A5103166668- Information Retrieval and Search Behavior
- Data Management and Algorithms
- Topic Modeling
- Advanced Image and Video Retrieval Techniques
- Caching and Content Delivery
- Web Data Mining and Analysis
- SARS-CoV-2 and COVID-19 Research
- Construction Project Management and Performance
- interferon and immune responses
- Image Retrieval and Classification Techniques
- Advanced Graph Neural Networks
- Resource-Constrained Project Scheduling
- IoT and Edge/Fog Computing
- Data Mining Algorithms and Applications
- Cancer Cells and Metastasis
- NF-κB Signaling Pathways
- BIM and Construction Integration
- Semantic Web and Ontologies
- Higher Education and Teaching Methods
- Advanced Database Systems and Queries
- Pharmacological Effects of Natural Compounds
- Complex Network Analysis Techniques
- Natural Language Processing Techniques
- Multi-Criteria Decision Making
- Computational Drug Discovery Methods
Gannan Medical University
2021-2024
Microsoft Research (United Kingdom)
2023
RMIT University
2014-2019
National Institute of Development Administration
2019
University of Southern California
2013-2015
Shanghai Jiao Tong University
2014
Nanjing University
2014
State Key Laboratory of Virology
2010
Wuhan University
2008-2010
Nanjing Institute of Industry Technology
2008
Severe acute respiratory syndrome coronavirus (SARS-CoV) encodes a highly basic nucleocapsid (N) protein which can inhibit the synthesis of type I interferon (IFN), but molecular mechanism this antagonism remains to be identified. In study, we demonstrated that N SARS-CoV could IFN-beta (IFN-β) induced by poly(I:C) or Sendai virus. However, found not IFN-β production overexpression downstream signaling molecules two important induction pathways, toll-like receptor 3 (TLR3)- and RIG-I-like...
Over the past 20 years, severe acute respiratory syndrome coronavirus (SARS-CoV), Middle East CoV (MERS-CoV), and SARS-CoV-2 emerged, causing human diseases throughout globe. Developing broad-spectrum drugs would be invaluable in responding to new, emerging coronaviruses address unmet urgent clinical needs. Main protease (Mpro; also known as 3CLpro) has a major role life cycle is one of most important targets for anti-coronavirus agents. We show that natural product, noncovalent inhibitor,...
Direct answering of questions that involve multiple entities and relations is a challenge for text-based QA. This problem most pronounced when answers can be found only by joining evidence from documents. Curated knowledge graphs (KGs) may yield good answers, but are limited their inherent incompleteness potential staleness. paper presents QUEST, method answer complex directly textual sources on-the-fly, computing similarity joins over partial results different Our completely unsupervised,...
G protein-coupled receptors (GPCRs) are seven-transmembrane proteins that transmit diverse extracellular signals across a membrane. Herpesvirus genomes encode multiple GPCRs implicated in viral pathogenesis. Kaposi sarcoma-associated herpesvirus GPCR (kGPCR) activates proliferative pathways and, when expressed endothelium mice, sufficiently induces angiogenic tumor resembling human Kaposi’s sarcoma. IKKε, an IκB kinase (IKK)-related kinase, is inflammation-driven tumorigenesis. We report...
As one of the most popular services over online platforms, social recommendation has attracted increasing research efforts recently. Among all tasks, an important is item high speed media streams. Existing stream techniques are not effective for handling users with diverse interests. Meanwhile, approaches recommending items to a particular user efficient when applied huge number In this paper, we propose novel framework Specifically, first Bi-Layer Hidden Markov Model (BiHMM) that adaptively...
Customer services are critical to all companies, as they may directly connect the brand reputation. Due a great number of customers, e-commerce companies often employ multiple communication channels answer customers' questions, for example, Chatbot and Hotline. On one hand, each channel has limited capacity respond requests; on other customers have different preferences over these channels. The current production systems mainly built based business rules that merely consider tradeoffs...
Human gamma herpesviruses, including Kaposi's sarcoma-associated herpesvirus (KSHV) and Epstein-Barr virus (EBV), are capable of inducing tumors, particularly in immune-compromised individuals. Due to the stringent host tropism, rodents resistant infection by human creating a significant barrier for vivo study viral genes that contribute tumorigenesis. The closely-related murine 68 (γHV68) efficiently infects laboratory mouse strains establishes robust persistent without causing apparent...
