检索结果-内蒙古大学图书馆

MindLLM: Pre-training Lightweight Large Language Model from Scratch, Evaluations and Domain Applications

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Yang, Yizhe Sun, Huashan Li, Jiawei Liu, Runheng Li, Yinghao Liu, Yuhang Gao, Yang Huang, Heyan School of Computer Science Beijing Institute of Technology Beijing Engineering Research Center of High Volume Language Information Processing and Cloud Computing Applications Beijing Institute of Technology Southeast Academy of Information Technology China

Large Language Models (LLMs) have demonstrated remarkable performance across various natural language tasks, marking significant strides towards general artificial intelligence. While general artificial intelligence is leveraged by developing increasingly large-scale models, there could be another branch to develop lightweight custom models that better serve certain domains, taking into account the high cost of training and deploying LLMs and the scarcity of resources. In this paper, we present MindLLM, a novel series of bilingual lightweight large language models, trained from scratch, alleviating such burdens by offering models with 1.3 billion and 3 billion parameters. A thorough account of experiences accrued during large model development is given, covering every step of the process, including data construction, model architecture, evaluation, and applications. Such insights are hopefully valuable for fellow academics and developers. MindLLM consistently matches or surpasses the performance of other open-source larger models on some public benchmarks. We also introduce an innovative instruction tuning framework tailored for smaller models to enhance their capabilities efficiently. Moreover, we explore the application of MindLLM in specific vertical domains such as law and finance, underscoring the agility and adaptability of our lightweight models. Copyright © 2023, The Authors. All rights reserved.

关键词： Benchmarking

A Systematic Review of Machine Learning Approaches for Detecting Deceptive Activities on Social Media: Methods, Challenges, and Biases

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Liu, Yunchong Shen, Xiaorui Zhang, Yeyubei Wang, Zhongyan Tian, Yexin Dai, Jianglai Cao, Yuchen School of Engineering and Applied Science University of Pennsylvania United States Khoury college of computer science Northeastern University United States Center for Data Science New York University United States Georgia Institute of Technology College of Computing United States Department of EECS University of California Berkeley United States

Social media platforms like Twitter, Facebook, and Instagram have facilitated the spread of misinformation, necessitating automated detection systems. This systematic review evaluates 36 studies that apply machine learning (ML) and deep learning (DL) models to detect fake news, spam, and fake accounts on social media. Using the Prediction model Risk Of Bias ASsessment Tool (PROBAST), the review identified key biases across the ML lifecycle: selection bias due to non-representative sampling, inadequate handling of class imbalance, insufficient linguistic preprocessing (e.g., negations), and inconsistent hyperparameter tuning. Although models such as Support Vector Machines (SVM), Random Forests, and Long Short-Term Memory (LSTM) networks showed strong potential, over-reliance on accuracy as an evaluation metric in imbalanced data settings was a common flaw. The review highlights the need for improved data preprocessing (e.g., resampling techniques), consistent hyperparameter tuning, and the use of appropriate metrics like precision, recall, F1 score, and AUROC. Addressing these limitations can lead to more reliable and generalizable ML/DL models for detecting deceptive content, ultimately contributing to the reduction of misinformation on social media. © 2024, CC BY.

关键词： Adversarial machine learning

Quantum memory assisted entangled state verification with local measurements

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Chen, Siyuan Xie, Wei Xu, Ping Wang, Kun Hefei National Research Center for Physical Sciences Microscale School of Physical Sciences University of Science and Technology of China Hefei230026 China College of Physics Jilin University Changchun130012 China School of Computer Science and Technology University of Science and Technology of China Hefei230027 China Institute for Quantum Information State Key Laboratory of High Performance Computing College of Computer Science and Technology National University of Defense Technology Changsha410073 China

We consider the quantum memory assisted quantum state verification task, where an adversary prepare independent multipartite entangled states and send to the local verifiers, who then store several copies in the quantum memory and measure them collectively to make decision. We establish an exact analytic formula for optimizing two-copy state verification, where the verifiers store two copies, and give a globally optimal two-copy strategy for multi-qubit graph states involving only Bell measurements. When the verifiers can store arbitrarily many copies, we present a dimension expansion technique that designs efficient verification strategies for this case, showcasing its application to efficiently verifying GHZ-like states. These strategies become increasingly advantageous with growing memory resources, ultimately approaching the theoretical limit of efficiency. Our findings demonstrate that quantum memories enhance state verification efficiency, sheding light on error-resistant strategies and practical applications of large-scale quantum memory-assisted verification. © 2023, CC BY.

