Il s'agit de notre dernière sélection de publications et de brevets mondiaux en anglais sur le traitement du langage naturel (NLP), parmi de nombreuses revues scientifiques en ligne, classées et axées sur le traitement du langage naturel, la tokenisation, le stemming, la lemmatisation, la partie de discours, la reconnaissance des entités nommées et l'analyse des sentiments.
The Role of Technology and Quality
Published on 2025-04-09 by Xiomara Z��iga-Santill�n, Diego Tapia-N��ez, Rosa Espinoza-Toalombo, Erika Romero-C�rdenas, Edwuin Carrasquero-Rodr�guez @MDPI
Abstract: This study evaluates the competitiveness of small- and medium-sized manufacturing enterprises (PYMES) through a meta-analysis that explores the role of technology and quality. Through a comprehensive literature review, 24 eligible studies, selected after applying specific criteria in a systematic search of the Scopus database, were identified and analyzed. A random-effects model was used to combine the effect sizes of the selected studies, and the heterogeneity among them was assessed using wide[...]
Our summary: The Role of Technology and Quality: Meta-analysis of 24 studies on small- and medium-sized manufacturing enterprises explores the impact of technology and quality on competitiveness. Analysis reveals positive relationship between innovation and competitiveness, with quality identified as top priority and state-of-the-art technologies as a risk factor in digital transformation.
technology, quality, competitiveness, innovation
Publication
Graph-to-Text Generation with Bidirectional Dual Cross-Attention and Concatenation
Published on 2025-03-11 by Elias Lemuye Jimale, Wenyu Chen, Mugahed A. Al-antari, Yeong Hyeon Gu, Victor Kwaku Agbesi, Wasif Feroze, Feidu Akmel, Juhar Mohammed Assefa, Ali Shahzad @MDPI
Abstract: Graph-to-text generation (G2T) involves converting structured graph data into natural language text, a task made challenging by the need for encoders to capture the entities and their relationships within the graph effectively. While transformer-based encoders have advanced natural language processing, their reliance on linearized data often obscures the complex interrelationships in graph structures, leading to structural loss. Conversely, graph attention networks excel at capturing graph struc[...]
Our summary: Proposal of a novel mechanism for integrating transformer-based and graph attention encoders to improve graph-to-text generation tasks, achieving higher BLEU and METEOR scores on benchmark datasets, showcasing potential for future research.
Graph-to-text generation, Bidirectional Dual Cross-Attention, Concatenation, Transformer-based encoders
Publication
The Rise of Reasoning Large Language Models for Consumer Complaint Detection and Classification
Published on 2025-03-07 by Konstantinos I. Roumeliotis, Nikolaos D. Tselikas, Dimitrios K. Nasiopoulos @MDPI
Abstract: Large language models (LLMs) have demonstrated remarkable capabilities in various natural language processing (NLP) tasks, but their effectiveness in real-world consumer complaint classification without fine-tuning remains uncertain. Zero-shot classification offers a promising solution by enabling models to categorize consumer complaints without prior exposure to labeled training data, making it valuable for handling emerging issues and dynamic complaint categories in finance. However, this task[...]
Our summary: The study evaluates the performance of leading large language models and reasoning models for zero-shot classification of consumer complaints in finance, highlighting their strengths and limitations. Integration of reasoning models into classification workflows can enhance complaint resolution automation and improve customer service efficiency.
reasoning models, large language models, zero-shot classification, consumer complaints
Publication
Perspectives Through an Academic Lens
Published on 2025-03-06 by Iulian �ntorsureanu, Simona-Vasilica Oprea, Adela B�ra, Drago? Vespan @MDPI
Abstract: In this paper, we investigated the role of generative AI in education in academic publications extracted from Web of Science (3506 records; 2019–2024). The proposed methodology included three main streams: (1) Monthly analysis trends; top-ranking research areas, keywords and universities; frequency of keywords over time; a keyword co-occurrence map; collaboration networks; and a Sankey diagram illustrating the relationship between AI-related terms, publication years and research ar[...]
