1. A brief review on benchmarking for large language models evaluation in healthcareLeona Cilar Budler, Hongyu Chen, Aokun Chen, Maxim Topaz, Wilson Tam, Jiang Bian, Gregor Štiglic, 2025, pregledni znanstveni članek Opis: This paper reviews benchmarking methods for evaluating large language models (LLMs) in healthcare settings. It highlights the importance of rigorous benchmarking to ensure LLMs' safety, accuracy, and effectiveness in clinical applications. The review also discusses the challenges of developing standardized benchmarks and metrics tailored to healthcare-specific tasks such as medical text generation, disease diagnosis, and patient management. Ethical considerations, including privacy, data security, and bias, are also addressed, underscoring the need for multidisciplinary collaboration to establish robust benchmarking frameworks that facilitate LLMs' reliable and ethical use in healthcare. Evaluation of LLMs remains challenging due to the lack of standardized healthcare-specific benchmarks and comprehensive datasets. Key concerns include patient safety, data privacy, model bias, and better explainability, all of which impact the overall trustworthiness of LLMs in clinical settings. Ključne besede: artificial intelligence, benchmarking, chatbots, healthcare, large language models, natural language processing Objavljeno v DKUM: 12.05.2025; Ogledov: 0; Prenosov: 1
Povezava na datoteko |
2. A hierarchical knowledge graph embedding framework for link predictionShuang Liu, Chengwang Hou, Jiana Meng, Peng Chen, Simon Kolmanič, 2024, izvirni znanstveni članek Ključne besede: knowledge graph embedding, knowledge graph completion, negative sampling, link prediction Objavljeno v DKUM: 04.02.2025; Ogledov: 0; Prenosov: 3
Celotno besedilo (1,76 MB) |
3. A cyclic di-GMP network is present in gram-positive Streptococcus and gram-negative Proteus speciesYing Liu, Changhan Lee, Fengyang Li, Janja Trček, Heike Bähre, Rey-Ting Guo, Chun-Chi Chen, Alexey Chernobrovkin, Roman Zubarev, Ute Römling, 2020, izvirni znanstveni članek Opis: The ubiquitous cyclic di-GMP (c-di-GMP) network is highly redundant with numerous GGDEF domain proteins as diguanylate cyclases and EAL domain proteins as c-di-GMP specific phosphodiesterases comprising those domains as two of the most abundant bacterial domain superfamilies. One hallmark of the c-di-GMP network is its exalted plasticity as c-di-GMP turnover proteins can rapidly vanish from species within a genus and possess an above average transmissibility. To address the evolutionary forces of c-di-GMP turnover protein maintenance, conservation, and diversity, we investigated a Gram-positive and a Gram-negative species, which preserved only one single clearly identifiable GGDEF domain protein. Species of the family Morganellaceae of the order Enterobacterales exceptionally show disappearance of the c-di-GMP signaling network, but Proteus spp. still retained one diguanylate cyclase. As another example, in species of the bovis, pyogenes, and salivarius subgroups as well as Streptococcus suis and Streptococcus henryi of the genus Streptococcus, one candidate diguanylate cyclase was frequently identified. We demonstrate that both proteins encompass PAS (Per-ARNT-Sim)-GGDEF domains, possess diguanylate cyclase catalytic activity, and are suggested to signal via a PilZ receptor domain at the C-terminus of type 2 glycosyltransferase constituting BcsA cellulose synthases and a cellulose synthase-like protein CelA, respectively. Preservation of the ancient link between production of cellulose(-like) exopolysaccharides and c-di-GMP signaling indicates that this functionality is even of high ecological importance upon maintenance of the last remnants of a c-di-GMP signaling network in some of today’s free-living bacteria. Ključne besede: cellulose biosynthesis, cyclic di-GMP signaling, EAL domain, GGDEF domain, Proteus mirabilis, Streptococcus gallolyticus subsp. gallolyticus Objavljeno v DKUM: 28.01.2025; Ogledov: 0; Prenosov: 13
Celotno besedilo (12,81 MB) Gradivo ima več datotek! Več... |
4. Improved relation extraction through key phrase identification using community detection on dependency treesShuang Liu, Xunqin Chen, Jiana Meng, Niko Lukač, 2025, izvirni znanstveni članek Opis: A method for extracting relations from sentences by utilizing their dependency trees to identify key phrases is presented in this paper. Dependency trees are commonly used in natural language processing to represent the grammatical structure of a sentence, and this approach builds upon this representation to extract meaningful relations between phrases. Identifying key phrases is crucial in relation extraction as they often indicate the entities and actions involved in a relation. The method uses community detection algorithms on the dependency tree to identify groups of related words that form key phrases, such as subject-verb-object structures. The experiments on the Semeval-2010 task8 dataset and the TACRED dataset demonstrate that the proposed method outperforms existing baseline methods. Ključne besede: community detection algorithms, dependency trees, relation extraction Objavljeno v DKUM: 17.01.2025; Ogledov: 0; Prenosov: 5
