After the health and mental care activities, as well as skill development programs, UITers will be equipped with academic knowledge through the specialized seminar "ARTIFICIAL INTELLIGENCE IN BIOINFORMATICS" held at 8:30 am on March 28, 2024, at Hall E (12th floor, E building, University of Information Technology) with the guest speaker being Professor TS. Do Phuc. Students, postgraduate students, and doctoral students can register to attend here: Registration Form For more details, please visit: Event Details Đông Xanh - Collaborative Communication Staff, University of Information Technology

Thesis Title: Development of pattern mining algorithms on quantitative databases [Ph.D. Thesis in Computer Science] Author: Bui Danh Huong Abstract: The dissertation presents an overview and theoretical basis study of common weighted pattern mining on quantitative databases. Based on this, the thesis proposes new efficient methods to solve problems and overcome current challenges in weighted pattern mining such as excessively large result pattern sets, redundant rule issues, or user-directed and real-time pattern mining challenges. Four specific problems focused on in the thesis are weighted frequent pattern mining, weighted closed frequent pattern mining, top-rank-k weighted frequent pattern mining, and stream data-based weighted frequent pattern mining. The algorithms proposed in the dissertation demonstrate superior efficiency over existing algorithms in terms of runtime, memory usage, and scalability when mining various forms of weighted patterns in the aforementioned problems. The research results of the dissertation have been published in 4 SCIE journals including Expert Systems with Applications (Q1, 2018), Applied Intelligence (Q2, 2020), Knowledge-Based Systems (Q1, 2020), IEEE Access (Q1, 2021), and presented in 3 domestic and international conferences (SMC-2016, FAIR-2016, and @-2017). *NOVEL CONTRIBUTIONS OF THE THESIS The scientific contributions of the dissertation include: Proposal of WN-tree structure and WN-list data structure, leading to the NFWI algorithm for efficient weighted frequent pattern mining. The WN-list structure has several advantages such as linear complexity of intersection operation, self-pruning capability, and fast computation of pattern weight support based on the WN-list of the pattern. Proposal oed on the ancestor relationship of WN-list, leading to the NFWCI algorithm for efficient weighted closed frequent pattern mining based on the WN-list structure. Proposal of TFWIN+ algorithm for efficient top-rank-k weighted frequent pattern mining based on the WN-list structure and early threshold increasing and branch pruning strategies. Proposal of SWN-tree structure, an improvement from the WN-tree structure, for effective storage and maintenance of information on data windows when sliding on data streams. This leads to the FWPODS algorithm for efficient stream data-based weighted frequent pattern mining based on the sliding window model. APPLICATIONS/ POTENTIAL APPLICATIONS IN PRACTICE OR OPEN ISSUES NEEDING FURTHER RESEARCH Future research directions will focus on solving several pattern mining problems on weighted databases such as mining maximal weighted itemsets, mining weighted closed frequent patterns on growing databases, mining weighted frequent patterns on uncertain databases, and deploying solutions for weighted frequent pattern mining on multi-core and distributed systems. We will also investigate applications utilizing the weighted pattern mining platform such as graph mining, social network mining, text mining, and IoT data mining. Interested readers are invited to visit the Library to read the hard copy or access the full text remotely at the following address: https://ir.vnulib.edu.vn/handle/VNUHCM/15989 For any inquiries regarding access accounts, please contact via email: thuvien@uit.edu.vn For more details, please refer to: https://www.facebook.com/LibUIT.Fanpage/posts/pfbid02CFrU9pEpP3iVrnYyjXXm3mA6Pm3LgBJSM93qMwUF8tpntkdbeoc9t55iVjyP6xRul Written by: Ha Bang Translated by: Ngoc Diem

