The increasing use of regional languages in digital communication has intensified the need for accurate, context-aware spelling detection tools that can handle the language’s rich morphology and complex orthographic patterns. This paper presents a transformer-based framework for Assamese spelling error detection, utilizing three multilingual pre-trained models—IndicBERT v2, XLM-RoBERTa, and mBERT—fine-tuned on a curated dataset of 115,925 word-level samples. Each word is represented as a tuple (word, label, corrected_word), enabling both binary error detection and lexicon-assisted correction. The models were trained using standardised preprocessing, subword tokenisation, and a weighted loss function to address class imbalance. Experimental results demonstrate that XLM-RoBERTa achieves the highest detection performance, with 94.84
High-utility co-location patterns (HUCPs) can effectively reveal valuable knowledge about spatial features and instances hidden in spatial datasets. However, current HUCPs do not consider user reference features. For example, users may be concerned with which facilities within 500 m of a school form a HUCP, but they are not concerned with the utility of the school itself. Therefore, the utility of the school should not be considered when calculating the utility value of the pattern. To address the shortcomings of traditional HUCPs, this work proposes mining high-utility core co-location patterns (HUCCPs), which treat user reference features as core features. Measuring the interestingness of HUCCPs still does not satisfy the downward closure property. If the generating-testing candidate-based methods are used, the candidate search space will be large, resulting in low mining efficiency. This paper adopts a clique-query-based mining method to improve mining efficiency. First, maximal cliques are enumerated from the neighboring instances graph, and then the maximal cliques are combined into a compact hash structure. The keys of this hash serve as the initial candidates. Only those candidates that contain core features are considered. The utility value of a candidate is calculated by querying the hash value, thereby improving the mining efficiency. Extensive experiments were conducted on the real Yelp dataset, and the experimental results demonstrate the effectiveness and efficiency of the proposed method.
Federated data mining has recently been proposed to securely and effectively mine data from multiple sources in the era of big data. In this paper, we address the federated utility mining problem within a horizontal framework comprising multiple clients and a server. We adopt the concept of the two-phase utility mining algorithm and consider the federated mining property to propose a federated two-phase high-utility itemset mining algorithm for discovering globally high-utility itemsets across multiple clients. According to the user-defined utility threshold, each client identifies high transaction-weighted utility itemsets as local candidate itemsets, as the original two-phase utility mining algorithm does, and sends them to the server. The server then aggregates the uploaded local candidate itemsets to determine global candidates, retrieves their actual utility values from clients, and identifies global high-utility itemsets. Thus, each client shares only candidate itemsets and their utility values with the server, rather than the entire database. The transmission can also be encrypted. The proposed approach thus preserves data privacy during the mining process. The results of the numerical experiments demonstrate that the proposed algorithm can remove unnecessary candidates and produce results nearly identical to those of centralized utility mining.
Sequential pattern mining aims to discover frequent ordered patterns in sequence databases, however, treating all items equally ignores their varying importance. Weighted sequential pattern mining (WSPM) addresses this limitation, yet existing algorithms incur high computational costs and severe memory overhead, especially at low support thresholds. We propose WSUI (Weighted Sequential pattern mining Using Index structures), a scalable algorithm that combines compact index structures with tight upper-bound pruning. WSUI introduces frequency-based automatic weighting to eliminate subjective weight assignment, extends the Data-IDList structure with sequence-level metadata for efficient evaluation, and employs a memory-efficient Pseudo-IDList to avoid materialization. Experiments show that WSUI outperforms the state-of-the-art EWSPM in terms of both runtime and memory usage.