Celotno besedilo (3,12 MB) |
5. Review of pallet pooling studiesChunhua Chen, Jianwei Ren, A Yunga, 2019, pregledni znanstveni članek Opis: Development of sharing economy has brought new opportunities to the development of green logistics. Pallet pooling is a typical type of sharing economy. It has been recently promoted in almost every country in the world. Researchers have reached a common agreement that establishing an effective pallet pool is of great importance to the pallets use efficiency. It is also believed that the implementation of pallet pool will substantially contribute to the economic and social development. Detailed analysis of pallet pooling such as benefit analysis, mode choose, pallet allocation model & algorithm, quality control, etc. are discussed in this paper. Based on literature review, several important problems are listed for future study. These problems are information management, supply chain management, rental pricing, pallets tracking and allocation, quality control, sustainable development (carbon emissions), and “internet plus pallet pooling”. Some suggestions on China’s pallet pooling development are proposed as well. Ključne besede: logistics, pallet pool, rental, internet plus, China Objavljeno v DKUM: 22.08.2024; Ogledov: 51; Prenosov: 8
Celotno besedilo (452,49 KB) Gradivo ima več datotek! Več... |
6. Segregation dynamics driven by network leadersWen-Xuan Wang, Yuhao Feng, Siru Chen, Wenzhe Xu, Xinjian Zhuo, Huijia Li, Matjaž Perc, 2022, izvirni znanstveni članek Opis: Network segregation - a critical problem in real-life networks - can reveal the emergence of conflicts or signal an impending collapse of the whole system. However, the strong heterogeneity of such networks and the various definitions for key nodes continue to pose challenges that limit our ability to foresee segregation and to determine the main drivers behind it. In this paper, we show that a multi-agent leader-follower consensus system can be utilized to define a new index, named leadership, to identify key leaders in real-life networks. And then, this paper explores the emergence of network segregation that is driven by these leaders based on the removal or the rewiring of the relations between different nodes in agreement with their contribution distance. We finally show that the observed leaders-driven segregation dynamics reveals the dynamics of heterogeneous attributes that critically influence network structure and its segregation. Thus, this paper provides a theoretical method to study complex social interactions and their roles in network segregation, which ultimately leads to a closed-form explanation for the emergence of imbalanced network structure from an evolutionary perspective. Ključne besede: complex networks, network segregation, multi-agent leader–follower consensus system, key leaders identification, leader, segregation, social physics Objavljeno v DKUM: 08.07.2024; Ogledov: 122; Prenosov: 24
Celotno besedilo (3,18 MB) Gradivo ima več datotek! Več... |
7. CNN-Based Vessel Meeting Knowledge Discovery From AIS Vessel TrajectoriesPeng Chen, Shuang Liu, Niko Lukač, 2023, izvirni znanstveni članek Opis: How to extract a collection of trajectories for different vessels from the raw AIS data to discover vessel meeting knowledge is a heavily studied focus. Here, the AIS database is created based on the raw AIS data after parsing, noise reduction and dynamic Ramer-Douglas-Peucker compression. Potential encountering trajectory pairs will be recorded based on the candidate meeting vessel searching algorithm. To ensure consistent features extracted from the trajectories in the same time period, time alignment is also adopted. With statistical analysis of vessel trajectories, sailing segment labels will be added to the input feature. All motion features and sailing segment labels are combined as input to one trajectory similarity matching method based on convolutional neural network to recognize crossing, overtaking or head-on situations for each potential encountering vessel pair, which may lead to collision if false actions are adopted. Experiments on AIS data show that our method is effective in classifying vessel encounter situations to provide decision support for collision avoidance. Ključne besede: AIS Data, CNN, Dynamic Rammer-Douglas-Peucker, knowledge discovery, maneuvering pattern, traffic pattern, trajectory Objavljeno v DKUM: 19.03.2024; Ogledov: 538; Prenosov: 417
Celotno besedilo (3,84 MB) Gradivo ima več datotek! Več... |
