A5028351629- Topic Modeling
- Natural Language Processing Techniques
- Text and Document Classification Technologies
- Advanced Text Analysis Techniques
- Advanced Graph Neural Networks
- Data Quality and Management
- Multimodal Machine Learning Applications
- Domain Adaptation and Few-Shot Learning
- Anomaly Detection Techniques and Applications
- Sentiment Analysis and Opinion Mining
- Authorship Attribution and Profiling
- Advanced Algorithms and Applications
- Fungal Biology and Applications
- Artificial Intelligence in Law
- Medicinal plant effects and applications
- Acoustic Wave Resonator Technologies
- Adversarial Robustness in Machine Learning
- Traffic Prediction and Management Techniques
- Image Processing and 3D Reconstruction
- Advanced MEMS and NEMS Technologies
- Fire Detection and Safety Systems
- Financial Distress and Bankruptcy Prediction
- Biomedical Text Mining and Ontologies
- Advanced Chemical Sensor Technologies
- Misinformation and Its Impacts
Westlake University
2022-2025
Institute of Information Engineering
2020-2024
Chinese Academy of Sciences
2020-2024
University of Chinese Academy of Sciences
2020-2024
Zhejiang University
2022-2023
University of Jinan
2015-2019
Event detection (ED), a key subtask of information extraction, aims to recognize instances specific event types in text. Previous studies on the task have verified effectiveness integrating syntactic dependency into graph convolutional networks. However, these methods usually ignore label information, which conveys rich and useful linguistic knowledge for ED. In this paper, we propose novel architecture named Edge-Enhanced Graph Convolution Networks (EE-GCN), simultaneously exploits...
Event extraction (EE) is a crucial research task for promptly apprehending event information from massive textual data. With the rapid development of deep learning, EE based on learning technology has become hotspot. Numerous methods, datasets, and evaluation metrics have been proposed in literature, raising need comprehensive updated survey. This article fills gap by reviewing state-of-the-art approaches, especially focusing general domain models. We introduce new literature classification...
This paper studies the multimodal named entity recognition (MNER) and relation extraction (MRE), which are important for content analysis various applications. The core of MNER MRE lies in incorporating evident visual information to enhance textual semantics, where two issues inherently demand investigations. first issue is modality-noise, task-irrelevant each modality may be noises misleading task prediction. second modality-gap, representations from different modalities inconsistent,...
Event detection tends to struggle when it needs recognize novel event types with a few samples.The previous work attempts solve this problem in the identify-then-classify manner but ignores trigger discrepancy between types, thus suffering from error propagation.In paper, we present unified model which converts task few-shot tagging double-part scheme.To end, first propose Prototypical Amortized Conditional Random Field (PA-CRF) label dependency scenario, approximates transition scores...
Aspect sentiment quad prediction (ASQP) aims to predict the elements for a given sentence, which is critical task in field of aspect-based analysis. However, data imbalance issue has not received sufficient attention ASQP task. In this paper, we divide into two-folds, quad-pattern and aspect-category imbalance, propose an Adaptive Data Augmentation (ADA) framework tackle issue. Specifically, augmentation process with condition function adaptively enhances tail patterns aspect categories,...
Artificial Intelligence (AI) systems are increasingly intertwined with daily life, assisting users in executing various tasks and providing guidance on decision-making. This integration introduces risks of AI-driven manipulation, where such may exploit users' cognitive biases emotional vulnerabilities to steer them toward harmful outcomes. Through a randomized controlled trial 233 participants, we examined human susceptibility manipulation financial (e.g., purchases) conflict resolution)...
Multi-Modal Relation Extraction (MMRE) aims at identifying the relation between two entities in texts that contain visual clues. Rich content is valuable for MMRE task, but existing works cannot well model finer associations among different modalities, failing to capture truly helpful information and thus limiting extraction performance. In this paper, we propose a novel framework better deeper correlations of text, entity pair, image/objects, so as mine more termed DGF-PT. We first...
