-specific reversal of priming in action metaphors (literal facilitation vs. metaphorical inhibition, primarily evident in error rates), originating from a dynamic mismatch between embodied simulations and semantic integration, a process potentially involving increased inhibitory control demands, while remaining consistent with broader processing costs at the behavioral level. These findings offer insights into the double-edged role of L2 embodiment in language processing -- helping Literal while hurting Metaphorical, providing implications for theories of embodied cognition and bilingualism and also informing practical pedagogy in L2 acquisition.
This study investigates the intersection and divergence between Chinese fine brushwork flower-and-bird paintings (gongbi hua) and European oil paintings, focusing on their representation of nature and symbolic communication. It examines shared thematic elements and aesthetic principles, such as the pursuit of harmony and beauty, while acknowledging distinct approaches shaped by differing cultural contexts. The research employs a comparative analysis methodology, scrutinizing composition, perspective, and material application in both traditions. Findings reveal that while both art forms explore similar subject matter, their execution and underlying philosophies diverge significantly. Chinese paintings emphasize meticulous detail and symbolic representation, whereas European oil paintings prioritize realism and the exploration of light and shadow. The analysis further explores how cultural values influence artistic choices, shaping the interpretation and appreciation of these art forms. The study contributes to a deeper understanding of cross-cultural artistic expression and provides insights into the diverse ways in which humans perceive and represent the natural world. It highlights the importance of considering both universal artistic concerns and culturally specific techniques in art historical analysis.
PurposeThis paper investigates how high-speed rail (HSR) influences socioeconomic inequality by providing the first systematic bibliometric review of research trends, methodological approaches and thematic structures. It examines whether HSR fosters balanced regional development or reinforces spatial disparities.Design/methodology/approachUsing the Bibliometrix R package, 237 records were retrieved from the Web of Science (1985–2024). Citation indicators, keyword co-occurrence and collaboration networks were combined with natural language processing (NLP) to classify studies by territorial scale, methodology, economic variables and inequality outcomes.FindingsThe paper offers the first structured overview of how the literature conceptualizes the link between HSR and inequality. It highlights persistent gaps – scarcity of city-level analyses, limited socioeconomic indicators and reliance on Chinese case studies – providing a foundation for more comparative and interdisciplinary research.Originality/valueThis paper contributes by offering a structured overview of how the literature has conceptualized and measured the relationship between HSR and inequality. By identifying persistent research gaps – such as the scarcity of city-level analyses, limited use of socioeconomic indicators, and overreliance on Chinese case studies – it provides a foundation for more comparative and interdisciplinary approaches. The study informs policymakers and researchers on how to design future infrastructure projects that balance efficiency with equity.
Transportation engineering, Railroad engineering and operation
Huaiyuan Mandarin is a Mandarin dialect that has three low-tone sandhi rules. T1 (low-falling) and T3 (low falling-rising) sandhis involve changing the first low tone to a mid-rising tone when two low tones occur consecutively, which may lead to neutralization between the sandhi tones and T2 (mid-rising). Huaiyuan half-third sandhi involves abridging the rising portion of the first T3 when it is followed by a non-low tone, which may bring about neutralization between half-T3 and T1. Given the complexity, this study investigates the tonal neutralization between Huaiyuan sandhi tones and their corresponding non-sandhi tones in disyllabic words. For each tone sandhi, 10 sandhi words and 10 non-sandhi words differing only in the first underlying tones were compared. Acoustic and identification results showed that sandhi-T3 and T2 were completely neutralized, while sandhi-T1 and T2, as well as half-T3 and T1 were not neutralized in production or perception. Discrimination results revealed that native listeners outperformed non-Huaiyuan listeners in differentiating between half-T3 and T1, suggesting higher-level linguistic knowledge was used by the native listeners for the perception of Huaiyuan sandhi words. These results indicate different degrees of phonetic and phonological motivations among the three low-tone sandhis in Huaiyuan Mandarin.