Recent network-constrained clustering approaches achieve improved clustering performance over earlier methods such as NS-TBC; however, they still incur considerable computational overhead due to unprioritized expansion and full adjacency-matrix traversal, leading to redundant edge processing and high time complexity. To address these limitations, this paper proposes Fast-NSTBC (Fast Network-Space Topological-Based Clustering), an optimized topology-based clustering algorithm that prioritizes nearest-point expansion and utilizes an adjacency-list representation. These enhancements reduce unnecessary edge expansion and restrict traversal to only adjacent points, rather than all points in the dataset. Experimental results on eight real-world datasets show that Fast-NSTBC significantly outperforms the state-of-the-art NS-TBC, reducing computational cost by 86
Mental health concerns are increasingly reflected in online personal narratives, where individuals express emotional struggles in subtle and culturally grounded ways. This paper presents a study on binary emotional distress classification in Vietnamese narrative texts. We introduce VNMH1550, a manually annotated dataset of 1,550 user-submitted confessions collected from the “Tâm sự” section of VnExpress, one of the most widely read online newspapers in Vietnam. Annotation guidelines were developed in collaboration with a clinical psychologist, and trained annotators achieved strong agreement (AAR ≥ 0.8), underscoring the task’s complexity. We benchmark traditional machine learning, deep learning, transformer-based pretrained models, and a zero-shot large language model for distress detection. Among classical models, Random Forest achieved the best macro-F1 (69.66
Erasable itemset mining can support users in analyzing the production planning in factories. During a financial crisis, manufacturers may halt production lines for low-profit products to minimize the overall impact on profit. In this paper, we propose erasable itemset mining with itemset-range constraints. The conditional requirement for assigning subsets and supersets is used to prevent the generation of unrelated patterns and better satisfy users’ needs. To address this issue, the proposed method employs bit vectors and data shrinking for execution performance. The bit vectors encode product material usages, thereby accelerating the calculation of itemsets’ gain values. The subset and superset constraints are used to eliminate unnecessary candidates during mining. Additionally, three data-shrinking strategies are applied to reduce the dataset size. Experimental results indicate that the proposed approach outperforms the previous work in terms of execution time and memory consumption.
Classifying Polish medical texts in conditions of low data availability places high demands on effectiveness, reliability, and calibration. This article examines the interaction between data augmentation (BT) and self-distillation (SD) for the mDeBERTa-v3-base model, analysing their impact on effectiveness, calibration and two dimensions of stability: resistance to linguistic noise and OOD degradation. Our experiments reveal a key trade-off: we show that BT alone improves calibration but is unstable on out-of-distribution data. In contrast, SD worsens raw calibration but provides exceptional robustness to linguistic perturbations. The combination of both methods (BT+SD) resolves this conflict, achieving the best balance: highest effectiveness, best probabilistic quality and double stability. The model inherits the stability of OOD while maintaining the noise resistance inherited from SD. We conclude that BT+SD is the most balanced strategy, combining the semantic benefits of BT with the stability of SD, making it a promising method for reliable NLP systems in clinical applications.
With the exponential growth of textual data on the internet, sentiment analysis has emerged as a crucial technique for enhancing the performance of recommendation systems by extracting user opinions from textual data such as reviews, comments, and social media posts. Sentiment analysis is the process that aims to identify and extract the sentiments and opinions expressed in a text. It is a research field whose main objective is to extract information about the polarity of a text, that is, to determine whether the author expresses positive, negative, or neutral sentiments towards the subject in question. While traditional recommendation systems usually rely on explicit ratings and user-item inter-actions, sentiment analysis can provide additional deeper insights into user preferences by assessing the polarity and emotional tone of text. Nowadays, sentiment analysis is gaining increasing attention in many domains, including politics, education, marketing, and economics, among others. This is due to the fact that opinions have a considerable influence as they significantly contribute to decision-making processes. This has led to the emergence of various sentiment analysis methods, including lexicon-based methods, machine learning-based methods, hybrid methods, and deep learning-based methods. Despite the-se advancements, the field still faces significant challenges, such as developing robust models capable of handling cultural and contextual variations, as well as the need for larger and more diverse datasets. This article presents an in-depth study on sentiment analysis approaches in different languages and its role in recommendation systems, with a particular focus on sentiment analysis in Arabic.
Emotional text-to-speech (TTS) has advanced rapidly in recent years, yet building such systems for languages without emotional data remains a fundamental challenge. The scarcity of annotated emotional corpora makes it difficult to model expressive prosody and transfer emotional nuances, particularly in low-resource languages. To address this problem, we propose a cross-lingual emotional TTS framework that transfers emotional expressiveness from a source language with rich emotional resources to a target language with scarce emotional data. Our approach builds on a VITS-based architecture and introduces three key innovations. First, we condition the model with emotion embeddings and zero-shot speaker embeddings, enabling expressive and speaker-specific synthesis across languages. Second, we adopt a dual-tokenizer design that maintains distinct phoneme vocabularies for the source and target languages, thereby reducing accent leakage and improving pronunciation robustness in cross-lingual adaptation. Finally, to enhance emotion controllability, we incorporate an auxiliary emotion classification objective. Experimental results show that our method synthesizes emotionally expressive speech in the target language despite the absence of target-language emotional data, with naturalness and emotional clarity comparable to systems trained on fully emotional corpora. This work demonstrates a practical and effective pathway for developing emotional TTS in low-resource settings.
