A5023705881- Natural Language Processing Techniques
- Topic Modeling
- Auction Theory and Applications
- Consumer Market Behavior and Pricing
- Scheduling and Optimization Algorithms
- Supply Chain and Inventory Management
- Semantic Web and Ontologies
- Machine Learning and Algorithms
- Industrial Technology and Control Systems
- Business Process Modeling and Analysis
- Resource-Constrained Project Scheduling
- Evaluation and Optimization Models
- Biomedical Text Mining and Ontologies
- Scheduling and Timetabling Solutions
- Prosthetics and Rehabilitation Robotics
- Advanced Computational Techniques and Applications
- Simulation and Modeling Applications
- Muscle activation and electromyography studies
- Intelligent Tutoring Systems and Adaptive Learning
- Advanced Graph Neural Networks
- Bayesian Modeling and Causal Inference
- Optimization and Search Problems
- Software Reliability and Analysis Research
- Educational Technology and Assessment
- Text Readability and Simplification
Fiberhome Technology Group (China)
2024
Harbin Institute of Technology
2019-2024
North China Electric Power University
2021-2024
Northwestern Polytechnical University
2012-2021
Xi'an Aeronautical University
2021
Xidian University
2004-2020
University of Washington
2019
Jilin Medical University
2014
Changchun University
2014
University of Glasgow
2013
Despite considerable advancements with deep neural language models, the enigma of text degeneration persists when these models are tested as generators. The counter-intuitive empirical observation is that even though use likelihood training objective leads to high quality for a broad range understanding tasks, using decoding bland and strangely repetitive. In this paper, we reveal surprising distributional differences between human machine text. addition, find strategies alone can...
Large language models (LLMs) have shown great potential across various industries due to their remarkable ability generalize through instruction tuning. However, the limited availability of domain-specific data significantly hampers performance on specialized tasks. While existing methods primarily focus selecting training from general datasets that are similar target domain, they often fail consider joint distribution instructions, resulting in inefficient learning and suboptimal knowledge...
Understanding causality has vital importance for various Natural Language Processing (NLP) applications. Beyond the labeled instances, conceptual explanations of can provide deep understanding causal fact to facilitate reasoning process. However, such explanation information still remains absent in existing resources. In this paper, we fill gap by presenting a human-annotated explainable CAusal REasoning dataset (e-CARE), which contains over 20K questions, together with natural language...
Abstract Cloud computing has gained popularity in recent years, but with its rise comes concerns about data security. Unauthorized access and attacks on cloud-based data, applications, infrastructure are major challenges that must be addressed. While machine learning algorithms have improved intrusion detection systems cloud security, they often fail to consider the entire life cycle of file processing, making it difficult detect certain issues, especially insider attacks. To address these...
Li Du, Xiao Ding, Kai Xiong, Ting Liu, Bing Qin. Proceedings of the 59th Annual Meeting Association for Computational Linguistics and 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.
Li Du, Xiao Ding, Ting Liu, Bing Qin. Proceedings of the 59th Annual Meeting Association for Computational Linguistics and 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.
Component assignment problem is a common challenge of reliability optimization, which non-deterministic polynomial hard widely used in the linear consecutive k-out-of-n systems. In consideration advantages quantum computing and importance measure, this article proposed novel algorithm, Birnbaum importance–based genetic to improve efficiency accuracy for solving component problem. First, model optimization systems established. Second, detailed procedure algorithm introduced solve Moreover,...
Abstract An outcome of upward social comparisons that has been largely overlooked is its effect on non‐transactional behaviours (i.e., word mouth). Previous research identified three different emotional reactions to comparisons: admiration, benign envy and malicious envy. Despite the fact their consumption previously analysed, it remains unclear how these affect mouth intention. This study carries out an experimental design demonstrates admiration positively influence behaviour. However,...
Li Du, Xiao Ding, Ting Liu, Zhongyang Li. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint (EMNLP-IJCNLP). 2019.
We focus on the task of stock market prediction based financial text which contains information that could influence movement market. Previous works mainly utilize a single semantic unit text, such as words, events, sentences, to predict tendency However, interaction different-grained within can be useful for context knowledge supplement and predictive selection, then improve performance prediction. To facilitate this, we propose constructing heterogeneous graph with nodes from task. A novel...
Acquiring high-quality temporal common sense (TCS) knowledge from free-form text is a crucial but challenging problem for event-centric natural language understanding, due to the reporting bias problem: people rarely report commonly observed events highlight special cases. For example, one may "I get up bed in 1 minute", we can observe "It takes me an hour every morning'' text. Models directly trained upon such corpus would capture distorted TCS knowledge, which could influence model...
Many open-domain question answering problems can be cast as a textual entailment task, where and candidate answers are concatenated to form hypotheses. A QA system then determines if the supporting knowledge bases, regarded potential premises, entail In this paper, we investigate neural-symbolic approach that integrates natural logic reasoning within deep learning architectures, towards developing effective yet explainable models. The proposed model gradually bridges hypothesis premises...
Collecting auction data on jinmajia.com, we make an analysis bidding behaviors in simultaneous ascending-bid auctions. Our results indicate that it is not optimal for bidders to use the late- and cross-bidding strategies large auctions with soft ending rule.
The maintenance of the entire life cycle wind turbines is important for a power project. main reason that whether can perform their best during operation period one key factors to measure success or failure farm investment. However, with increasing number installed turbines, how in an efficient way will become challenging problem. To lead better services, first step analyze process workflow really works real-life scenarios. Process mining such technique which analysis over event logs...
Through pretraining on a corpus with various sources, Large Language Models (LLMs) have gained impressive performance. However, the impact of each component remains opaque. As result, organization is still empirical and may deviate from optimal. To address this issue, we systematically analyze 48 datasets 5 major categories data LLMs measure their impacts using benchmarks about nine model capabilities. Our analyses provide results contribution multiple corpora performances LLMs, along joint...
Although achieving promising performance, recent analyses show that current generative large language models (LLMs) may still capture dataset biases and utilize them for generation, leading to poor generalizability harmfulness of LLMs. However, due the diversity over-optimization problem, previous prior-knowledge-based debiasing methods fine-tuning-based not be suitable To address this issue, we explore combining active learning with causal mechanisms propose a casual-guided (CAL) framework,...
Multi-stage amplifiers are widely applied in analog circuits. However, their large number of components, complex transfer functions, and intricate pole-zero distributions necessitate extensive manpower for derivation param sizing to ensure stability. In order achieve efficient the function simplify difficulty circuit design, we propose AmpAgent: a multi-agent system based on language models (LLMs) efficiently designing such from literature with process performance porting. AmpAgent is...