Driven by the “Chinese culture going global” strategy and the Belt and Road Initiative, cultural instruction has increasingly become a crucial component of international Chinese language education. Based on 487 CSSCI and core journal articles from the CNKI database published between 2016 and 2025, this study employs CiteSpace 6.3.R1 to analyze the research hotspots and evolutionary paths of this field from the perspectives of keyword co-occurrence, clustering, burst detection, and author–institution collaboration. The findings reveal that the research focus in cultural teaching has shifted from “traditional cultural dissemination” to “intercultural integration,” with key hotspots centering on tea culture, Confucius Institutes, and professional degree programs. Collaborative networks show a growing trend toward clustering, and several keywords exhibit strong burst intensity and long duration, indicating sustained academic attention in this field. This study helps clarify the knowledge structure of cultural instruction and provides theoretical support and practical guidance for future topic development, paradigm construction, and pedagogical innovation.
Universal phoneme recognition typically requires analyzing long speech segments and language-specific patterns. Many speech processing tasks require pure phoneme representations free from contextual influence, which motivated our development of CUPE - a lightweight model that captures key phoneme features in just 120 milliseconds, about one phoneme's length. CUPE processes short, fixed-width windows independently and, despite fewer parameters than current approaches, achieves competitive cross-lingual performance by learning fundamental acoustic patterns common to all languages. Our extensive evaluation through supervised and self-supervised training on diverse languages, including zero-shot tests on the UCLA Phonetic Corpus, demonstrates strong cross-lingual generalization and reveals that effective universal speech processing is possible through modeling basic acoustic patterns within phoneme-length windows.
Toheeb Aduramomi Jimoh, Tabea De Wille, Nikola S. Nikolov
Natural Language Processing (NLP) is becoming a dominant subset of artificial intelligence as the need to help machines understand human language looks indispensable. Several NLP applications are ubiquitous, partly due to the myriad of datasets being churned out daily through mediums like social networking sites. However, the growing development has not been evident in most African languages due to the persisting resource limitations, among other issues. Yorùbá language, a tonal and morphologically rich African language, suffers a similar fate, resulting in limited NLP usage. To encourage further research towards improving this situation, this systematic literature review aims to comprehensively analyse studies addressing NLP development for Yorùbá, identifying challenges, resources, techniques, and applications. A well-defined search string from a structured protocol was employed to search, select, and analyse 105 primary studies between 2014 and 2024 from reputable databases. The review highlights the scarcity of annotated corpora, the limited availability of pre-trained language models, and linguistic challenges like tonal complexity and diacritic dependency as significant obstacles. It also revealed the prominent techniques, including rule-based methods, among others. The findings reveal a growing body of multilingual and monolingual resources, even though the field is constrained by socio-cultural factors such as code-switching and the desertion of language for digital usage. This review synthesises existing research, providing a foundation for advancing NLP for Yorùbá and in African languages generally. It aims to guide future research by identifying gaps and opportunities, thereby contributing to the broader inclusion of Yorùbá and other under-resourced African languages in global NLP advancements.
Obtaining valuable information from massive data efficiently has become our research goal in the era of Big Data. Text summarization technology has been continuously developed to meet this demand. Recent work has also shown that transformer-based pre-trained language models have achieved great success on various tasks in Natural Language Processing (NLP). Aiming at the problem of Chinese news text summary generation and the application of Transformer structure on Chinese, this paper proposes a Chinese news text summarization model (CNsum) based on Transformer structure, and tests it on Chinese datasets such as THUCNews. The results of the conducted experiments show that CNsum achieves better ROUGE score than the baseline models, which verifies the outperformance of the model.
Framing is an essential device in news reporting, allowing the writer to influence public perceptions of current affairs. While there are existing automatic news framing detection datasets in various languages, none of them focus on news framing in the Chinese language which has complex character meanings and unique linguistic features. This study introduces the first Chinese News Framing dataset, to be used as either a stand-alone dataset or a supplementary resource to the SemEval-2023 task 3 dataset. We detail its creation and we run baseline experiments to highlight the need for such a dataset and create benchmarks for future research, providing results obtained through fine-tuning XLM-RoBERTa-Base and using GPT-4o in the zero-shot setting. We find that GPT-4o performs significantly worse than fine-tuned XLM-RoBERTa across all languages. For the Chinese language, we obtain an F1-micro (the performance metric for SemEval task 3, subtask 2) score of 0.719 using only samples from our Chinese News Framing dataset and a score of 0.753 when we augment the SemEval dataset with Chinese news framing samples. With positive news frame detection results, this dataset is a valuable resource for detecting news frames in the Chinese language and is a valuable supplement to the SemEval-2023 task 3 dataset.
