https://dspace.ctu.edu.vn/jspui/handle/123456789/2593 Viện Hàn lâm Khoa học và Công nghệ Việt Nam Sun, 15 Mar 2026 03:07:08 GMT 2026-03-15T03:07:08Z https://dspace.ctu.edu.vn/jspui/handle/123456789/119587 Title: On the relational dependency coalitional games Authors: Vu, Duc Nghia; Demetrovics, Janos; Tran, Thanh Dai; Vu, Duc Thi Abstract: Cooperating game theory is becoming increasingly popular in AI, data science, and game theory applications in sharing and circular economy. Social media shows us the impact of many influencers on millions (even hundreds of millions) of followers, which raised the need to have a new model of coalition game, in which the influence or dependence of players on others are not equal, some have more than others. In this paper, we introduce a relational dependency game with new properties of minimal winning coalitions and maximal losing coalitions and their in-depth relationship with different approaches from simple games. In this new model, unlike simple games, all winning coalitions have the same payoff but losing coalitions have different payoffs, which coincides with Leo Tolstoy’s philosophy: all happy families are alike, but each unhappy family is unhappy in its way. The algorithm to find a minimal winning coalition among maximal losing coalitions is addressed in this paper. In this new model, unlike a simple game, we present the relational dependency coalition game model in which players depend on or do not depend on one another when they share a common interest in achieving a specific goal or outcome. The players must find a minimal winning coalition on which all players of the game depend on achieving the highest payoff. Closure operations and choice functions arise naturally in this game when there is a one-to-one correspondence between the winning coalition/losing coalition and the closure operation/choice function. And the game becomes more complex when relational independence lives with dependency among players. How to have a structural representation of relational independence along with dependency and how to describe a minimal winning collation on a simple hypergraph is also addressed in the paper. Mon, 01 Jan 2024 00:00:00 GMT https://dspace.ctu.edu.vn/jspui/handle/123456789/119587 2024-01-01T00:00:00Z https://dspace.ctu.edu.vn/jspui/handle/123456789/119574 Title: Language-adversarial training for indic multilingual speaker verification Authors: Hoang, Long Vu; Nguyen, Van Huy; Ngo, Thi Thu Huyen; Pham, Viet Thanh Abstract: Speaker verification now reports a reasonable level of accuracy in its applications in voice-based biometric systems. Recent research on deep neural networks and predicting speaker identity based on speaker embeddings has gained remarkable success. However, results are limited when it comes to verifying multilingual speakers. In this paper, we propose an ensemble system submitted to the I-MSV Challenge 2022. The system is built upon the ECAPA-TDNN and RawNet2 models with additional adversarial training layers. Probabilistic Linear Discriminant Analysis (PLDA) back-end scoring and Large Margin Cosine Loss (LMCL) are implemented to further obtain more discriminative features. Experimental results show that on the Constraint Private Test set of the task, our proposed model achieved remarkable results, ranking third with an Equal Error Rate (EER) of 2.9734%. Mon, 01 Jan 2024 00:00:00 GMT https://dspace.ctu.edu.vn/jspui/handle/123456789/119574 2024-01-01T00:00:00Z https://dspace.ctu.edu.vn/jspui/handle/123456789/119573 Title: Spatio-temporal graph learning with epidemiological factors for HIV epidemic short-term prediction Authors: Pham, Thanh Dat; Nguyen, Van Duong; Tran, Tan Thanh; Nguyen, Viet Anh Abstract: HIV/AIDS is a major epidemic in the 21st century, with high mortality rates and no effective preventive vaccine. It significantly impacts the economy, mental well-being and health systems and shortens national lifespans. Early detection helps reduce transmission and allocate medical resources effectively. However, predicting outbreaks remains challenging due to the influence of temporal, spatial and epidemiological factors, which complicate the spread of the disease across regions and pose difficulties for predictive models. Very few studies use deep learning models to tackle the HIV epidemic. To address this gap, we suggest using a graph data structure to simulate HIV transmission between neighboring areas and integrate epidemiological factors into this framework. We develop a spatio-temporal graph neural network model to predict short-term infection trends. This model incorporates important factors from HIV modeling, including temporal dynamics, geographic regions, and epidemiological variables such as age groups, career groups, gender groups, risk population groups, and transmission routes within an area. Our approach uses self-attention in the graph architecture to gather node-level information across the infection graph at each step during time series processing. We employ a GRU mechanism to update the graph information over time, allowing for a comprehensive evaluation of transmission probabilities between regions and improving predictive accuracy. Our proposed model was tested on HIV datasets from districts in Ho Chi Minh City, Viet Nam, and demonstrated superior performance compared to existing spatio-temporal models applied to the same dataset. Mon, 01 Jan 2024 00:00:00 GMT https://dspace.ctu.edu.vn/jspui/handle/123456789/119573 2024-01-01T00:00:00Z https://dspace.ctu.edu.vn/jspui/handle/123456789/119572 Title: A data challenge for Vietnamese abstractive multi-document summarization Authors: Tran, Mai Vu; Le, Hoang Quynh; Can, Duy Cat; Nguyen, Quoc An Abstract: This paper provides an overview of the Vietnamese abstractive multi-document summarization shared task (AbMuSu) for Vietnamese news, which is hosted at the 9th annual workshop on Vietnamese Language and Speech Processing (VLSP 2022). The main goal of this shared task is to develop automated summarization systems that can generate abstractive summaries for a given set of documents on a specific topic. The input consists of several news documents on the same topic, and the output is a related abstractive summary. The focus of the AbMuSu shared task is solely on Vietnamese news summarization. To this end, a human-annotated dataset comprising 1,839 documents in 600 clusters, collected from Vietnamese news in 8 categories, has been developed. Participating models are evaluated and ranked based on their ROUGE2-F1 score, which is the most common evaluation metric for document summarization problems. Mon, 01 Jan 2024 00:00:00 GMT https://dspace.ctu.edu.vn/jspui/handle/123456789/119572 2024-01-01T00:00:00Z