关键词： Quantum entanglement

Prompt-based Logical Semantics Enhancement for Implicit Discourse Relation Recognition

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Chenxu Jian, Ping Huang, Mu School of Computer Science and Technology Beijing Institute of Technology Beijing China Beijing Engineering Research Center of High Volume Language Information Processing and Cloud Computing Applications Beijing Institute of Technology Beijing China

Implicit Discourse Relation Recognition (IDRR), which infers discourse relations without the help of explicit connectives, is still a crucial and challenging task for discourse parsing. Recent works tend to exploit the hierarchical structure information from the annotated senses, which demonstrate enhanced discourse relation representations can be obtained by integrating sense hierarchy. Nevertheless, the performance and robustness for IDRR are significantly constrained by the availability of annotated data. Fortunately, there is a wealth of unannotated utterances with explicit connectives, that can be utilized to acquire enriched discourse relation features. In light of such motivation, we propose a Prompt-based Logical Semantics Enhancement (PLSE) method for IDRR. Essentially, our method seamlessly injects knowledge relevant to discourse relation into pre-trained language models through prompt-based connective prediction. Furthermore, considering the prompt-based connective prediction exhibits local dependencies due to the deficiency of masked language model (MLM) in capturing global semantics, we design a novel self-supervised learning objective based on mutual information maximization to derive enhanced representations of logical semantics for IDRR. Experimental results on PDTB 2.0 and CoNLL16 datasets demonstrate that our method achieves outstanding and consistent performance against the current state-of-the-art models. Copyright © 2023, The Authors. All rights reserved.

关键词： Semantics

Twenty Years of Personality computing: Threats, Challenges and Future Directions

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Celli, Fabio Kartelj, Aleksandar Ðorðević, Miljan Suhartono, Derwin Filipović, Vladimir Milutinović, Veljko Spathoulas, Georgios Vinciarelli, Alessandro Kosinski, Michal Lepri, Bruno Maggioli SpA Italy Faculty of Mathematics University of Belgrade Serbia School of Electrical Engineering University of Belgrade Serbia School of Computer Science Bina Nusantara University Indonesia Dept. of Information Security Norwegian University of Science and Technology Norway School of Computing Science University of Glasgow United Kingdom Graduate School of Business Stanford University United States Center for Augmented Intelligence Fondazione Bruno Kessler Italy

Personality computing is a field at the intersection of Personality Psychology and computer science. Started in 2005, research in the field utilizes computational methods to understand and predict human personality traits. The expansion of the field has been very rapid and, by analyzing digital footprints (text, images, social media, etc.), it helped to develop systems that recognize and even replicate human personality. While offering promising applications in talent recruiting, marketing and healthcare, the ethical implications of Personality computing are significant. Concerns include data privacy, algorithmic bias, and the potential for manipulation by personality-aware Artificial Intelligence. This paper provides an overview of the field, explores key methodologies, discusses the challenges and threats, and outlines potential future directions for responsible development and deployment of Personality computing technologies. © 2025, CC BY-NC-ND.

关键词：

A Reference Vector-Assisted Many-Objective Optimization Algorithm with Adaptive Niche Dominance Relation

学校读者我要写书评

暂无评论

Intelligent Automation & Soft computing 2024年第2期39卷 189-211页

作者： Fangzhen Ge Yating Wu Debao Chen Longfeng Shen School of Computer Science and Technology Huaibei Normal UniversityHuaibei340604China School of Physic and Electronic Information Huaibei Normal UniversityHuaibei340604China Anhui Engineering Research Center for Intelligent Computing and Application on Cognitive Behavior(ICACB) Huaibei Normal UniversityHuaibei340604China Anhui Province Key Laboratory of Intelligent Computing and Applications Huaibei Normal UniversityHuaibei340604China Institute of Artificial Intelligence Hefei Comprehensive National Science CenterHefei230000China

It is still a huge challenge for traditional Pareto-dominatedmany-objective optimization algorithms to solve manyobjective optimization problems because these algorithms hardly maintain the balance between convergence and diversity and can only find a group of solutions focused on a small area on the Pareto front,resulting in poor performance of those *** this reason,we propose a reference vector-assisted algorithmwith an adaptive niche dominance relation,for short *** new dominance relation forms a niche based on the angle between candidate *** comparing these solutions,the solutionwith the best convergence is found to be the non-dominated solution to improve the selection *** reproduction,a mutation strategy of k-bit crossover and hybrid mutation is used to generate high-quality *** 23 test problems with up to 15-objective,we compared the proposed algorithm with five state-of-the-art *** experimental results verified that the proposed algorithm is competitive.

关键词： Many-objective optimization evolutionary algorithm Pareto dominance reference vector adaptive niche

RLG-RAG: Guiding the Knowledge Retrieval and Evaluation in Retrieval-Augmented Generation Framework by Reasoning Logic 25

学校读者我要写书评

暂无评论

RLG-RAG: Guiding the Knowledge Retrieval and Evaluation in R...