The fusion of language and vision in large vision-language models (LVLMs) has revolutionized deep learning-based object detection by enhancing adaptability, contextual reasoning, and generalization beyond traditional architectures. This in-depth review presents a structured exploration of the state-of-the-art in LVLMs, systematically organized through a three-step research review process. First, we discuss the functioning of vision language models (VLMs) for object detection, describing how these models harness natural language processing (NLP) and computer vision (CV) techniques to revolutionize object detection and localization. We then explain the architectural innovations, training paradigms, and output flexibility of recent LVLMs for object detection, highlighting how they achieve advanced contextual understanding for object detection. The review thoroughly examines the approaches used in integration of visual and textual information, demonstrating the progress made in object detection using VLMs that facilitate more sophisticated object detection and localization strategies. This review presents comprehensive visualizations demonstrating LVLMs' effectiveness in diverse scenarios including localization and segmentation, and then compares their real-time performance, adaptability, and complexity to traditional deep learning systems. Based on the review, its is expected that LVLMs will soon meet or surpass the performance of conventional methods in object detection. The review also identifies a few major limitations of the current LVLM modes, proposes solutions to address those challenges, and presents a clear roadmap for the future advancement in this field. We conclude, based on this study, that the recent advancement in LVLMs have made and will continue to make a transformative impact on object detection and robotic applications in the future.
Recently, the Large Language Model-based Phoneme-to-Grapheme (LLM-P2G) method has shown excellent performance in speech recognition tasks and has become a feasible direction to replace the traditional WFST decoding method. This framework takes into account both recognition accuracy and system scalability through two-stage modeling of phoneme prediction and text generation. However, the existing LLM-P2G adopts the Top-K Marginalized (TKM) training strategy, and its candidate phoneme sequences rely on beam search generation, which has problems such as insufficient path diversity, low training efficiency, and high resource overhead. To this end, this paper proposes a sampling marginalized training strategy (Sampling-K Marginalized, SKM), which replaces beam search with random sampling to generate candidate paths, improving marginalized modeling and training efficiency. Experiments were conducted on Polish and German datasets, and the results showed that SKM further improved the model learning convergence speed and recognition performance while maintaining the complexity of the model. Comparative experiments with a speech recognition method that uses a projector combined with a large language model (SpeechLLM) also show that the SKM-driven LLM-P2G has more advantages in recognition accuracy and structural simplicity. The study verified the practical value and application potential of this method in cross-language speech recognition systems.
Abstract In recent years, there has been an increasing focus on women’s science fiction in China. A prevailing perception among readers and critics suggests that women’s sensibilities enable them to convey more nuanced emotions in their works. To examine this viewpoint within the realm of contemporary Chinese science fiction, a quantitative approach based on affective computing was employed. This approach allowed for a systematic evaluation of indicators such as emotional arc, emotional richness, and twistiness. The findings reveal that while individual writers may exhibit distinct emotional writing styles, overall, there is no significant disparity in emotional narratives between male and female science fiction writers.
History of scholarship and learning. The humanities, Social Sciences
BackgroundLung cancer, characterized by its high morbidity and lethality, necessitates thorough research to enhance our understanding of its pathogenesis and discover novel therapeutic approaches. Recent studies increasingly demonstrate that lung cancer cells can modulate the tumor microenvironment, promoting tumor growth, and metastasis through the release of exosomes. Exosomes are small vesicles secreted by cells and contain a variety of bioactive molecules such as proteins, nucleic acids, and metabolites. This paper presents a comprehensive review of exosome research in lung cancer and its progress through bibliometric analysis.MethodsPublications related to exosomes in lung cancer patients were systematically searched on the Web of Science Core Collection (WoSCC) database. Bibliometric analysis was performed using VOSviwers, CiteSpace, and the R package “Bibliometrics”. Publications were quantitatively analyzed using Microsoft Office Excel 2019. The language of publication was restricted to “English” and the search strategy employed TS=(exosomes or exosomes or exosomes) and TS=(lung cancer). The search period commenced on January 1, 2004, and concluded on November 12, 2023, at noon. The selected literature types included Articles and Reviews.ResultsThe study encompassed 1699 papers from 521 journals across 71 countries and 2105 institutions. Analysis revealed a consistent upward trend in lung cancer exosome research