Our summary: Role of generative AI in education analyzed through academic publications from Web of Science, including trends, sentiment analysis, and topic modeling. AI applications in specialized fields and emerging topics identified.
generative AI, education, academic publications, Web of Science
Publication
A Contextual and Transformer-Based Approach for Improved Detection
Published on 2025-03-06 by Parul Dubey, Pushkar Dubey, Pitshou N. Bokoro @MDPI
Abstract: Sarcasm detection is a crucial task in natural language processing (NLP), particularly in sentiment analysis and opinion mining, where sarcasm can distort sentiment interpretation. Accurately identifying sarcasm remains challenging due to its context-dependent nature and linguistic complexity across informal text sources like social media and conversational dialogues. This study utilizes three benchmark datasets, namely, News Headlines, Mustard, and Reddit (SARC), which contain diverse sarcastic[...]
Our summary: Utilizing transformer-based models and contextual summarization to enhance sarcasm detection accuracy and efficiency. Leveraging benchmark datasets to address the challenges of identifying sarcasm in informal text sources. Performance evaluation based on accuracy, F1 score, and Jaccard coefficient showcasing the effectiveness of the proposed methodology.
Transformer-Based, Sarcasm Detection, Contextual Approach, Improved Detection
Publication
Voice-based conversation artificial intelligence for point-of-sale systems
Patent published on the 2025-03-06 in WO under Ref WO2025049246 by PREDICTSPRING INC [US] (Mangtani Nitin [us], Garimella Sandilya [us], Chadalawada Viswanth [us], Sahu Madhav [us])
Abstract: A conversational artificial intelligence (Al) point-of-sale (POS) system uses generative Al to help a store associate complete tasks in a retail environment. This POS system combines four technologies: speech recognition; natural language processing; acting in response to verbal command in the context of a retail store and a POS device; and providing bi-directional and fully conversational responses. The Al, which runs in an app on a smartphone, tablet, or other mobile device, uses a machine lea[...]
Our summary: Conversational AI POS system uses generative AI to assist store associates in retail tasks, combining speech recognition, natural language processing, and bi-directional responses. The AI, running on mobile devices, utilizes ML models trained on support tickets and product documentation to understand dependencies and provide appropriate responses.
voice-based conversation, artificial intelligence, point-of-sale systems, generative AI
Patent
Image interpretation model development
Patent published on the 2025-03-06 in WO under Ref WO2025048865 by SYNTHESIS HEALTH INC [CA] (Reicher Murray [ca], Kaura Deepak [ca])
Abstract: An image classification model, e.g., a neural network model, may be trained on a set of training medical imaging exams each including a training report and a training medical image. A model generation module or device may, for each of the training medical imaging exams: use finding item criteria to reorganize text of the training report into a list of finding items, each associated with text extracted from the training report text, use natural language processing to analyze the resultant text as[...]
Our summary: Training a neural network model on medical imaging exams with associated finding items and classifications extracted from training reports.
image interpretation, model development, neural network, medical imaging
Patent
A Radical-Based Token Representation Method for Enhancing Chinese Pre-Trained Language Models
Published on 2025-03-05 by Honglun Qin, Meiwen Li, Lin Wang, Youming Ge, Junlong Zhu, Ruijuan Zheng @MDPI
Abstract: In the domain of natural language processing (NLP), a primary challenge pertains to the process of Chinese tokenization, which remains challenging due to the lack of explicit word boundaries in written Chinese. The existing tokenization methods often treat each Chinese character as an indivisible unit, neglecting the finer semantic features embedded in the characters, such as radicals. To tackle this issue, we propose a novel token representation method that integrates radical-based features int[...]
Our summary: Radical-based token representation method improves Chinese language model performance by incorporating radical features, enhancing accuracy on multiple NLP tasks.
Tokenization, Chinese, Radical-based, Language models
Publication
FR 
EN 
DE 
ES 
IT