Emotion detection is an essential tool for gaining insights into user or customer feedback on products, whether it is through social media comments, product reviews, or news discussions. However, accurately identifying emotions in text is a challenging task due to the complexity and variability of human language. To address this challenge, machine learning and data analysis techniques are often employed. This paper presents an enhanced voting rule-based framework that systematically combines predictions from six transformer models through adaptive fusion strategies. By combining the outputs of multiple BERT models through hierarchical decision rules, this approach leverages the complementary strengths of diverse transformer architectures to enhance overall classification accuracy. The voting mechanism systematically reduces prediction errors and model-specific biases that individual models might introduce, ensuring that the final prediction is more reliable by reflecting the consensus of multiple models rather than relying on a single model. Experimental evaluation on the SemEval 2025 Task 11 dataset demonstrates our proposal has achieved Macro F1 of 0.7462 and Micro F1 of 0.7770, consistent improvements over individual transformer models (Macro F1 is up 1.22
Image recontextualization involves transforming an image from its original setting to a new context while preserving the details of the subject, including broad applications in media, marketing, and e-commerce. Recent advances in deep generative modeling, particularly diffusion-based approaches, have significantly improved recontextualization techniques. However, existing methods such as IP-Adapter and Magic Clothing struggle with balancing contextual integration and precise control over image generation, while HyperLoRA suffers from information loss due to CLIP-based encoding. On the other hand, fine-tuning methods like Dreambooth or textual inversion require time-consuming adaptation for a single subject and struggle to preserve the details of the item. In this work, we introduce Parallel U-Net for Image Recontextualization (PUIR), a novel approach that leverages dual U-Net architectures, namely, one U-Net is dedicated to extracting features from the conditioning image and another U-Net allows generating the target image in a new context. By employing self-attention layers to fuse parallel features from both U-Nets, PUIR enhances detail preservation and improves image fidelity. Moreover, we develop a method for enabling multi-subject recontextualization while requiring only single-item training pairs. During the denoising, we propose to use a novel noise-mixing strategy together with a masking approach to generate the target image composed of many subjects. We demonstrate the effectiveness of PUIR in fashion applications. Our experiments show that PUIR outperforms HyperLoRA in maintaining item-specific details and overall image quality.
Human pose estimation is one of the most extensively studied problems in computer vision. However, it is noted that a crucial evaluation benchmark is still very difficult and challenging due to large variation in poses, number of persons interacting, and frequent occlusions. Recently, attention modules like SE, CBAM, and ELA have shown great potential for feature extraction improvements particularly for HPE models and more broadly for computer vision tasks. This paper studies the effects that integrating these attention modules at primary levels of feature extraction in the OmniPose model brings about. Three variants of OmniPose with SEBlock, CBAM, and ELA included are described, which were trained under exactly similar conditions to make a fair comparison among them. SE yields the best overall accuracy, ELA balances accuracy and training efficiency at larger batch sizes, while CBAM provides improved stability but sacrifices final accuracy. Such results underscore the fact that the effectiveness of attention mechanisms largely varies with where they are positioned within the architecture of a model, rather than how they are designed internally. This work initiates a shifted mindset for the research community towards laying a foundation for more appropriately designed attention mechanisms and thus opening up their usage to disparate models for HPE at various integration levels.
Visual grounding on synthetic aperture radar (SAR) imagery aims to localize objects described by natural language expressions. However, existing SAR grounding studies remain limited to single-category scenarios and lack relational descriptions, preventing models from generalizing to realistic multi-object scenes. To address this issue, we introduce SAAVG, a new multi-class SAR visual grounding dataset built from multiple high-resolution SAR detection sources and enriched with automatically generated and manually verified referring expressions. Building upon the LQVG framework, we propose an Asymmetric Hierarchical Fusion mechanism that deepens the Vision–Language Interaction pathway through iterative cross-modal refinement, motivated by the dominant role of this pathway in visual grounding performance. We examine two variants, Shared-Weight and Per-Scale, to characterize their behavior across SAR and optical domains. Experiments on DIOR-RSVG, SARVG1.0, and SAAVG show that the Shared-Weight variant achieves state-of-the-art results on SARVG1.0 (Pr@0.5 92.01 https://github.com/duongb1/SAAVG .