Companion Proceedings of the ACM on Web Conference 2025

作者： Kehan Xu Kun Zhang Wei Huang Jingyuan Li Yuanzhuo Wang School of Information Science and Engineering Yanshan University Qinhuangdao China Pattern Recognition Center Tencent WeChat AI Beijing China School of Computer and Artificial Intelligence Beijing Technology and Business University Beijing China Institute of Computing Technology Chinese Academy of Sciences (CAS) Beijing China

ISBN: (纸本)9798400713316

The knowledge retrieval, integration, and evaluation processes in the RAG method lack the guidance of reasoning logic, leading to ongoing challenges in maintaining factual consistency. To address these issues, this paper proposes the RLG-RAG framework, which constructs a reasoning graph based on user queries to guide the knowledge retrieval, integration, and evaluation processes. By fully representing the reasoning logic of RAG, RLG-RAG dynamically models and integrates knowledge relationships during retrieval and defines a precise scope of relevant knowledge through sufficiency evaluation. This reduces inference-irrelevant knowledge that large language models may obtain. Experimental analyses on accuracy, factual consistency, and robustness demonstrate that RLG-RAG resists interference and provides accurate, factually consistent answers. The project URL is https://***/10.5281/zenodo.14852250.

关键词： knowledge evaluation

China in the eyes of news media:a case study under COVID-19 epidemic

学校读者我要写书评

暂无评论

Frontiers of Information technology & Electronic Engineering 2021年第11期22卷 1443-1457页

作者： Hong HUANG Zhexue CHEN Xuanhua SHI Chenxu WANG Zepeng HE Hai JIN Mingxin ZHANG Zongya LI National Engineering Research Center for Big Data Technology and System Huazhong University of Science and TechnologyWuhan430074China Services Computing Technology and System Lab Huazhong University of Science and TechnologyWuhan430074China Cluster and Grid Computing Lab Huazhong University of Science and TechnologyWuhan430074China School of Computer Science and Technology Huazhong University of Science and TechnologyWuhan430074China School of Journalism and Information Communication Huazhong University of Science and TechnologyWuhan430074China

As one of the early COVID-19 epidemic outbreak areas,China attracted the global news media’s attention at the beginning of *** the epidemic period,Chinese people united and actively fought against the ***,in the eyes of the international public,the situation reported about China is not *** better understand how the international public portrays China,especially during the epidemic,we present a case study with big data *** aim to answer three questions:(1)What has the international media focused on during the COVID-19 epidemic period?(2)What is the media’s tone when they report China?(3)What is the media’s attitude when talking about China?In detail,we crawled more than 280000 pieces of news from 57 mainstream media agencies in 22 countries and made some interesting *** example,international media paid more attention to Chinese livelihood during the COVID-19 epidemic *** March and April,“progress of Chinese vaccines,”“specific drugs and treatments,”and“virus outbreak in U.S.”became the media’s most common *** terms of news attitude,Cuba,Malaysia,and Venezuela had a positive attitude toward China,while France,Canada,and the United Kingdom had a negative *** study can help understand China’s image in the eyes of the international media and provide a sound basis for image analysis.

关键词： Country image COVID-19 epidemic Topic mining Entity Tone of news Emotion

Research on the design of indoor air quality monitoring system based on cloud-side-end collaboration

学校读者我要写书评

暂无评论

Research on the design of indoor air quality monitoring syst...

International Conference on computer Engineering and Applications (ICCEA)

作者： Liang Yu Yuting Lin Haonan Xu Jiabao Yang Guotong Zhao College of Computer Guangdong University of Science and Technology DongGuan China AIoT Edge Computing Engineering Technology Research Center of Dongguan City Guangdong AIoT Application Innovation Joint Laboratory School of Information Technology & Engineering Guangzhou College of Commerce Guangzhou China

ISBN: (数字)9798350386776

ISBN: (纸本)9798350386783

Given that individuals spend more than 90% of their time indoors, research indicates a direct correlation between indoor air quality and overall health and well-being. With the rise of respiratory illnesses and increasing concerns about indoor air quality, traditional methods such as using potted plants, air purifiers, and basic air conditioning units with purification capabilities are insufficient for real-time monitoring. To address these challenges, this study proposes an intelligent indoor air quality monitoring system that leverages cloud-edge-end collaboration. The system includes a cost-effective air quality monitor on the device side to collect and localize indoor air quality data. This data is then transmitted to the edge-side ceiling light via WiFi-Mesh. A machine learning model deployed using the CNN-LSTM algorithm on the Edge Impluse platform is used to analyze and predict indoor air quality levels over time. The system uploads this information to a cloud MQTT server using the Cat.1 protocol and sends notifications to subscribed users via a mobile app, enabling individuals to monitor and adjust indoor air quality as needed. This proactive approach can help reduce air pollution, decrease disease incidence, and enhance overall quality of life. Physical and deployment tests demonstrate the system's high performance, with a training accuracy of 97.5%, F1-SCORE of 95%, and prediction accuracy of 94.1%, making it a reliable tool for indoor air quality management.

关键词： Protocols Accuracy Atmospheric modeling Collaboration Predictive models Real-time systems Servers