Event Causality Identification (ECI), which aims to detect whether a causality relation exists between two given textual events, is an important task for event understanding. However, the ECI ignores crucial structure and cause-effect component information, making it struggle downstream applications. In this paper, we explore novel task, namely Extraction (ECE), aiming extract pairs with their structured information from plain texts. The ECE more challenging since each can contain multiple...
In few-shot relational triple extraction (FS-RTE), one seeks to extract triples from plain texts by utilizing only few annotated samples. Recent work first extracts all entities and then classifies their relations. Such an entity-then-relation paradigm ignores the entity discrepancy between To address it, we propose a novel task decomposition strategy, Relation-then-Entity, for FS-RTE. It detects relations occurred in sentence corresponding head/tail of detected instantiate this further...
Event detection (ED) is a pivotal task for information retrieval, which aims at identifying event triggers and classifying them into pre-defined types. In real-world applications, events are usually annotated with numerous fine-grained types, often arises long-tail type nature co-occurrence nature. Existing studies explore the correlations without full utilization, may limit capability of detection. This paper simultaneously incorporates both type-level instance-level correlations, proposes...
Phellinus baumii, a fungus that grows on mulberry trees and is used in traditional Chinese medicine, exerts therapeutic effects against various diseases, including cancer. Polyphenols, generally considered to be antioxidants, have antitumor proapoptotic effects. In this study, we identified the composition of baumii polyphenol (PBP) characterized its 17 chemical components by UPLC-ESI-QTOF-MS. Furthermore, clarify potential mechanism PBP Lung Cancer Cells, network pharmacology experimental...
Event Detection (ED) aims to recognize instances of specified types event triggers in text. Different from English ED, Chinese ED suffers the problem word-trigger mismatch due uncertain word boundaries. Existing approaches injecting information into character-level models have achieved promising progress alleviate this problem, but they are limited by two issues. First, interaction between characters and lexicon words is not fully exploited. Second, ignore semantic provided labels. We thus...
Event extraction (EE) is a crucial research task for promptly apprehending event information from massive textual data. With the rapid development of deep learning, EE based on learning technology has become hotspot. Numerous methods, datasets, and evaluation metrics have been proposed in literature, raising need comprehensive updated survey. This article fills gap by reviewing state-of-the-art approaches, especially focusing general domain models. We introduce new literature classification...
Event detection (ED), a key subtask of information extraction, aims to recognize instances specific event types in text. Previous studies on the task have verified effectiveness integrating syntactic dependency into graph convolutional networks. However, these methods usually ignore label information, which conveys rich and useful linguistic knowledge for ED. In this paper, we propose novel architecture named Edge-Enhanced Graph Convolution Networks (EE-GCN), simultaneously exploits...
Document-level Event Extraction (DEE) is particularly tricky due to the two challenges it poses: scattering-arguments and multi-events. The first challenge means that arguments of one event record could reside in different sentences document, while second reflects document may simultaneously contain multiple such records. Motivated by humans' reading cognitive extract information interests, this paper, we propose a method called HRE (Human Reading inspired Extractor for Document Events),...
Aspect sentiment quad prediction (ASQP) aims to predict the elements for a given sentence, which is critical task in field of aspect-based analysis. However, data imbalance issue has not received sufficient attention ASQP task. In this paper, we divide into two-folds, quad-pattern and aspect-category imbalance, propose an Adaptive Data Augmentation (ADA) framework tackle issue. Specifically, augmentation process with condition function adaptively enhances tail patterns aspect categories,...
Chinese Spelling Correction (CSC) aims to detect and correct the misspelled characters in texts. Recent studies have achieved great success by incorporating phonetic information for task predictions. Still, existing methods suffer from two limitations: 1) The differentiated between textual Pinyin pronunciation are underexplored. 2) predictions performed with over-emphasises on either or sequence, ignoring balanced modeling adaptive fusion. In this work, we proposed a method alleviate issues...