A5101690738- Image Retrieval and Classification Techniques
- Topic Modeling
- Advancements in Battery Materials
- Natural Language Processing Techniques
- Text and Document Classification Technologies
- Recommender Systems and Techniques
- Complex Network Analysis Techniques
- Web Data Mining and Analysis
- Advanced Image and Video Retrieval Techniques
- Video Analysis and Summarization
- Supercapacitor Materials and Fabrication
- Advanced Battery Materials and Technologies
- Digital Media Forensic Detection
- Data Mining Algorithms and Applications
- Robotic Path Planning Algorithms
- Advanced Steganography and Watermarking Techniques
- Robot Manipulation and Learning
- Chaos-based Image/Signal Encryption
- Sentiment Analysis and Opinion Mining
- Music and Audio Processing
- Sustainable Supply Chain Management
- Corporate Social Responsibility Reporting
- Software Reliability and Analysis Research
- Software System Performance and Reliability
- Advanced Graph Neural Networks
Harbin Institute of Technology
2015-2025
Southwest Jiaotong University
2023-2025
Fujian Normal University
2024
Nanjing University of Information Science and Technology
2023
Beijing Institute of Technology
2022
Beijing Institute of Fashion Technology
2022
Jinan University
2019-2022
China Academy of Information and Communications Technology
2021
Huawei Technologies (China)
2016-2021
South China University of Technology
2019
With the urgent market demand for high-energy-density batteries, alloy-type or conversion-type anodes with high specific capacity have gained increasing attention to replace current low-specific-capacity graphite-based anodes. However, and large initial irreversible compared anodes, which consume most of Li+ in corresponding cathode severely reduces energy density full cells. Therefore, practical application these high-capacity it is develop a commercially available prelithiation technique...
This paper presents a character-level seque-nce-to-sequence learning method, RNNembed. method allows the system to read raw characters, instead of words generated by preprocessing steps, into pure single neural network model under an end-to-end framework. Specifically, we embed recurrent encoder–decoder framework and generate sequence representation as input. The dimension input feature space can be significantly reduced well avoiding need handle unknown or rare in sequences. In language...
Microblogging service has emerged to be a dominant web medium for billions of individuals sharing and spreading instant news information, therefore monitoring the event evolution on microblog sphere is crucial providing both better user experience deeper understanding real-time events. In this paper we explore problem generating storylines from microblogs input queries. This challenging due sparse, dynamic social nature microblogs. Given query an ongoing event, propose sketch storyline by...
Due to the high energy density and low production cost, silicon oxide (SiOx) is recognized as one of most promising anode materials for lithium-ion batteries. However, initial coulombic efficiency rapid capacity attenuation SiOx anodes severely restricts its commercial application. Here, we propose a scalable strategy pre-lithiation followed by thermal passivation improve anodes, boosting large-scale applications strategy. First, hollow porous SiOx@C spheres (Hp-SiOx@C) with adjustable shell...
Environmental, Social, and Governance (ESG) ratings of firms are divergent across data providers, which creates uncertainty about corporate sustainability. Using ESG rating received by Chinese listed companies, we find that issued different agencies generally have a low correlation. Empirical tests show disagreement inhibits firms' green innovation behavior. Corporate resources intentions potential channels through affects innovation. Further analysis finds the information environment,...
Purpose Labor-related risks resulting from layoffs may pique auditors’ scrutiny. Although previous research has enriched the understanding of economic consequences layoffs, authors know relatively little about relationship between and audit fees. This study aims to investigate whether auditors are concerned corporate layoff events their pricing decisions under influence events. Design/methodology/approach examines effect storms on fees using news reports mainstream financial media in China...
Ferroelectric fluorite oxides such as hafnium (HfO2)-based materials are one of the most promising candidates for future large-scale integrated circuits (ICs), while zirconia (ZrO2)-based materials, which have same structure HfO2 and more abundant resources lower cost raw usually thought to be anti- or ferroelectric-like. Here, we report a remanent polarization (Pr) ∼15 μC/cm2 in orthorhombic ZrO2 thin films at 77 K. This ferroelectricity arises from an electric field-induced...
A new triple wing harmonium (TWH) model that integrates text metadata into a low-dimensional semantic space is proposed for the application of content-based movie recommendation. The considered here include synopsis, actor list, and user comments. We develop TWH projecting these multiple textual features latent topics with different probability distribution assumptions. contrastive divergence (CD) algorithm used efficient learning inference. Experimental results suggest method performs...
Both sentiment analysis and topic classification are frequently used in customer care marketing. They can help people understand the brand perception opinions from social media, such as online posts, tweets, forums, blogs. As such, recent years, many solutions have been proposed for both tasks. However, we believe that following two problems not addressed adequately: (1) Conventional usually treat tasks isolation. When closely related (e.g., posts about "customer care" often a "negative"...
In home-based elderly care service, how to precisely recognize activities is a key issue in the design and implementation of context-aware service for people. Existing research works reveal that those approaches ignore characteristics activity diversity, similarity features people at home, so recognition accuracy are not high enough real-life applications. Thus, this paper, we first study types service. Then, propose two-stage home method based on random forest similarity. The uses improved...
The advent of Big Data era drives data analysts from different domains to use mining techniques for analysis. However, performing analysis in a specific domain is not trivial; it often requires complex task configuration, onerous integration algorithms, and efficient execution distributed environments.Few efforts have been paid on developing effective tools facilitate conducting tasks.
Advanced manufacturing such as aerospace, semi-conductor, and flat display device often involves complex production processes, generates large volume of data. In general, the data comes from products with different levels quality, assembly line flows equipments, processing craft massive controlling parameters. The scale complexity is beyond analytic power traditional IT infrastructures. To achieve better performance, it imperative to explore underlying dependencies exploit insights improve...
In music information retrieval (MIR) an important research topic, which has attracted much attention recently, is the utilization of user-assigned tags, artist-related style, and mood labels, can be extracted from listening web sites, such as Last.fm (http://www.last.fm/) All Music Guide (http://www.allmusic.com/). A fundamental problem in area how to understand relationships among artists/songs these different pieces information. Co-clustering simultaneously clustering two types data (e.g.,...
Multi-document summarization aims to distill the most representative information from a set of documents generate summary. Given as input, existing multi-document approaches utilize different sentence selection techniques extract sentences document The submodularity hidden in textual-unit similarity motivates us incorporate this property into our solution tasks. In poster, we propose new principled and versatile framework for tasks using submodular function [8].
Given a collection of web images with the corresponding textual descriptions, in this paper, we propose novel cross-domain learning method to classify these multimedia objects by transferring correlation knowledge among different information sources. Here, is extracted from unlabeled through unsupervised and applied perform supervised classification tasks. To mine more meaningful knowledge, instead using commonly used visual words traditional bag-of-visual-words (BoW) model, discover higher...
The tremendous growth and popularity of social media platforms like Twitter, Facebook, etc. provides business organizations an opportunity to monitor the feedback from its customers, identify their problems take corrective measures. In this paper, we describe a system automatically analyze customer through various detect issues faced by customers. Business can use engage with customers help alleviate them. uses statistical event detection techniques for identifying issues. offers batch...
To address the significant challenges encountered by silicon-based anodes in high-performance lithium-ion batteries (LIBs), including poor cycling stability, low initial coulombic efficiency (ICE), and insufficient interface compatibility, this work innovatively prepares Si/SiOx@F-C composites via situ coating fluorine-doping carbon layer on Si/SiOx surface through high-temperature pyrolysis. The Si/SiO