A5101765506- Multimodal Machine Learning Applications
- Natural Language Processing Techniques
- Topic Modeling
- Data Stream Mining Techniques
- Advanced Image and Video Retrieval Techniques
- Advanced Text Analysis Techniques
- Machine Learning and Data Classification
- Anomaly Detection Techniques and Applications
- Video Analysis and Summarization
- Academic integrity and plagiarism
- Data Management and Algorithms
- Handwritten Text Recognition Techniques
- Fungal Biology and Applications
- Tensor decomposition and applications
- Microbial Natural Products and Biosynthesis
- Satellite Communication Systems
- Digital Media and Visual Art
- Generative Adversarial Networks and Image Synthesis
- Wireless Communication Networks Research
- Data Mining Algorithms and Applications
- Aesthetic Perception and Analysis
- Stock Market Forecasting Methods
- Sentiment Analysis and Opinion Mining
- Advanced Graph Neural Networks
- Artificial Intelligence in Healthcare
Hunan University
2022-2024
Hainan Normal University
2024
Simon Fraser University
2020-2021
University of Toronto
2021
Beijing University of Posts and Telecommunications
2021
Tongji University
2015-2019
Statistical Constraints (SCs) play an important role in statistical modeling and analysis. This paper brings the concept to data cleaning studies how leverage SCs for error detection. provide a novel approach that has various application scenarios works harmoniously with downstream modeling. Entailment relationships between integrity constraints analytical insight into SCs. We develop SCODED, SC-Oriented Data Error Detection system, comprising two key components: (1) SC Violation : checks...
Aspect-based financial sentiment analysis, which aims to classify the text instance into a pre-defined aspect class and predict score for mentioned target. In this paper, we propose neural network model, Attention-based LSTM model with Aspect information (ALA), solve opinion mining problem introduced by WWW 2018 shared task. The proposed can adapt dataset so that effectively understand semantic of short text. We evaluate our 10-fold cross-validation, compare variety related deep models.
The wide sharing and rapid dissemination of digital artworks has aggravated the issues plagiarism, raising significant concerns in cultural preservation copyright protection. Yet, modes plagiarism are formally uncharted, causing rough detection practices with duplicate checking. This work is thus devoted to understanding artwork poster design as running case, for building more dedicated techniques. As first study such, we elaborate on 8 elements that form unique posters 6 judgement criteria...
Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI
Abstract Existing research treated all sentences in the text on an equal basis during training process and did not consider that key tend to have a stronger influence. We propose Convolutional Neural Network sentiment classification model based enhancement. The proposed can identify generate representation these reduce noise improve accuracy of classification. experiment results show improves compared with other classic models.
Mohan Zhang, Luchen Tan, Zihang Fu, Kun Xiong, Jimmy Lin, Ming Li, Zhengkai Tu. Proceedings of the 16th Conference European Chapter Association for Computational Linguistics: Main Volume. 2021.
Error detection is key for data quality management. AI techniques can leverage user domain knowledge to identifying sets of erroneous records that conflict with knowledge. To represent a wide range knowledge, several recent papers have developed and utilized soft approximate constraints (ACs) relation expected satisfy only certain degree, rather than completely. We introduce error localization, new AI-based technique enhancing ACs.
This paper proposes a modular tensor sparsity preserving projection (MTSPP) algorithm. algorithm uniformly partitions the high-dimensional matrix data and builds third order data, determines weight of sparse reconstruction all samples applies it to tensor. Experiments finally indicate that MTSPP improves robust performance global representation-based dimension reduction by weighted representation spatial relationship characteristics within module between modules.
We focus on the problem of learning distributed representations for entity search queries, named entities, and their short descriptions. With our representation models, query, description can be represented as low-dimensional vectors. Our goal is to develop a simple but effective model that make query related entities similar in vector space. Hence, we propose three kinds strategies, difference between them mainly lies how deal with relationship an its description. analyze strengths...
In the paraphrase generation task, source sentences often contain phrases that should not be altered. Which phrases, however, can context dependent and vary by application. Our solution to this challenge is provide user with explicit tags placed around any arbitrary segment of text mean "don't change me!" when generating a paraphrase; model learns explicitly copy these output. The contribution work novel data technique using distant supervision allows us start pretrained sequence-to-sequence...
Next generation satellite communications (NGSC) evolved by 5G NR combination is the hot issue in recent researches. For conditions of new frequency (Ku/Ka/Q/V) and much more broadband (>100MHz) for mobile communications, obvious nonlinearity Input Multiplexing (IMUX) filter, High Power Amplifier (HPA), Output (OMUX) filter will cause serious signals degradation. Traditional behavioral modeling methods IMUX HPA, OMUX (IHO) are relatively independent each other, which difficult to...