1. Solving Ordinary Differential Equations in Python Joakim Sundnes in Simula SpringerBriefs on Computing (2024) => Link ebook: https://link.springer.com/book/10.1007/978-3-031-46768-4 2. Algorithms for Sparse Linear Systems Jennifer Scott, Miroslav Tůma in Nečas Center Series (2023) => Link ebook: https://link.springer.com/book/10.1007/978-3-031-25820-6 3. Space–Time Conservation Element and Solution Element Method Advances and Applications in Engineering Sciences Chih-Yung Wen, Yazhong Jiang… in Engineering Applications of Computational Methods (2023) => Link ebook: https://link.springer.com/book/10.1007/978-981-99-0876-9 4. Computational Physiology Simula Summer School 2022 − Student Reports Kimberly J. McCabe in Simula SpringerBriefs on Computing (2023) => Link ebook: https://link.springer.com/book/10.1007/978-3-031-25374-4 5. Computational Physiology Simula Summer School 2021 − Student Reports Kimberly J. McCabe in Simula SpringerBriefs on Computing (2022) => Link ebook: https://link.springer.com/book/10.1007/978-3-031-05164-7 6. Anisotropy Across Fields and Scales Prof. Evren Özarslan… in Mathematics and Visualization (2021) => Link ebook: https://link.springer.com/book/10.1007/978-3-030-56215-1 7. Introduction to Scientific Programming with Python Dr. Joakim Sundnes in Simula SpringerBriefs on Computing (2020) => Link ebook: https://link.springer.com/book/10.1007/978-3-030-50356-7 8. Programming for Computations - Python A Gentle Introduction to Numerical Simulations with Python 3.6 Prof. Dr. Svein Linge… in Texts in Computational Science and Engineering (2020) => Link ebook: https://link.springer.com/book/10.1007/978-3-030-16877-3 For more details, please visit: Library Facebook post Hạ Băng - Collaborative communication staff, University of Information Technology

Article: "Object Detection in Aerial Images Using the Dino Method" Students involved: Le Huu Do - KTPM2022.1 - Lead author Ngo Huong Giang - KTPM2022.1 - Co-author Supervisors: MSc. Vo Duy Nguyen MSc. Tran Thi Hong Yen MSc. Nguyen Thanh Hiep Summary of the paper: Recently, data collection from unmanned aerial vehicles has significantly contributed to the development of tasks related to processing aerial images. Three basic tasks in this field include classification, object detection, and segmentation. However, particularly in the task of object detection, researchers often focus on improving traditional models, i.e., models belonging to either one-stage or two-stage categories. However, we have also witnessed the emergence and remarkable development of end-to-end object detection models. Among these methods, DINO (DETR with Improved deNoising anchor boxes) stands out as the first end-to-end method to achieve SOTA (State of the Art) results on the leaderboard. We conducted experiments on the VisDrone dataset with various backbones to comprehensively evaluate the performance of the DINO method and obtained many significant results. This is a promising competitive result for other end-to-end methods. "We would like to sincerely thank the School of Software Engineering, Multimedia Communications Department, and the UIT-Together research group for providing us with the opportunity to research and complete this paper." The 26th National Conference on Electronics, Communications, and Information Technology (REV-ECIT) 2023. Date of organization: December 16, 2023. For more information, please visit: Link Đông Xanh - Collaborative Communication Officer, University of Information Technology

The University Training Department (P.ĐTĐH) announces the mid-term exam (GK) organization plan for the second semester of the academic year 2023-2024 as follows:    - Centralized exam time: From April 15, 2024, to April 20, 2024 (during the centralized exam period, students are exempt from classes).    - The time for courses managed by the course management unit to organize separate exams: from April 22, 2024, to April 27, 2024. (During this time, students will attend classes as usual). For detailed information, please visit: https://se.uit.edu.vn/tin-tuc/10-thong-bao-hoc-vu/11620-kế-hoạch-tổ-chức-thi-giữa-kỳ-học-kỳ-2-năm-học-2023-2024-2023-2024.html Ha Bang - Collaborator of the University of Information Technology's Communication Department