We present an in-depth comparative analysis of the effectiveness distributions sets human-created and automatically-created query variations used to represent same information need. The automatic are generated using Bing's click graph. Experiments performed with TREC datasets show that for retrieval can result in similar human variations, although two types be appreciably different several important respects --- e.g., their similarities corresponding retrieved lists.
Sophisticated ranking mechanisms make use of term dependency features in order to compute similarity scores for documents. These often include exact phrase occurrences, and proximity estimates. Both cases build on the intuition that if multiple query terms appear near each other, document is more likely be relevant query. In this paper we examine processes used these statistics. Two distinct input structures can -- inverted files direct files. Inverted must store position offsets terms,...
Bag-of-words retrieval models are widely used, and provide a robust trade-off between efficiency effectiveness. These often make simplifying assumptions about relations query terms, treat term statistics independently. However, terms rarely independent, previous work has repeatedly shown that dependencies can be critical to improving the effectiveness of ranked results. Among all term-dependency models, Markov Random Field (MRF) [Metzler Croft, SIGIR, 2005] model received most attention in...
Information Retrieval has a long history of applying either discriminative or generative modeling to retrieval and ranking tasks. Recent developments in transformer architectures multi-task learning techniques have dramatically improved our ability train effective neural models capable resolving wide variety tasks using these paradigms. In this paper, we propose novel approach which can be used produce more models. The key idea is improve the quality underlying model by cross-training task...
The generative theory for relevance and its operational manifestation --- the model are based on premise that a single query is used to represent an information need retrieval. In this work, we extend devise novel techniques modeling using as set of variations representing same need. Our new approach fusion at term level, or document-list level. We theoretically analyze connections between these methods provide empirical support their equivalence TREC datasets. Specifically, our inducing...
Colorectal cancer (CRC) is the second most common cause of cancer-related mortality and lies third in terms morbidity due to limited number effective druggable targets. Since stem cells (CSCs) are considered be one roots tumorigenesis, outgrowth metastasis, targeting CSCs may a promising strategy reverse malignant phenotypes CRC. Cyclin-dependent kinase 12 (CDK12) has been reported involved self-renewal various cancers, rendering it an attractive potential target against consequently limit...
Corydalis bungeana Turcz. (CBT) is frequently used to treat inflammatory illnesses, the mechanisms underlying its use ulcerative colitis (UC) remain unclear. A dextran sulfate sodium (DSS)-induced UC mice model was established. The disease activity index (DAI), colonic length, histological inspection by hematoxylin-eosin staining, cytokines levels in colon, proteomics and intestinal flora were investigated evaluate effect of CBT. results showed that CBT can significantly reduce DAI, increase...
Increasing test collection sizes and limited judgment budgets create measurement challenges for IR batch evaluations, that are greater when using deep effectiveness metrics than shallow metrics, because of the increased likelihood unjudged documents will be encountered. Here we study problem metric score adjustment, with goal accurately estimating system performance sets, assuming dynamic adjustment is required per topic due to variability in number relevant documents. We seek induce...
Fields are a valuable auxiliary source of information in semi-structured HTML web documents. So, it is no surprise that ranking models have been designed to leverage this improve search effectiveness. We present the first (initial) study utilizing field-based relevance modeling framework. play two different, and integrated, roles our models: sources for inducing units on which applied ranking. Our preliminary results suggest can precision at top ranks; specifically, greater extent than...
To satisfy interaction between users and service providers (SP) in lots of rural areas without enough communication infrastructure, a dedicated return channel for the terrestrial digital TV is demanded. Therefore, we propose specification on Dedicated Return Channel (DRC) based SC-OFDMA technique future broadcast, promote it to Advanced Television Systems Committee (ATSC). efficiently utilize wireless resources uplink support large number wide coverage areas, scheduling among multiple most...
No abstract available.
The dominant retrieval models in information systems today are variants of TF×IDF, and typically use bag-of-words processing order to balance recall precision. However, the size collections continues increase, number results produced by these exceeds documents that can be reasonably assessed. To address this need, researchers commercial providers now looking at more expensive computational improve quality returned. One such method is incorporate term proximity into ranking model. We explore...