A5052849527- Topic Modeling
- Biomedical Text Mining and Ontologies
- Natural Language Processing Techniques
- Advanced Text Analysis Techniques
- Text and Document Classification Technologies
- Sentiment Analysis and Opinion Mining
- Machine Learning in Healthcare
- Misinformation and Its Impacts
- Text Readability and Simplification
- Machine Learning in Bioinformatics
- Computational Drug Discovery Methods
- 3D Surveying and Cultural Heritage
- Traditional Chinese Medicine Studies
- Advanced Image and Video Retrieval Techniques
- vaccines and immunoinformatics approaches
- 3D Shape Modeling and Analysis
- Advanced Computational Techniques and Applications
- Sustainable Development and Environmental Policy
- Remote Sensing and LiDAR Applications
- Complex Network Analysis Techniques
- Opinion Dynamics and Social Influence
- Web Data Mining and Analysis
- Advanced Graph Neural Networks
- Genetics, Bioinformatics, and Biomedical Research
- Image Retrieval and Classification Techniques
Hong Kong Polytechnic University
2020-2025
Soochow University
2012-2019
This article describes our work on the BioCreative-V chemical-disease relation (CDR) extraction task, which employed a maximum entropy (ME) model and convolutional neural network for at inter- intra-sentence level, respectively. In work, between entity concepts in documents was simplified to mentions. We first constructed pairs of chemical disease mentions as instances training testing stages, then we trained applied ME Finally, merged classification results from mention level document...
Abstract The coronavirus disease 2019 (COVID-19) pandemic has been severely impacting global society since December 2019. related findings such as vaccine and drug development have reported in biomedical literature—at a rate of about 10 000 articles on COVID-19 per month. Such rapid growth significantly challenges manual curation interpretation. For instance, LitCovid is literature database COVID-19-related PubMed, which accumulated more than 200 with millions accesses each month by users...
Understanding the relations between chemicals and diseases is crucial in various biomedical tasks such as new drug discoveries therapy developments. While manually mining these from literature costly time-consuming, a procedure often difficult to keep up-to-date. To address issues, BioCreative-V community proposed challenging task of automatic extraction chemical-induced disease (CID) order benefit biocuration. This article describes our work on CID relation tasks. We built machine learning...
Abstract This study investigates whether sentiment analysis, a natural language processing technique, can be used to examine accuracy in interpreting. The data were obtained from parallel bidirectional corpus of original speeches delivered at the United Nations and their simultaneous renditions provided by professional interpreters. Specifically, this explores how much conveyed across languages via accurate renditions, interpreting direction affects conveyance sentiment, analysis may help...
Abstract Extraction of causal relations between biomedical entities in the form Biological Expression Language (BEL) poses a new challenge to community text mining due complexity BEL statements. We propose simplified statements [Simplified (SBEL)] facilitate extraction and employ BERT (Bidirectional Encoder Representation from Transformers) improve performance relation (RE). On one hand, statement is transformed into an intermediate form—SBEL statement, which then further decomposed two...
Causal relation extraction of biomedical entities is one the most complex tasks in text mining, which involves two kinds information: entity relations and functions. One feasible approach to take function detection as independent sub-tasks. However, this separate learning method ignores intrinsic correlation between them leads unsatisfactory performance. In paper, we propose a joint model, combines exploit their commonality capture inter-relationship, so improve performance causal...
Automatically understanding chemical-disease relations (CDRs) is crucial in various areas of biomedical research and health care. Supervised machine learning provides a feasible solution to automatically extract between entities from scientific literature, its success, however, heavily depends on large-scale corpora manually annotated with intensive labor tremendous investment. We present an attention-based distant supervision paradigm for the BioCreative-V CDR extraction task. Training...
The Coronavirus Disease 2019 (COVID-19) pandemic has shifted the focus of research worldwide, and more than 10 000 new articles per month have concentrated on COVID-19-related topics. Considering this rapidly growing literature, efficient precise extraction main topics COVID-19-relevant is great importance. manual curation information for biomedical literature labor-intensive time-consuming, as such procedure insufficient difficult to maintain. In response these complications, BioCreative...
The COVID-19 pandemic has been severely impacting global society since December 2019. Massive research undertaken to understand the characteristics of virus and design vaccines drugs. related findings have reported in biomedical literature at a rate about 10,000 articles on per month. Such rapid growth significantly challenges manual curation interpretation. For instance, LitCovid is database COVID-19-related PubMed, which accumulated more than 200,000 with millions accesses each month by...
Biomedical event extraction is an information task to obtain events from biomedical text, whose targets include the type, trigger, and respective arguments involved in event. Traditional usually adopts a pipelined approach, which contains trigger identification, argument role recognition, finally construction either using specific rules or by machine learning. In this paper, we propose n-ary relation method based on BERT pre-training model construct Binding events, order capture semantic...
In recent years, biomedical event extraction has been dominated by complicated pipeline and joint methods, which need to be simplified. addition, existing work not effectively utilized trigger word information explicitly. Hence, we propose MLSL, a method based on multi-layer sequence labeling for extraction. MLSL does introduce prior knowledge complex structures. Moreover, it explicitly incorporates the of candidate words into learn interaction relationships between argument roles. Based...
Abstract Background The COVID-19 pandemic has increasingly accelerated the publication pace of scientific literature. How to efficiently curate and index this large amount biomedical literature under current crisis is great importance. Previous indexing mainly performed by human experts using Medical Subject Headings (MeSH), which labor-intensive time-consuming. Therefore, alleviate expensive time consumption monetary cost, there an urgent need for automatic semantic technologies emerging...