Paper: "VlogQA: Task, Dataset, and Baseline Models for Vietnamese Spoken-Based Machine Reading Comprehension" Paper Link: https://arxiv.org/abs/2402.02655 Students involved: Ngo Phuoc Thinh - KHDL 2019 Dang Tran Anh Khoa - KHDL 2019 Supervisors: Mr. Luu Thanh Son Mr. Nguyen Van Kiet Assoc. Prof. Nguyen Luu Thuy Ngan Abstract: This paper presents the development process of a Vietnamese spoken language corpus for machine reading comprehension (MRC) tasks and provides insights into the challenges and opportunities associated with using real-world data for machine reading comprehension tasks. The existing MRC corpora in Vietnamese mainly focus on formal written documents such as Wikipedia articles, online newspapers, or textbooks. In contrast, the VlogQA consists of 10,076 question-answer pairs based on 1,230 transcript documents sourced from YouTube – an extensive source of user-uploaded content, covering the topics of food and travel. By capturing the spoken language of native Vietnamese speakers in natural settings, an obscure corner overlooked in Vietnamese research, the corpus provides a valuable resource for future research in reading comprehension tasks for the Vietnamese language. Regarding performance evaluation, our deep-learning models achieved the highest F1 score of 75.34% on the test set, indicating significant progress in machine reading comprehension for Vietnamese spoken language data. In terms of EM, the highest score we accomplished is 53.97%, which reflects the challenge in processing spoken-based content and highlights the need for further improvement. The team would like to extend sincere thanks to Assoc. Prof. Nguyen Luu Thuy Ngan, Mr. Luu Thanh Son, and Mr. Nguyen Van Kiet - those who have guided and supported us throughout the project. Thanks to the inspiration, ideas, and care of our supervisor, we have successfully completed the project. Detailed Information: https://www.facebook.com/UIT.Fanpage/posts/pfbid02gNuPjFNGYLvocEvtpqAD1r... Hạ Băng - Media Collaborator, University of Information Technology English version: Phan Huy Hoang  

Paper: "XLMR4MD: New Vietnamese Dataset and Framework for Detecting the Consistency of Description and Permission in Android Applications Using Large Language Models" Students involved: Nguyen Ngoc Qui - IT2018 Supervisors: Dr. Nguyen Tan Cam Mr. Nguyen Van Kiet Abstract: Google Play and other app platforms host a variety of Android applications along with their metadata. Among this metadata, description and privacy policies help explain the functionality of the applications. They also describe the permissions of the application, especially those related to sensitive information. Detecting inconsistencies between the description of the application, privacy information, and permissions extracted from the application's source code helps users decide whether to install and use the application. In this study, we propose a new method based on a pre-trained language model to detect inconsistencies between permissions extracted from the application description, privacy policies, and permissions extracted from the application's source code (APK file). Related work focuses on models of large-scale datasets, especially for languages with rich resources like English. However, a language with low resources, specifically Vietnamese, requires more datasets for this task. To address this issue, we propose the ViDPApp dataset (Description and Privacy Policies of Applications in Vietnamese domains) - a manually labeled dataset with over 12,000 sentences with inter-annotator agreement (IAA) of over 85%. Additionally, we propose XLMR4MD, a framework using large language models, outperforming other machine learning models (LSTM, Bi-GRU-LSTM-CNN, WikiBERT, DistilBERT, mBERT, and PhoBERT). The framework achieves the best F1 score of 84.04% in detecting inconsistencies between Android application permissions and descriptions. This framework can be fine-tuned for 100 different languages, thereby extending to languages with low resources like Vietnamese. This dataset is available for research purposes. "I would like to extend my sincerest thanks to Dr. Nguyen Tan Cam and Mr. Nguyen Van Kiet. With their willingness and dedication, both of you have guided us throughout the research process, pointing out significant limitations in the research process. The support of both of you has been a great source of encouragement, helping me overcome challenges and accomplish the research work to the best of my ability. I sincerely appreciate the guidance and valuable knowledge that both of you have imparted to me. Once again, I thank both of you very much!" Detailed Information: https://www.facebook.com/UIT.Fanpage/posts/pfbid033awLFtwtmYu8h7nzLMoiWU... Hạ Băng - Media Collaborator, University of Information Technology English version: Phan Huy Hoang  

