A5058291454- Topic Modeling
- Sentiment Analysis and Opinion Mining
- Advanced Text Analysis Techniques
- Text and Document Classification Technologies
- Natural Language Processing Techniques
- Computational and Text Analysis Methods
- Recommender Systems and Techniques
- Spam and Phishing Detection
- Complex Network Analysis Techniques
- Web Data Mining and Analysis
- Speech and Audio Processing
- Advanced Graph Neural Networks
- Image Retrieval and Classification Techniques
- Online Learning and Analytics
- Face and Expression Recognition
- Speech Recognition and Synthesis
- Advanced Image and Video Retrieval Techniques
- Misinformation and Its Impacts
- Music and Audio Processing
- Advanced Data Compression Techniques
- Metaheuristic Optimization Algorithms Research
- Expert finding and Q&A systems
- Domain Adaptation and Few-Shot Learning
- Mobile Crowdsensing and Crowdsourcing
- Advanced Multi-Objective Optimization Algorithms
Sun Yat-sen University
2015-2024
Guangzhou University
2024
City University of Hong Kong
2012-2022
University of Hradec Králové
2022
Peking University
2022
Khon Kaen University
2022
University of Trnava
2022
Conference Board
2022
Kansai University
2022
Kasetsart University
2022
Social emotion classification is important for numerous applications, such as public opinion measurement, corporate reputation estimation, and customer preference analysis. However, topics that evoke a certain in the general are often context-sensitive, making it difficult to train universal classifier all collections. A multilabeled sentiment topic model, namely, contextual model (CSTM), can be used adaptive social classification. The CSTM distinguishes context-independent from both...
With the rapid development of Internet, an increasing number users enjoy to shop online and express their reviews on products services. Analysis these can not only help potential make rational decisions when purchasing but also improves quality Hence, sentiment analysis for has become important meaningful research domain.
With the growing availability and popularity of sentiment-rich resources like blogs online reviews, new opportunities challenges have emerged regarding identification, extraction, organization sentiments from user-generated documents or sentences. Recently, many studies exploited lexicon-based methods supervised learning algorithms to conduct sentiment analysis tasks separately; however, former approaches ignore contextual information sentences latter ones do not take embedded in words into...
Social emotion classification aims to predict the aggregation of emotional responses embedded in online comments contributed by various users. Such a task is inherently challenging because extracting relevant semantics from free texts classical research problem. Moreover, are typically characterized sparse feature space, which makes corresponding very difficult. On other hand, though deep neural networks have been shown be effective for speech recognition and image analysis tasks their...
With the development of social network platforms, discussion forums, and question answering websites, a huge number short messages that typically contain few words for an individual document are posted by online users. In these messages, emotions frequently embedded communicating opinions, expressing friendship, promoting influence. It is quite valuable to detect from but corresponding task suffers sparsity feature space. this article, we first generate term groups co-occurring in same...
Copy number variation (CNV) is an important source of genetic in organisms and a main factor that affects phenotypic variation. A comprehensive study chicken CNV can provide valuable information on diversity facilitate future analyses associations between economically traits chickens. In the present study, F2 full-sib population (554 individuals), established from cross Xinghua White Recessive Rock chickens, was used to explore genome. Genotyping performed using 60K SNP BeadChip. total 1,875...
Cross-domain sentiment classification aims to tag sentiments for a target domain by labeled data from source domain. Due the difference between domains, accuracy of trained classifier may be very low. In this paper, we propose boosting-based learning framework named TR-TrAdaBoost cross-domain classification. We firstly explore topic distribution documents, and then combine it with unigram TrAdaBoost. The captures information which is valuable Experimental results indicate that represents...
Traditional methods of annotating the sentiment an unlabeled document are based on lexicons or machine learning algorithms, which have shown low computational cost competitive performance. However, these ignore semantic composition problem displaying in several ways such as negative reversing and intensification. In this paper, we propose a new method for classification using intensive supplementary information, so to exploit linguistic feature words conjunction with context information....
Broodiness is a polygenic trait controlled by small number of autosomal genes. Vasoactive intestinal peptide receptor-1 (VIPR-1) gene could be candidate chicken broodiness, and its genomic variations genetic effects on broodiness traits were analyzed in this study. The partial cloning sequencing the VIPR-1 showed that average nucleotide diversity was 0.00669 ± 0.00093 Red Jungle Fowls (RJF), 0.00582 0.00026 domestic chickens. One hundred twenty-eight variation sites identified 11,136-bp...
As a novel paradigm for data mining and dimensionality reduction, Non-negative Matrix Tri-Factorization (NMTF) has attracted much attention due to its notable performance elegant mathematical derivation, it been applied plethora of real-world applications, such as text co-clustering. However, the existing NMTF-based methods usually involve intensive matrix multiplications, which exhibits major limitation high computational complexity. With explosion at both size feature dimension texts,...
This paper presents an approach to the recognition of speech signal using frequency spectral information with mel for improvement feature representation in a HMM based approach. The exploits observation given resolution which results overlapping resulting limit. Resolution decomposition mapping system. Simulation show quality metrics wrt. computational time, learning accuracy