8. Research on vehicle re-identification algorithm based on fusion attention methodPeng Chen, Shuang Liu, Simon Kolmanič, 2023, izvirni znanstveni članek Opis: The specific task of vehicle re-identification is how to quickly and correctly match the same vehicle in different scenarios. In order to solve the problem of inter-class similarity and environmental interference in vehicle images in complex scenes, one fusion attention method is put forward based on the idea of obtaining the distinguishing features of details-the mechanism for the vehicle re-identification method. First, the vehicle image is preprocessed to restore the image's attributes better. Then, the processed image is sent to ResNet50 to extract the features of the second and third layers, respectively. Then, the feature fusion is carried out through the two-layer attention mechanism for a network model. This model can better focus on local detail features, and global features are constructed and named SDLAU-Reid. In the training process, a data augmentation strategy of random erasure is adopted to improve the robustness. The experimental results show that the mAP and rank-k indicators of the model on VeRi-776 and the VehicleID are better than the results of the existing vehicle re-identification algorithms, which verifies the algorithm's effectiveness. Ključne besede: vehicle re-identification, attention mechanism, key-point, local feature, feature fusion Objavljeno v DKUM: 06.02.2024; Ogledov: 348; Prenosov: 30
Celotno besedilo (5,28 MB) Gradivo ima več datotek! Več... |
9. Measurement of two-particle correlations of hadrons in ▫$e^+ e^-$▫ collisions at BelleY.-C. Chen, Marko Bračko, Samo Korpar, Peter Križan, Tomaž Podobnik, Marko Starič, Luka Šantelj, 2022, izvirni znanstveni članek Opis: The measurement of two-particle angular correlation functions in high-multiplicity e+e− collisions at √s=10.52 GeV is reported. In this study, the 89.5 fb−1 of hadronic e+e− annihilation data collected by the Belle detector at KEKB are used. Two-particle angular correlation functions are measured in the full relative azimuthal angle (Δϕ) and three units of pseudorapidity (Δη), defined by either the electron beam axis or the event-shape thrust axis, and are studied as a function of charged-particle multiplicity. The measurement in the thrust axis analysis, with mostly outgoing quark pairs determining the reference axis, is sensitive to the region of additional soft gluon emissions. No significant anisotropic collective behavior is observed with either coordinate analyses. Near-side jet correlations appear to be absent in the thrust axis analysis. The measurements are compared to predictions from various event generators and are expected to provide new constraints to the phenomenological models in the low-energy regime. Ključne besede: Particle & resonance production, Quark & gluon jets, Quark-gluon plasma Objavljeno v DKUM: 24.08.2023; Ogledov: 396; Prenosov: 9
Celotno besedilo (634,71 KB) Gradivo ima več datotek! Več... |
10. Corrosion mechanisms for cemented soils in three different sulfate solutionsPengju Han, Chao Ren, Xiaohong Bai, Frank Chen, 2015, izvirni znanstveni članek Opis: In order to simulate and study the corrosion effects on the compressive strength of cemented soils that could be exposed in a polluted environment, a series of tests were conducted on cemented soil blocks cured with different concentrations of H2SO4, MgSO4, and Na2SO4 solutions. The test results show that the corrosion degree generally increases with the corrosion time and the solution concentration, while the compressive strength decreases with the increasing corrosion degree. The corrosion degree is highest for the Na2SO4 solution, followed by the MgSO4 and H2SO4 solutions. Namely, when the SO4 2- ion exists in a solution, the corrosion degree for the positive ions follows this descending order: Na+, Mg2+, and H+. X-ray diffraction (XRD) phase analyses were performed for the cemented soil samples after corrosion and ionic concentrations. The results show that the compressive strength decreases with an increase of the Mg2+ concentration in the MgSO4 solution and the Na+ concentration in the Na2SO4 solution. At the same time, the strength increases with an increase of the pH value of the H2SO4 solution. Based on the chemical analysis results, the corrosion of H2SO4 or MgSO4 solutions on cemented soils is deemed as a composite action involving the combined resolving and crystallizing corrosion processes. Furthermore, the corrosion of the Na2SO4 solution of cemented soil is a composite action consisting of dissolving and crystallizing. Ključne besede: cemented soil, compressive strength, corrosion mechanism, sulfate, pollution, solution Objavljeno v DKUM: 15.06.2018; Ogledov: 1757; Prenosov: 178
Celotno besedilo (305,84 KB) Gradivo ima več datotek! Več... |