Paper: "ViLexNorm: A Lexical Normalization Corpus for Vietnamese Social Media Text" Paper Link: https://arxiv.org/abs/2401.16403 Students involved: Nguyen Thanh Nhi - KHDL2021  - Co-first author Le Thanh Phong - KHDL2021  - Co-first author Supervisor: Mr. Nguyen Van Kiet Abstract: Lexical normalization, a fundamental task in Natural Language Processing (NLP), involves the transformation of words into their canonical forms. This process has been proven to benefit various downstream NLP tasks greatly. In this work, we introduce ViLexNorm, the first-ever corpus developed for the Vietnamese lexical normalization task. The corpus comprises over 10,000 pairs of sentences meticulously annotated by human annotators, sourced from public comments on Vietnam's most popular social media platforms. Various methods were used to evaluate our corpus, and the best-performing system achieved a result of 57.74% using the Error Reduction Rate (ERR) metric (van der Goot, 2019a) with the Leave-As-Is (LAI) baseline. For extrinsic evaluation, employing the model trained on ViLexNorm demonstrates the positive impact of the Vietnamese lexical normalization task on other NLP tasks. Our corpus is publicly available exclusively for research purposes. "We would like to express our sincere gratitude to Mr. ThS. Nguyen Van Kiet for dedicating much time and enthusiasm to guide the group throughout the research and paper publication process. We also thank Mr. ThS. Luu Thanh Son for supporting the group during the experimentation phase. Additionally, we extend our thanks to the lecturers and assistants of the Faculty of Information Science and Engineering for their valuable insights during the execution of this project." The 18th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2024), ranked A, is the flagship European conference dedicated to European and international researchers, covering a wide spectrum of research in Computational Linguistics and Natural Language Processing. Detailed Information: https://www.facebook.com/UIT.Fanpage/posts/pfbid02qcuaXwc8yjKigg7AZ6QBSq... Hạ Băng - Media Collaborator, University of Information Technology English version: Phan Huy Hoang  

Title of the paper: "Detection of ships in aerial images using the RTMDET-ROTATED method" Paper link: https://drive.google.com/file/d/1tq60Dj2A9eYTDZgJchOvbMcs8xGJMMPW/view?usp=sharing&fbclid=IwAR1oGFU26L95yI6Ak1Tm4gWcXfW2_TtkK8fDLK_eAwXbMKdhf4bGN0WNcQ4 Student: Tran Nguyen Chi Huy - KTPM2021.KTPM20211 - Main author Supervisors: M.Sc Vo Duy Nguyen M.Sc Dinh Nguyen Anh Dung Paper Summary: Currently, thanks to the development of unmanned aerial vehicles, detecting objects in images captured from above has become a prominent topic in the field of computer vision. However, traditional object detection methods still cannot indicate the orientation of objects and the bounding boxes overlap if the image has a dense object density. Therefore, the problem of rotated object detection has emerged. Among them, detecting ships is one of the issues that attract much attention due to its high application in reality, especially when the sovereignty of national waters and maritime safety face many challenges and changes. To overcome these difficulties and meet the demand, many studies have focused on developing effective methods. In this study, we experimentally evaluate the real-time object detection model, called RTMDet (Real-Time Models for Object Detection), with two backbones CSPNeXt and CSPDarkNet on the DOTA-ship dataset. This dataset was created to meet the specific needs of ship detection-related applications. DOTA-ship also reproduces most of the challenges in real-life situations. Finally, we analyze the results and provide detailed comments, analyzing RTMDet when trained on the DOTA-ship dataset, thereby suggesting directions for future research to improve methods in the future. "I sincerely express my gratitude to the Faculty of Software Engineering, Multimedia Communication Department, and the UIT-Together research group for creating conditions that have helped me research and complete this paper." The 26th National Conference on Electronics, Communications, and Information Technology, REV-ECIT 2023, was held on December 16, 2023.\ For more details, visit: https://www.facebook.com/UIT.Fanpage/posts/pfbid08JWb82MfCT... Ha Bang - Media Collaborator, University of Information Technology Nhat Hien - Translation Collaborator, University of Information Technology

