A5101820549- Sentiment Analysis and Opinion Mining
- Topic Modeling
- Text and Document Classification Technologies
- Natural Language Processing Techniques
- Advanced Text Analysis Techniques
- Coding theory and cryptography
- Cryptographic Implementations and Security
- Mental Health via Writing
- Chaos-based Image/Signal Encryption
- Humor Studies and Applications
- Cryptography and Data Security
- Machine Learning and ELM
- Language, Metaphor, and Cognition
- Biomedical Text Mining and Ontologies
- Cryptography and Residue Arithmetic
- Digital Mental Health Interventions
- Domain Adaptation and Few-Shot Learning
- Suicide and Self-Harm Studies
- Stock Market Forecasting Methods
- Information Retrieval and Search Behavior
- Image Retrieval and Classification Techniques
- Expert finding and Q&A systems
- Web Data Mining and Analysis
- Mobile Crowdsensing and Crowdsourcing
- Advanced Data Storage Technologies
Dalian University of Technology
2015-2024
PRG S&Tech (South Korea)
2021
Dalian University
2019-2020
Wuhan Engineering Science & Technology Institute
2015-2016
Zhejiang Science and Technology Information Institute
2015-2016
Jilian Technology Group (China)
2011
PLA Information Engineering University
2007-2011
Jinan University
2009
Hubei Normal University
2006
Northwest University
2005
Background As a common mental disease, depression seriously affects people’s physical and health. According to the statistics of World Health Organization, is one main reasons for suicide self-harm events in world. Therefore, strengthening detection can effectively reduce occurrence or so as save more people families. With development computer technology, some researchers are trying apply natural language processing techniques detect who depressed automatically. Many existing feature...
Metaphors are frequently used to convey emotions. However, there is little research on the construction of metaphor corpora annotated with emotion for analysis emotionality metaphorical expressions. Furthermore, most studies focus English, and few in other languages, particularly Sino-Tibetan languages such as Chinese, from texts, although likely be many differences emotional expressions usages across different languages. We therefore construct a significant new corpus metaphor, 5,605...
Clinical notes record the health status, clinical manifestations and other detailed information of each patient. The International Classification Diseases (ICD) codes are important labels for electronic records. Automatic medical assignment to through deep learning model can not only improve work efficiency accelerate development informatization but also facilitate resolution many issues related insurance. Recently, neural network-based methods have been proposed automatic code assignment....
Yufeng Diao, Hongfei Lin, Di Wu, Liang Yang, Kan Xu, Zhihao Jian Wang, Shaowu Zhang, Bo Dongyu Zhang. Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing. 2018.
Pre-trained distributed word representations have been proven useful in various natural language processing (NLP) tasks. However, the effect of words’ geometric structure on has not carefully studied yet. The existing methods underestimate words whose distances are close Euclidean space, while overestimating with a much greater distance. In this paper, we propose vector refinement model to correct pre-trained embedding, which brings similarity space closer semantics by using manifold...
Journal Article Cryptanalysis of WG Family Stream Ciphers Get access Lin Ding, Ding * 1Zhengzhou Information Science and Technology Institute, 450000 Zhengzhou, China *Corresponding author: dinglin_cipher@163.com Search for other works by this author on: Oxford Academic Google Scholar Chenhui Jin, Jin Jie Guan, Guan Shaowu Zhang, Zhang Ting Cui, Cui Dong Han, Han Wei Zhao 2Science on Communication Security Laboratory, 610041 Chengdu, The Computer Journal, Volume 58, Issue 10, October 2015,...
Background Suicide has become the fifth leading cause of death worldwide. With development internet, social media an imperative source for studying psychological illnesses such as depression and suicide. Many methods have been proposed suicide risk assessment. However, most existing cannot grasp key information text. To solve this problem, we propose efficient method to extract core from posts Objective We developed a multifeature fusion recurrent attention model Methods used bidirectional...
Metaphorical expressions are frequently used to convey emotions in human communication. However, there is limited research on the detection of emotionality metaphorical expressions, although a number studies have focused sentiment analysis and metaphor separately. We, therefore, attempt identify Chinese texts. We first construct manual corpus with an annotation scheme, which contains annotations metaphor, emotional categories. then use as train-and-test set automatically three methods. The...
Metaphor is the central issue of language and thinking. identification plays a significant preliminary role in field machine translation, reading comprehension, automatic summarization, making it focus natural processing. Recently, research into Chinese verb metaphor has become widespread concern. The main problem that usage semantic resources relatively simple, there lack deep support. Therefore, this paper proposes word representation method suitable for classification tasks, which...
The Grain v1 stream cipher is one of the seven finalists in final eSTREAM portfolio. Though many attacks have been published, no recovery attack better than exhaustive key search on full single setting has found yet. In this paper, new state utilizing weak normality order employed keystream output function are proposed. These remarkable advantages offline time, online time and memory complexities, which all search. success probability each 0.632. proposed primarily depends function. This...
Abstract Linear cryptanalysis is one of the most important methods and remains vital in cryptography research up to now. Many derivatives, improvements, generations linear have been introduced attack ciphers different designs. This paper defines a new characteristic cryptographic function, called 2‐bit‐fixed characteristic, as generation characteristic. Compared with it restricts input more severely obtain larger output bias. Using this gives multidimensional on 27‐round PRESENT computation...
Observing that the Haar wavelet is sensitive to step edges, and CDF(2.2) which a subset of Cohen-Daubechies-Feauveau family performs well for smooth signals, we present new adaptive transform via lifting scheme noise reduction. Unlike many popular adapt scale-by-scale, proposed can point-by-point. Experiment results show method has advantages both CDF(2,2) wavelets edge-dominated regions. Moreover, interval switch thresholding parameter given by experiments.
An extreme learning machine (ELM) is a useful technique for learning; however, the existing methods cannot exploit geometric structure information or discriminate of data space well. Therefore, we propose globality‐locality preserving maximum variance (GLELM) based on manifold learning. Based characteristics traditional ELM method, GLELM introduces basic principles linear discriminant analysis (LDA) and local preservation projection (LPP) into ELM, fully taking account contained in sample....