https://elib.vku.udn.vn/handle/123456789/2662 2026-04-16T09:56:18Z 2026-04-16T09:56:18Z Vu, Manh Hung Nguyen, Van Quang Tran, Thi Thao Pham, Van Truong https://elib.vku.udn.vn/handle/123456789/2752 2023-09-26T02:36:03Z 2023-07-01T00:00:00Z

Title: A New ConvMixer-Based Approach for Diagnosis of Fault Bearing Using Signal Spectrum Authors: Vu, Manh Hung; Nguyen, Van Quang; Tran, Thi Thao; Pham, Van Truong Abstract: It has been reported that nearly 40% of electrical machine failures are caused by bearing problems. That is why identifying bearing failure is crucial. Deep learning for diagnosing bearing faults has been widely used, like WDCNN, Conv-mixer, and Siamese models. However, good diagnosis takes a significant quantity of training data. In order to overcome this, we propose a new approach that can dramatically improve training performance with a small data set. In particular, we propose to integrate the ConvMixer models to the backbone of Siamese network, and use the few-short learning for more accurate classification even with limited training data. Various experimental results with raw signal inputs and signal spectrum inputs are conducted, and compared with those from traditional models using the same data set provided by Case Western Reserve University (CWRU). Description: Lecture Notes in Networks and Systems (LNNS, volume 734); CITA: Conference on Information Technology and its Applications; pp: 3-14.

2023-07-01T00:00:00Z Le, Trieu Phong Tran, Thi Phuong https://elib.vku.udn.vn/handle/123456789/2751 2023-09-26T02:33:56Z 2023-07-01T00:00:00Z

Title: Differentially-Private Distributed Machine Learning with Partial Worker Attendance: A Flexible and Efficient Approach Authors: Le, Trieu Phong; Tran, Thi Phuong Abstract: In distributed machine learning, multiple machines or workers collaborate to train a model. However, prior research in cross-silo distributed learning with differential privacy has the drawback of requiring all workers to participate in each training iteration, hindering flexibility and efficiency. To overcome these limitations, we introduce a new algorithm that allows partial worker attendance in the training process, reducing communication costs by over 50% while preserving accuracy on benchmark data. The privacy of the workers is also improved because less data are exchanged between workers. Description: Lecture Notes in Networks and Systems (LNNS, volume 734); CITA: Conference on Information Technology and its Applications; pp: 15-24.

2023-07-01T00:00:00Z Ngo, Q. Hung Nguyen, D. Hien Le, Khac Nhien An https://elib.vku.udn.vn/handle/123456789/2750 2023-09-26T02:32:02Z 2023-07-01T00:00:00Z

Title: Building Legal Knowledge Map Repository with NLP Toolkits Authors: Ngo, Q. Hung; Nguyen, D. Hien; Le, Khac Nhien An Abstract: Today, the legal document system is increasingly strict with different levels of influence and affects activities in many different fields. The increasing number of legal documents interwoven with each other also leads to difficulties in searching and applying in practice. The construction of knowledge maps that involve one or a group of legal documents is an effective approach to represent actual knowledge domains. A legal knowledge graph constructed from laws and legal documents can enable a number of applications, such as question answering, document similarity, and search. In this paper, we describe the process of building a system of knowledge maps for the Vietnamese legal system from the source of about 325,000 legal documents that span all fields of social life. This study also proposes an integrated ontology to represent the legal knowledge from legal documents. This model integrates the ontology of relational knowledge and the graph of key phrases and entities in the form of a concept graph. It can express the semantics of the content of a given legal document. In addition, this study also describes the process of building and exploiting natural language processing tools to build a VLegalKMaps system, which is a repository of Vietnamese legal knowledge maps. We also highlight open challenges in the realization of knowledge graphs in a technical legal system that enables this approach at scale. Description: Lecture Notes in Networks and Systems (LNNS, volume 734); CITA: Conference on Information Technology and its Applications; pp: 25-36.

2023-07-01T00:00:00Z Le, Tran Duc Le, Ba Luong Dinh, Truong Duy Pham, Van Dai https://elib.vku.udn.vn/handle/123456789/2749 2023-09-26T02:30:11Z 2023-07-01T00:00:00Z

Title: Classification of Ransomware Families Based on Hashing Techniques Authors: Le, Tran Duc; Le, Ba Luong; Dinh, Truong Duy; Pham, Van Dai Abstract: The primary objective of this research is to propose a novel method for analyzing malware through the utilization of hashing techniques. The proposed approach integrates the use of Import Hash, Fuzzy Hash, and Section Level Fuzzy Hash (SLFH) to create a highly optimized, efficient, and accurate technique to classify ransomware families. To test the proposed methodology, we collected a comprehensive dataset from reputable sources and manually labelled each sample to augment the reliability and precision of our analysis. During the development of the proposed methodology, we introduced new steps and conditions to identify ransomware families, resulting in the highest performance level. The major contributions of this research include the combination of various hashing techniques and the proposal of a hash comparison strategy that facilitates the comparison of section hashes between ransomware and the pre-build database. Description: Lecture Notes in Networks and Systems (LNNS, volume 734); CITA: Conference on Information Technology and its Applications; pp: 37-49.

2023-07-01T00:00:00Z