Title: MULTIVARIATE SPATIO-TEMPORAL DATA VISUALIZATION MODEL [Doctoral Dissertation in Computer Science] Author: Nguyen Thi Hong Abstract: Data visualization is a mapping of data into information and/or knowledge. In this process, data is represented as a visual graph suitable for human visual perception in transforming the graph into information or knowledge. The dissertation "Multivariate Spatio-Temporal Data Visualization Model" has systematized the process of transforming data into corresponding knowledge with human visual perception and the process of representing multivariate data of research cases to structure into a framework for visualizing multivariate spatio-temporal data. However, visualizing multivariate data faces significant challenges due to technical limitations of display environments and biological constraints of the human visual system. The dissertation has analyzed and decomposed this challenge into three research questions to build a visualization framework: (1) How to represent multivariate data on a 2D display environment?; (2) How to display multidimensional graphs with desired visual properties?; (3) How to overcome the limitations of size and resolution of 2D display environments? Surveying the relationships of data variables in applied visualization cases, the dissertation proposes a multidimensional block model for representing multivariate data. The non-spatial multidimensional block represents non-spatial multivariate data. The 2D spatial block, also known as the multidimensional map, represents data of moving objects on the ground and spatio-temporal data. The 3D spatial block represents aerial data. These multidimensional blocks visualize multivariate data to support users in answering analytical questions including elementary questions, global questions, and relational questions. When analyzing data according to relational questions of survey cases, users have extracted useful information from non-spatial multidimensional block representations of experimental datasets related to the hand-foot-mouth disease epidemic in Binh Duong province and from multidimensional map representations of experimental datasets related to the dengue fever epidemic in An Giang, Soc Trang, and Tien Giang provinces. Based on real-world experimental applications, the dissertation approaches from both sides of mapping data into information and/or knowledge to establish a multivariate data visualization system consisting of two components: technical visualization and visual perception. The technical visualization component transforms data into graphs by computers, while the visual perception component transforms graphs into information by humans. For the technical visualization component, the dissertation applies the philosophy of "the whole is greater than the sum of its parts" to structure a visualization framework transforming multivariate dataset into a graph suitable for users' visual perception and requirements. The visualization framework consists of two layers: visualization mapping and visual display. The visualization mapping layer is influenced by the characteristics of the data in transforming the multivariate dataset into a multidimensional graph. The visual display layer is influenced by the characteristics of visual perception and user requirements in enhancing the visual properties of the multidimensional graph. The visualization framework representing multidimensional data proposed by the dissertation "Multivariate Spatio-Temporal Data Visualization Model" is an interactive visualization framework. Graphic tools provided to users not only interact with visual graphs but also interact with all visualization functions to improve the visual properties of the graph according to users' preferences. The visualization framework has addressed three research questions related to challenges in visualizing multivariate spatio-temporal data. The dissertation has established a process of integrating visual perception variables as a strong solution to enhance the visual properties of multidimensional graphs. The integrated approach to visual perception applies Descartes' product of visual perception variables with layers of flat feature signs to enhance visual properties, where the emerging visual perception variables are subsets of a basic visual perception variable or subsets of the product of basic visual perception variables. The increase in the number of emerging visual perception variables resulting from the product of basic visual perception variables leads to an increase in the significant number of integrated visual perception variable flat feature layers. For those interested, please visit the Library to read the hard copy or access the full text remotely at the following address:https://ir.vnulib.edu.vn/handle/VNUHCM/15987 For more details, visit: https://www.facebook.com/LibUIT.Fanpage/posts/pfbid0W8uqZPuLQthjBekXkx... Ha Bang - Media Collaborator, University of Information Technology Nhat Hien - Translation Collaborator, University of Information Technology