over the years, with a notable surge in recent times. This surge indicates a growing interest and depth of inquiry into lung cancer exosomes. Major research institutions in China and the United States, including Nanjing Medical University, Shanghai Jiao Tong University, Chinese Academy Of Sciences, and Utmd Anderson Cancer Center, emerged as crucial research hubs. The annual publication count in this field witnessed a continuous rise, particularly in recent years. Key terms such as lung cancer, non-small cell lung cancer (NSCLC), microvesicles, intercellular communication, exosomal miRNAs, and oncology dominated the research landscape. Fields like cell biology, biochemistry, biotechnology, and oncology exhibited close relation with this research. Clotilde Théry emerged as the most cited author in the field, underlining her significant contributions. These results demonstrate the broad impact of exosome research in lung cancer, with key terms covering not only disease-specific aspects such as lung cancer and NSCLC but also basic biological concepts like microvesicles and intercellular communication. Explorations into exosomal microRNAs and oncology have opened new avenues for lung cancer exosome research. In summary, lung cancer exosome research is poised to continue receiving attention, potentially leading to breakthroughs in treatment and prevention.ConclusionPublications on lung cancer exosomes show a rising trend year by year, with China and the United States ranking first and second in terms of the number of publications. However, there is insufficient academic learning cooperation and exchanges between the two sides, and Chinese universities account for a large proportion of research institutions in this field. Jing Li is the most productive author, Clotilde Théry is the most co-cited author, and Cancers is the journal with the highest number of publications. The current focus in the field of lung cancer exosomes is on biomarkers, liquid biopsies, immunotherapy, and tumor microenvironment.
Neoplasms. Tumors. Oncology. Including cancer and carcinogens
Generating emotionally appropriate responses in conversations with large language models presents a significant challenge due to the complexities of human emotions and cognitive processes, which remain largely underexplored in their critical role in social interactions. In this study, we introduce a two-stage automatic data generation framework to create CAPE, a Chinese dataset named Cognitive Appraisal theory-based Emotional corpus. This corpus facilitates the generation of dialogues with contextually appropriate emotional responses by accounting for diverse personal and situational factors. We propose two tasks utilizing this dataset: emotion prediction and next utterance prediction. Both automated and human evaluations demonstrate that agents trained on our dataset can deliver responses that are more aligned with human emotional expressions. Our study shows the potential for advancing emotional expression in conversational agents, paving the way for more nuanced and meaningful human-computer interactions.
Sign language serves as the primary meaning of communication for the deaf-mute community. Different from spoken language, it commonly conveys information by the collaboration of manual features, i.e., hand gestures and body movements, and non-manual features, i.e., facial expressions and mouth cues. To facilitate communication between the deaf-mute and hearing people, a series of sign language understanding (SLU) tasks have been studied in recent years, including isolated/continuous sign language recognition (ISLR/CSLR), gloss-free sign language translation (GF-SLT) and sign language retrieval (SL-RT). Sign language recognition and translation aims to understand the semantic meaning conveyed by sign languages from gloss-level and sentence-level, respectively. In contrast, SL-RT focuses on retrieving sign videos or corresponding texts from a closed-set under the query-by-example search paradigm. These tasks investigate sign language topics from diverse perspectives and raise challenges in learning effective representation of sign language videos. To advance the development of sign language understanding, exploring a generalized model that is applicable across various SLU tasks is a profound research direction.
Mohamed Bayan Kmainasi, Rakif Khan, Ali Ezzat Shahroor
et al.
Large language models (LLMs) have shown remarkable abilities in different fields, including standard Natural Language Processing (NLP) tasks. To elicit knowledge from LLMs, prompts play a key role, consisting of natural language instructions. Most open and closed source LLMs are trained on available labeled and unlabeled resources--digital content such as text, images, audio, and videos. Hence, these models have better knowledge for high-resourced languages but struggle with low-resourced languages. Since prompts play a crucial role in understanding their capabilities, the language used for prompts remains an important research question. Although there has been significant research in this area, it is still limited, and less has been explored for medium to low-resourced languages. In this study, we investigate different prompting strategies (native vs. non-native) on 11 different NLP tasks associated with 12 different Arabic datasets (9.7K data points). In total, we conducted 197 experiments involving 3 LLMs, 12 datasets, and 3 prompting strategies. Our findings suggest that, on average, the non-native prompt performs the best, followed by mixed and native prompts.