Facial Emotion Recognition (FER) is crucial for applications in human-computer interaction and mental health. However, existing FER datasets often suffer from limitations such as class imbalance and limited diversity, hindering the development of robust models that generalize across different data distributions. This paper introduces FaceEmo-Set, a novel dataset designed to address these challenges through strategic data integration and balanced representation. Comprising ∼ 29k collected images, of which 25,200 remain after quality validation, across seven basic emotions (anger, disgust, fear, happiness, neutral, sadness, and surprise), FaceEmo-Set integrates diverse sources including movies, TV shows, GIFs, internet images, AI-generated content, and samples from established datasets. We evaluate FaceEmo-Set using a Vision Transformer (ViT) architecture on external benchmarks, comparing its cross-dataset generalization performance against established datasets (FER2013, RAF-DB, and RAVDESS). Our results demonstrate FaceEmo-Set’s balanced performance across all emotion categories, particularly excelling in challenging minority classes like disgust (0.75 and 0.70 recall on AffectNet and FER2013, respectively). When combined with existing datasets, FaceEmo-Set achieves significantly improved accuracy (65.27
In the rapidly evolving field of intelligent surveillance, real-time human detection is essential for enhancing security and operational efficiency. This study optimizes human detection and tracking by integrating YOLOv10 with ByteTrack, leveraging its high accuracy, low computational cost, and real-time processing capabilities. Among YOLO variants with fewer than 26 million parameters, YOLOv10-L achieves a mAP50 of 91
Encoder-decoder models for video captioning trained with token-level cross-entropy often exhibit weak semantic grounding and sensitivity to spurious correlations, particularly when audio provides misleading cues. We introduce EBM-EoD-VC, a unified framework that integrates sample-level energy-based modeling with distribution-level energy statistics into a standard Transformer pipeline. The model learns a joint embedding space regularized by energy distance, distance correlation, and partial distance correlation to control for audio as a confounder. On the compact MSVD dataset, a well-tuned cross-entropy baseline remains competitive. On the larger, noisier MSR-VTT dataset, our method achieves consistent gains in CIDEr and SPICE, especially when audio is used. These results demonstrate that energy-based regularization tightens video-caption alignment and improves semantic adequacy, with benefits most pronounced in open-domain, multimodal settings.
This paper introduces a method for evaluating the semantic fidelity of transformations between conceptual metamodels. Using a semi-formal system, we define FLAT method for quantification of enforcment, loss, alteration, and preservation (translation) of meaning through formal metrics. The method is validated by evaluating mappings from an Association-Oriented Metamodel (AOM) to UML, EER, and ORM, demonstrating its effectiveness in identifying semantic discrepancies.
Retrieval-Augmented Generation (RAG) systems for biomedicine frequently produce plausible but unverifiable mechanistic claims, undermining their utility for safety-critical applications such as adverse event prediction, drug repurposing, and clinical decision support. The core pathology is generation-before-verification: systems that permit language models to propose mechanistic links without requiring explicit, structured evidence paths. We address this by inverting the pipeline: structural traceability precedes and constrains generation. Our deterministic Graph-RAG method validates drug–disease associations if and only if an explicit mechanistic path exists: Drug (D) to Target (T) to Phenotype (P), where edges are constructed from curated interaction databases (ChEMBL, OpenTargets) and validated via MedCPT-encoded retrieval over PubMed. We introduce a hybrid weighted scoring mechanism where the LLM acts as a “negation detector” to penalize factually incorrect high-similarity vectors below an acceptance threshold. We evaluate against the Comparative Toxicogenomics Database (CTD) [1] on 50 rare drugs spanning easy, medium, and hard difficulty strata, comparing our deterministic pipeline against 4 frontier generative baselines: GPT-4.1, GPT-5.2 (December 2025 release), Claude Sonnet 4.5, and Opus 4.6, all equipped with web search tools (Tavily API). Across 400 total experiments (8 arms × 50 drugs), our pipeline achieves significantly higher Precision@1 (0.760–0.841 vs. 0.540–0.620 for websearch baselines, Bonferroni-corrected p < 0.05 ) and evidence quality (96 https://github.com/rangan2510/Drug-Evidence .
Computer programming knowledge is fundamental in information technology education. Supporting students in querying programming knowledge is therefore important for teaching and digital transformation. To develop such systems, programming knowledge must be modeled and organized into a knowledge-base, and an expressive query language is required to capture diverse learner information needs. Existing studies on knowledge representation and query languages remain incomplete, particularly for representing programming language knowledge. This paper proposes a combined knowledge representation model for computer programming together with a query language tailored to the proposed model. We also investigate query parsing and forward chaining inference mechanisms to query execution. The approach is implemented in a programming knowledge querying system and evaluated with real information technology students. Experimental results demonstrate the effectiveness and practical applicability of the proposed system.