Title: ANALYSIS OF SOCIAL NETWORKS BASED ON TOPIC MODELING AND APPLICATIONS [Doctoral Thesis in Computer Science] Author: Hồ Trung Thành Summary: Today, social networks play a crucial role in various fields such as economics, society, politics, education, and many others. Social networks enable individuals or communities to share messages, discuss, or provide feedback on topics of interest through social links. Among these, messages serve as implicit social links containing valuable information and knowledge. Each message has multiple characteristics, among which the topic of interest and the individuals involved in sending or receiving the topic are crucial. Unlike traditional approaches that view each message belonging to a single topic, the topic modeling approach reveals that each message contains a mixture of multiple topics, with each topic being of interest to multiple individuals or communities. The topic modeling approach is suitable for analyzing social networks. However, previous studies still have limitations in exploring, labeling, and analyzing the temporal variations of topics of interest of individuals and communities based on social links with temporal factors. The thesis aims to address the limitations of previous studies. Specifically, the thesis focuses on two main objectives: (i) Building a Time-Author-Recipient-Topic (TART) model based on the topic modeling approach. The tasks of the TART model include exploring topics of interest and analyzing the role of individuals in topics within messages exchanged on social networks; labeling topics; using temporal factors to divide factors such as the set of individuals sending and receiving topics, the set of topics, and identifying the temporal variations of topics of interest of individuals over time; analyzing changes in individuals' topics of interest; (ii) Developing a method for exploring communities on social networks based on the topic modeling approach with temporal factors and the Kohonen neural network method. The tasks of the community exploration method involve clustering individuals based on features such as topics and levels of interest to identify communities of individuals with common interests; analyzing the characteristic variations of communities on social networks. Through experiments on proposed models and methods using two Vietnamese message datasets (collected from social networks in universities and online newspapers) with a software system built for social network analysis, the thesis has achieved its objectives. The main contributions of the thesis are: (1) Developing a method combining exploration and labeling of topics of interest on social networks based on the topic modeling approach and hierarchical topic trees. (2) Building the TART model to explore the role of individuals based on the topic modeling approach with temporal factors. This model also plays an important role in exploring social links between individuals on social networks. (3) Developing a method for exploring individual communities based on the topic modeling approach. The community exploration method is a combination of the TART model and the Kohonen neural network method to discover communities of individuals with common interests at each time period. (4) Building social network analysis software to fully implement the six modules on the overall research framework of the thesis from data collection, preprocessing, topic exploration and labeling experiments, TART model experiments, and community exploration experiments. The research results and experiments of the thesis have been published in specialized journals, conferences, and indexed by reputable scientific databases such as Thomson Reuters, Scopus, IEEE, Springer. Interested readers please visit the Library to read the paper copy or access the full text remotely at the following address: https://ir.vnulib.edu.vn/handle/VNUHCM/15977 If you have any questions or need support about your access account, please contact via email: thuvien@uit.edu.vn Detailed Information: https://www.facebook.com/LibUIT.Fanpage/posts/pfbid02i7Na87gpQY13qawAEAA... Hạ Băng - Media Collaborator, University of Information Technology  English version: Phan Huy Hoang

Today, on March 1, 2024, the University of Information Technology will hold a session of the Doctoral Dissertation Evaluation Council for doctoral candidate Nguyen Van Kiet in the Computer Science field. With the topic "Research on Building an Automatic Reading Comprehension Model for Vietnamese Text," supervised by Associate Professor Dr. Nguyen Luu Thuy Ngan and Dr. Nguyen Gia Tuan Anh. It is known that the topic "Research on Building an Automatic Reading Comprehension Model for Vietnamese Text" has presented papers at international conferences in the first quarter with related research. Doctoral candidate Nguyen Van Kiet has excellently completed the research topic and received high praise from the Council. According to the evaluation of the members of the Council, the dissertation reflects the serious studying and research process of the doctoral candidate. The research results have high scientific value and practical application, meeting all the requirements of a doctoral dissertation. Detailed Information: https://www.facebook.com/share/p/7YAnCMbKFD8zozhU/?mibextid=WC7FNe Hạ Băng - Media Collaborator, University of Information Technology English version: Phan Huy Hoang