While Large Language Models (LLMs) have achieved remarkable success across various applications, they also raise concerns regarding self-cognition. In this paper, we perform a pioneering study to explore self-cognition in LLMs. Specifically, we first construct a pool of self-cognition instruction prompts to evaluate where an LLM exhibits self-cognition and four well-designed principles to quantify LLMs' self-cognition. Our study reveals that 4 of the 48 models on Chatbot Arena--specifically Command R, Claude3-Opus, Llama-3-70b-Instruct, and Reka-core--demonstrate some level of detectable self-cognition. We observe a positive correlation between model size, training data quality, and self-cognition level. Additionally, we also explore the utility and trustworthiness of LLM in the self-cognition state, revealing that the self-cognition state enhances some specific tasks such as creative writing and exaggeration. We believe that our work can serve as an inspiration for further research to study the self-cognition in LLMs.
Language models have shown unprecedented capabilities, sparking debate over the source of their performance. Is it merely the outcome of learning syntactic patterns and surface level statistics, or do they extract semantics and a world model from the text? Prior work by Li et al. investigated this by training a GPT model on synthetic, randomly generated Othello games and found that the model learned an internal representation of the board state. We extend this work into the more complex domain of chess, training on real games and investigating our model's internal representations using linear probes and contrastive activations. The model is given no a priori knowledge of the game and is solely trained on next character prediction, yet we find evidence of internal representations of board state. We validate these internal representations by using them to make interventions on the model's activations and edit its internal board state. Unlike Li et al's prior synthetic dataset approach, our analysis finds that the model also learns to estimate latent variables like player skill to better predict the next character. We derive a player skill vector and add it to the model, improving the model's win rate by up to 2.6 times.
Asha-Naima Ferrante,1,2 Barbara K Keller,1 Julian S Flury,1 Michael A Harnik,3 Martin grosse Holtforth,2,4 Maria M Wertli1,5 1Department of General Internal Medicine, University Hospital of Bern, Inselspital, University of Bern, Bern, 3010, Switzerland; 2Department of Psychology, University of Bern, Bern, 3012, Switzerland; 3Department of Anaesthesiology and Pain Medicine, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland; 4Psychosomatic Competence Center, University Hospital of Bern, Inselspital, University of Bern, Bern, 3010, Switzerland; 5Department of Internal Medicine, Kantonsspital Baden, Baden, 5404, SwitzerlandCorrespondence: Asha-Naima Ferrante, Department of General Internal Medicine, University Hospital, Inselspital, Freiburgstrasse 18, Bern, 3010, Switzerland, Email asha-naima.ferrante@students.unibe.chPurpose: To describe the details of a systematic review to assess the current evidence about the efficacy of communication strategies on the prevention of chronic postsurgical pain (CPSP).Methods: The protocol for this systematic review was based on the Cochrane Handbook methodology and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P) recommendations. A systematic search of the literature on electronic databases Medline, Embase, Cochrane Library, CINAHL, PsycINFO, and Web of Science (from the inception to 19 June 2022) was carried out using predefined search terms to identify relevant studies. This review will include randomized clinical trials or observational studies. The search strategy consisted of keywords and index terms related to “clinician”, “communication” or “post-surgical pain”. Inclusion criteria are as follows: randomized clinical trials or observational studies using a parallel group design that assess the efficacy of communication interventions in patients undergoing surgery and that assess pain and pain-related disability. We considered interventions that included any type of written, verbal, and non-verbal communication in combination with other interventions or without. Control groups may include no communication intervention or another intervention distinctly different. We excluded studies with follow-up duration of less than 3 months, patients aged < 18 years, and studies for which no reviewer had language proficiency (eg, Chinese, Korean). Descriptive statistics will be used to summarize quantitative findings. Meta-analysis will only be considered if at least three studies used the same outcome with comparable interventions, as we expect a wide heterogeneity of study population and settings.Conclusion: This systematic review and meta-analysis will be an important source for clinicians and researchers to understand the influence of communication to prevent CPSP.Study Registration: This protocol is registered with the International Prospective Register of Systematic Reviews (PROSPERO). Registration number: CRD42021241596.Keywords: chronic post-surgical pain, communication, education, pain prevention, systematic review